Greenhouse Gas Emissions and Fuel
Efficiency Standards for Medium- and
Heavy-Duty Engines and Vehicles -
Phase 2
Regulatory Impact Analysis
&EPA
United States
Environmental Protection
Agency
igNHTSA

-------

Greenhouse Gas Emissions and Fuel

Efficiency Standards for Medium- and

Heavy-Duty Engines and Vehicles -

Phase 2

Regulatory Impact Analysis

Office of Transportation and Air Quality
U.S. Environmental Protection Agency

And

National Highway Traffic Safety Administration
U.S. Department of Transportation
EPA-420-R-16-900

-------
TABLE OF CONTENTS
LIST OF ACRONYMS	ES-1
EXECUTIVE SUMMARY	ES-10
CHAPTER 1: INDUSTRY CHARACTERIZATION
1.1	Introduction	1-1
1.2	Trailers	1-1
1.3	Vocational Vehicles: Custom Chassis	1-8
CHAPTER 2: TECHNOLOGY AND COST
2.1	Overview of Technologies	2-1
2.2	Technology Principles - SI Engines	2-2
2.3	Technology Principles - CI Engines	2-9
2.4	Technology Principles - Class 4 to 8 Vehicles	2-19
2.5	Technology Application-HD Pickups and Vans	2-56
2.6	Technology Application-SI Engines	2-72
2.7	Technology Application and Estimated Costs - CI Engines	2-75
2.8	Technology Application and Estimated Costs - Tractors	2-94
2.9	Technology Application and Estimated Costs - Vocational Vehicles	2-148
2.10	Technology Application and Estimated Costs - Trailers	2-216
2.11	Technology Costs	2-265
2.12	Package Costs	2-350
CHAPTER 3: TEST PROCEDURES
3.1	Heavy-Duty Engine Test Procedure	3-1
3.2	Aerodynamic Assessment	3-5
3.3	Tire Rolling Resistance	3-47

-------
3.5	Tare Weights and Payload	3-71
3.6	Powertrain Test Procedures	3-74
3.7	Hybrid Powertrain Test Procedures	3-82
3.8	Axle Efficiency Test	3-85
3.9	Transmission Efficiency Test	3-86
3.10	HD Pickup Truck and Van Chassis Test Procedure	3-86
CHAPTER 4: VEHICLE SIMULATION MODEL
4.1	Purpose and Scope	4-1
4.2	Model Code Description	4-3
4.3	Validation of Phase 2 GEM Simulations	4-11
4.4	EPA and NHTSA HD Vehicle Compliance Model	4-27
4.5	Technology Improvements that Are Recognized in GEM without Simulation 4-42
CHAPTER 5: IMPACTS ON EMISSIONS AND FUEL CONSUMPTION
5.1	Executive Summary	5-1
5.2	Introduction	5-5
5.3	Program Analysis and Modeling Methods	5-6
5.4	Greenhouse Gas Emission and Fuel Consumption Impacts	5-26
5.5	Non-Greenhouse Gas Emission Impacts	5-42
CHAPTER 6: HEALTH AND ENVIRONMENTAL IMPACTS
6.1	Health and Environmental Effects of Non-GHG Pollutants	6-1
6.2	Impacts of the Rules on Concentrations of Non-GHG Pollutants	6-33
6.3	Changes in Atmospheric CO2 Concentrations, Global Mean Temperature,
Sea Level Rise, and Ocean pH Associated with the Program's GHG
Emissions Reductions	6-44

-------
CHAPTER 7: VEHICLE-RELATED COSTS, FUEL SAVINGS
7.1	Vehicle Costs, Fuel Savings and Maintenance Costs vs. the Dynamic Baseline
and Using Method A	7-1
7.2	Vehicle Costs, Fuel Savings and Maintenance Costs vs. the Flat Baseline
and using Method B	7-17
7.3	Key Parameters Used in the Estimation of Costs and Fuel Savings	7-48
CHAPTER 8: ECONOMIC AND OTHER IMPACTS
8.1	Framework for Benefits and Costs	8-1
8.2	Conceptual Framework for Evaluating Impacts	8-2
8.3	Analysis of the Rebound Effect	8-9
8.4	Impact on Class Shifting, Fleet Turnover, and Sales	8-22
8.5	Monetized GHG Impacts	8-26
8.6	Quantified and Monetized Non-GHG Health and Environmental Impacts 8-41
8.7	Additional Impacts	8-48
8.8	Petroleum, Energy and National Security Impacts	8-57
8.9	Summary of Benefits and Costs	8-71
8.10	Employment Impacts	8-78
8.11	Oil Price Sensitivity Analysis using Method B	8-90
APPENDIX 8.A TO CHAPTER 8 - SUPPLEMENTAL ANALYSIS OF
QUANTIFIED AND MONETIZED NON-GHG HEALTH AND
ENVIRONMENTAL IMPACTS	8-92
CHAPTER 9: SAFETY IMPACTS
9.1	Summary of Supporting HD Vehicle Safety Research	9-1
9.2	Safety Related Comments to the NPRM	9-9

-------
CHAPTER 10: CAFE MODEL FOR HD PICKUPS AND VANS
10.1	Overview of the CAFE Model	10-2
10.2	What Impacts Did NHTSA's "Method A" Analysis Show for Regulatory
Alternatives?	10-68
10.3	What Industry Impacts Did EPA's "Method B" Analysis Show for
Regulatory Alternatives?	10-119
CHAPTER 11: RESULTS OF THE PREFERRED AND ALTERNATIVE
11.1	What Are the Alternatives that the Agencies Considered?	11-1
11.2	How Do These Alternatives Compare in Overall GHG Emissions
Reductions and Fuel Efficiency?	11-19
CHAPTER 12: FINAL REGULATORY FLEXIBILITY ANALYSIS
12.1	Overview of the Regulatory Flexibility Act	12-1
12.2	Need for Rulemaking and Rulemaking Objectives	12-2
12.3	Definition and Description of Small Businesses	12-2
12.4	Summary of Small Entities to which the Rulemaking will Apply	12-3
12.5	Related Federal Rules	12-4
12.6	Projected Reporting, Recordkeeping, and Other Compliance Requirements 12-4
12.7	Regulatory Flexibilities	12-5
12.8	Projected Economic Effects of the Final Rulemaking	12-15
12.9	Summary of Economic Effects	12-19
CHAPTER 13: NATURAL GAS VEHICLES AND ENGINES
13.1	Detailed Lifecycle Analysis	13-1

-------

-------
Table of Contents, Acronym List, and Executive Summary
List of Acronyms
ng
Microgram
|im
Micrometers
2002$
U.S. Dollars in calendar year 2002
2009$
U.S. Dollars in calendar year 2009
A/C
Air Conditioning
ABS
Antilock Brake Systems
ABT
Averaging, Banking and Trading
AC
Alternating Current
ACES
Advanced Collaborative Emission Study
ALVW
Adjusted Loaded Vehicle Weight
AEO
Annual Energy Outlook
AES
Automatic Engine Shutdown
AHS
American Housing Survey
AMOC
Atlantic Meridional Overturning Circulation
AMT
Automated Manual Transmission
ANL
Argonne National Laboratory
APU
Auxiliary Power Unit
AQ
Air Quality
AQCD
Air Quality Criteria Document
AR4
Fourth Assessment Report
ARB
California Air Resources Board
ASL
Aggressive Shift Logic
ASPEN
Assessment System for Population Exposure Nationwide
AT
Automatic Transmissions
ATA
American Trucking Association
ATIS
Automated Tire Inflation System
ATRI
Alliance for Transportation Research Institute
AT SDR
Agency for Toxic Substances and Disease Registry
ATUS
American Time Use Survey
Avg
Average
BAC
Battery Air Conditioning
BenMAP
Benefits Mapping and Analysis Program
bhp
Brake Horsepower
bhp-hrs
Brake Horsepower Hours
BLS
Bureau of Labor Statistics
BSFC
Brake Specific Fuel Consumption
BTS
Bureau of Transportation Statistics
BTU
British Thermal Unit
CAA
Clean Air Act
CAAA
Clean Air Act Amendments

-------
CAD/CAE Computer Aided Design And Engineering
CAE	Computer Aided Engineering
CAFE	Corporate Average Fuel Economy
CARB	California Air Resources Board
CBI	Confidential Business Information
CCP	Coupled Cam Phasing
CCSP	Climate Change Science Program
Cd	Coefficient of Drag
CdA	Drag Area
CDC	Centers for Disease Control
CFD	Computational Fluid Dynamics
CFR	Code of Federal Regulations
CH4	Methane
CILCC	Combined International Local and Commuter Cycle
CITT	Chemical Industry Institute of Toxicology
CMAQ	Community Multiscale Air Quality
CO	Carbon Monoxide
CO2	Carbon Dioxide
CCheq	CO2 Equivalent
COFC	Container-on-Flatcar
COI	Cost of Illness
COPD	Chronic Obstructive Pulmonary Disease
CoV	Coefficient of Variation
CPS	Cam Profile Switching
CRC	Coordinating Research Council
CRGNSA	Columbia River Gorge National Scenic Area
CRR	Rolling Resistance Coefficient
CS	Climate Sensitivity
CSI	Cambridge Systematics Inc.
CSS	Coastal Sage Scrub
CSV	Comma-separated Values
CVD	Cardiovascular Disease
CVT	Continuously-Variable Transmission
CW	Curb Weight
D/UAF	Downward and Upward Adjustment Factor
DCP	Dual Cam Phasing
DCT	Dual Clutch Transmission
DE	Diesel Exhaust
DEAC	Cylinder Deactivation
DEER	Diesel Engine-Efficiency and Emissions Research
DEF	Diesel Exhaust Fluid
DHHS	U.S. Department of Health and Human Services
Diesel HAD Diesel Health Assessment Document

-------
DMC
DO
DOC
DOD
DOE
DOHC
DOT
DPF
DPM
DR
DRIA
DVVL
EC
EC
ECU
ED
EERA
EFR
EGR
EHPS
EIA
EISA
EMS-HAP
EO
EPA
EPS
ERG
ESC
EV
F
FEL
FET
FEV1
FHWA
FIA
FMCSA
FOH
FR
FTP
FVC
g
g/s
Direct Manufacturing Costs
Dissolved Oxygen
Diesel Oxidation Catalyst
Department of Defense
Department of Energy
Dual Overhead Camshaft Engines
Department of Transportation
Diesel Particulate Filter
Diesel Particulate Matter
Discount Rate
Draft Regulatory Impact Analysis
Discrete Variable Valve Lift
European Commission
Elemental Carbon
Electronic Control Unit
Emergency Department
Energy and Environmental Research Associates
Engine Friction Reduction
Exhaust Gas Recirculation
Electrohydraulic Power Steering
Energy Information Administration (part of the U.S. Department of
Energy)
Energy Independence and Security Act
Emissions Modeling System for Hazardous Air Pollution
Executive Order
Environmental Protection Agency
Electric Power Steering
Eastern Research Group
Electronic Stability Control
Electric Vehicle
Frequency
Family Emission Limit
Federal Excise Tax
Functional Expiratory Volume
Federal Highway Administration
Forest Inventory and Analysis
Federal Motor Carrier Safety Administration
Fuel Operated Heater
Federal Register
Federal Test Procedure
Forced Vital Capacity
Gram
Gram-per-second

-------
g/ton-mile	Grams emitted to move one ton (2000 pounds) of freight over one mile
gal	Gallon
gal/1000 ton- Gallons of fuel used to move one ton of payload (2,000 pounds) over
mile	1000 miles
GCAM	Global Change Assessment Model
GCW	Gross Combined Weight
GDP	Gross Domestic Product
GEM	Greenhouse gas Emissions Model
GEOS	Goddard Earth Observing System
GHG	Greenhouse Gases
GIFT	Geospatial Intermodal Freight Transportation
Greenhouse Gases, Regulated Emissions, and Energy Use in
GREET	Transportation
GSF1	Generic Speed Form one
GUI	Graphical User Interface
GVWR	Gross Vehicle Weight Rating
GWP	Global Warming Potential
HABs	Harmful Algal Blooms
HAD	Diesel Health Assessment Document
HC	Hydrocarbon
HD	Heavy-Duty
HDUDDS	Heavy Duty Urban Dynamometer Driving Cycle
HEG	High Efficiency Gearbox
HEI	Health Effects Institute
HES	Health Effects Subcommittee
HEV	Hybrid Electric Vehicle
HFC	Hydrofluorocarbon
HFET	Highway Fuel Economy Dynamometer Procedure
HHD	Heavy Heavy-Duty
HHDDT	Highway Heavy-Duty Diesel Transient
hp	Horsepower
hrs	Hours
HRV	Heart Rate Variability
HSC	High Speed Cruise Duty Cycle
HTUF	Hybrid Truck User Forum
hz	Hertz
IARC	International Agency for Research on Cancer
IATC	Improved Automatic Transmission Control
IC	Indirect Costs
ICCT	International Council on Clean Transport
ICD	International Classification of Diseases
ICF	ICF International
ICM	Indirect Cost Multiplier

-------
ICP
Intake Cam Phasing
IMAC
Improved Mobile Air Conditioning
IMPROVE
Interagency Monitoring of Protected Visual Environments
IPCC
Intergovernmental Panel on Climate Change
IRFA
Initial Regulatory Flexibility Analysis
IRIS
Integrated Risk Information System
ISA
Integrated Science Assessment
JAMA
Journal of the American Medical Association
k
Thousand
kg
Kilogram
KI
kinetic intensity
km
Kilometer
km/h
Kilometers per Hour
kW
Kilowatt
L
Liter
lb
Pound
LD
Light-Duty
LHD
Light Heavy-Duty
LLNL
Lawrence Livermore National Laboratory's
LRR
Lower Rolling Resistance
LSC
Low Speed Cruise Duty Cycle
LT
Light Trucks
LTCCS
Large Truck Crash Causation Study
LUB
Low Friction Lubes
LUC
Land Use Change
m2
Square Meters
m3
Cubic Meters
MAGIC C
Model for the Assessment of Greenhouse-gas Induced Climate Change
MCF
Mixed Conifer Forest
MD
Medium-Duty
MDPV
Medium-Duty Passenger Vehicles
mg
Milligram
MHD
Medium Heavy-Duty
MHEV
Mild Hybrid
mi
mile
min
Minute
MM
Million
MMBD
Million Barrels per Day
MMT
Million Metric Tons
MOVES
Motor Vehicle Emissions Simulator
mpg
Miles per Gallon
mph
Miles per Hour
MSAT
Mobile Source Air Toxic

-------
MRL
Minimal Risk Level
MT
Manual Transmission
MY
Model Year
N20
Nitrous Oxide
NA
Not Applicable
NAAQS
National Ambient Air Quality Standards
NAFA
National Association of Fleet Administrators
NAICS
North American Industry Classification System
NAS
National Academy of Sciences
NATA
National Air Toxic Assessment
NCAR
National Center for Atmospheric Research
NCI
National Cancer Institute
NCLAN
National Crop Loss Assessment Network
NEC
Net Energy Change Tolerance
NEI
National Emissions Inventory
NEMS
National Energy Modeling System
NEPA
National Environmental Policy Act
NESCAUM
Northeastern States for Coordinated Air Use Management
NESCCAF
Northeast States Center for a Clean Air Future
NESHAP
National Emissions Standards for Hazardous Air Pollutants
NHS
National Highway System
NHTSA
National Highway Traffic Safety Administration
NiMH
Nickel Metal-Hydride
NIOSH
National Institute of Occupational Safety and Health
Nm
Newton-meters
NMHC
Nonmethane Hydrocarbons
NMMAPS
National Morbidity, Mortality, and Air Pollution Study
NOx
Nitrogen Oxide
N02
Nitrogen Dioxide
NO A A
National Oceanic and Atmospheric Administration
NOx
Oxides of Nitrogen
NPRM
Notice of Proposed Rulemaking
NPV
Net Present Value
NRC
National Research Council
NRC-CAN
National Research Council of Canada
NREL
National Renewable Energy Laboratory
NTP
National Toxicology Program
NVH
Noise Vibration and Harshness
O&M
Operating and maintenance
03
Ozone
OAQPS
Office of Air Quality Planning and Standards
OC
Organic Carbon
OE
Original Equipment

-------
OEHHA	Office of Environmental Health Hazard Assessment
OEM	Original Equipment Manufacturer
OHV	Overhead Valve
OMB	Office of Management and Budget
OPEC	Organization of Petroleum Exporting Countries
ORD	EPA's Office of Research and Development
ORNL	Oak Ridge National Laboratory
OTAQ	Office of Transportation and Air Quality
Pa	Pascal
PAH	Polycyclic Aromatic Hydrocarbons
PEF	Peak Expiratory Flow
PEMS	Portable Emissions Monitoring System
PGM	Platinum Group Metal
PHEV	Plug-in Hybrid Electric Vehicles
PM	Particulate Matter
PM 10	Coarse Particulate IVIatter (diameter of 10 lim or less)
PM2.5	Fine Particulate Matter (diameter of 2.5 |im or less)
POM	Polycyclic Organic Matter
Ppb	Parts per Billion
Ppm	Parts per Million
Psi	Pounds per Square Inch
PTO	Power Take Off
R&D	Research and Development
RBM	Resisting Bending Moment
REL	Reference Exposure Level
RESS	Rechargeable Energy Storage System
RFA	Regulatory Flexibility Act
RfC	Reference Concentration
RFS2	Renewable Fuel Standard 2
RIA	Regulatory Impact Analysis
RPE	Retail Price Equivalent
Rpm	Revolutions per Minute
RSWT	Reduced-Scale Wind Tunnel
S	Second
SAB	Science Advisory Board
SAB-HES	Science Advisory Board - Health Effects Subcommittee
SAE	Society of Automotive Engineers
SAR	Second Assessment Report
SAV	Submerged Aquatic Vegetation
SBA	Small Business Administration
SBAR	Small Business Advocacy Review
SBREFA	Small Business Regulatory Enforcement Fairness Act
SCC	Social Cost of Carbon

-------
SCR
Selective Catalyst Reduction
SER
Small Entity Representation
SET
Supplemental Emission Test
SGDI
Stoichiometric Gasoline Direct Injection
SHEV
Strong Hybrid Vehicles
SI
Spark-Ignition
SIDI
Spark Ignition Direct Injection
S02
Sulfur Dioxide
SOx
Sulfur Oxides
SOA
Secondary Organic Aerosol
SOC
State of Charge
SOHC
Single Overhead Cam
SOx
Oxides of Sulfur
SPR
Strategic Petroleum Reserve
STB
Surface Transportation Board
Std.
Standard
STP
Scaled Tractive Power
SUV
Sport Utility Vehicle
SVOC
Semi-Volatile Organic Compound
SwRI
Southwest Research Institute
TAR
Technical Assessment Report
TC
Total Costs
TCp
Total Cost package
TDS
Turbocharging And Downsizing
THC
Total Hydrocarbon
TIAX
TIAX LLC
TMC
Technology & Maintenance Council
TOFC
Trailer-on-Flatcar
Ton-mile
One ton (2000 pounds) of payload over one mile
TPM
Tire Pressure Monitoring
TRBDS
Turbocharging and Downsizing
TRU
Trailer Refrigeration Unit
TSD
Technical Support Document
TSS
Thermal Storage
TTMA
Truck Trailer Manufacturers Association
TW
Test Weight
U/DAF
Upward and Downward Adjustment Factor
UCT
Urban Creep and Transient Duty Cycle
UFP
Ultra Fine Particles
LIRE
Unit Risk Estimate
USD A
United States Department of Agriculture
USGCRP
United States Global Change Research Program
UV
Ultraviolet

-------
UV-b
Ultraviolet-b
VHHD
Vocational Heavy Heavy-Duty
VIN
Vehicle Identification Number
VIUS
Vehicle Inventory Use Survey
VLHD
Vocational Light Heavy-Duty
VMHD
Vocational Medium Heavy-Duty
VMT
Vehicle Miles Traveled
VOC
Volatile Organic Compound
VSL
Vehicle Speed Limiter
VTRIS
Vehicle Travel Information System
VVL
Variable Valve Lift
VVT
Variable Valve Timing
WACAP
Western Airborne Contaminants Assessment Project
WBS
Wide Base Singles
WHR
Waste Heat Recovery
WHTC
World Harmonized Transient Cycle
WHVC
World Harmonized Vehicle Cycle
WRF
Weather Research Forecasting
WTP
Willingness-to-Pay
WTVC
World Wide Transient Vehicle Cycle
WVU
West Virginia University

-------
Executive Summary
The Environmental Protection Agency (EPA) and the National Highway Traffic Safety
Administration (NHTSA), on behalf of the Department of Transportation, are each finalizing
changes to our comprehensive Heavy-Duty National Program. The Program will further reduce
greenhouse gas emissions (GHG) and increase fuel efficiency for on-road heavy-duty vehicles,
responding to the President's directive on February 18, 2014, to take coordinated steps toward
the production of even cleaner vehicles. NHTSA's fuel consumption standards and EPA's
carbon dioxide (CO2) emissions standards are tailored to each of the three current regulatory
categories of heavy-duty vehicles: (1) Combination Tractors; (2) Heavy-duty Pickup Trucks and
Vans; and (3) Vocational Vehicles, as well as gasoline and diesel heavy-duty engines. In
addition, the agencies are adding new standards for combination trailers. EPA's
hydrofluorocarbon emissions standards that currently apply to air conditioning systems in
tractors, pickup trucks, and vans, will also be applied to vocational vehicles.
Table 1 presents the rule-related technology costs, maintenance costs, fuel savings, other
benefits, and net benefits in both present-value and annualized terms for Method A. This table
shows the costs and benefits relative to the dynamic baseline. Table 2 presents the rule-related
fuel savings, costs, benefits and net benefits in both present value terms and in annualized terms
as calculated for Method B relative to the flat baseline.

-------
Table 1 NHTSA's Estimated 2018-2029 Model Year Lifetime Discounted Costs,
Benefits, and Net Benefits using Method A, Relative to the Dynamic Baselinea, and
Assuming the 3% Discount Rate SC-GHG Values
(Billions of 2013 Dollars)
Lifetime Present Value - 3% Discount Rate
Vehicle Program
-$23.7
Maintenance
-$1.7
Fuel Savings
$149.1
Benefits (less costs by increased vehicle use)
$72.8
Net Benefitsb
$196.5
Annualized Value - 3% Discount Rate
Vehicle Program
-$0.9
Maintenance
-$0.1
Fuel Savings
$5.9
Benefits (less costs by increased vehicle use)
$2.9
Net Benefitsb
$7.8
Lifetime Present Value - 7% Discount Rate
Vehicle Program
-$16.1
Maintenance
-$0.9
Fuel Savings
$79.7
Benefits (less costs by increased vehicle use)
$54.6
Net Benefitsb
$117.3
Annualized Value - 7% Discount Rate
Vehicle Program
-$1.2
Maintenance
-$0.1
Fuel Savings
$5.8
Benefits (less costs by increased vehicle use)
$4.0
Net Benefitsb
$8.5
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for
an explanation of the flat baseline, la, and dynamic baseline, lb, please see Preamble
Section X.A.I
b Net benefits reflect the fuel savings plus benefits minus costs.

-------
Table 2 EPA's Estimated 2018-2029 Model Year Lifetime Discounted Costs, Benefits, and Net
Benefits using Method B and Relative to the Flat Baseline and Assuming the 3% Discount Rate SC-GHG
Values" (Billions of 2013 Dollars)
Lifetime Present Valuec - 3% Discount Rate
Vehicle Program
-$27
Maintenance
-$1.9
Fuel Savings
$169
Benefitsb
$88
Net Benefits'1
$229
Annualized Va
uee - 3% Discount Rate
Vehicle Program
-$1.4
Maintenance
-$0.1
Fuel Savings
$8.6
Benefitsb
$4.5
Net Benefits'1
$11.7
Lifetime Present Valuec - 7% Discount Rate
Vehicle Program
-$18
Maintenance
-$0.9
Fuel Savings
$87
Benefitsb
$62
Net Benefits'1
$131
Annualized Va
uee - 7% Discount Rate
Vehicle Program
-$1.4
Maintenance
-$0.1
Fuel Savings
$7.0
Benefitsb
$3.9
Net Benefits'1
$9.4
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an
explanation of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1
b EPA estimated the benefits associated with reductions in GHGs (CO2, CH4, and N20) using four
different values of a one ton reduction in each gas. The four values applied to each GHG are: model
average at 2.5% discount rate, 3%, and 5%; 95th percentile at 3% and each increases over time. For
the purposes of this overview presentation of estimated costs and benefits, however, the benefits
shown here use the central marginal value: the model average at 3% discount rate, in 2013 dollars.
Chapter 8.5 provides a complete list of values for the 4 estimates for each GHG. 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 (marginal values, i.e.
SC-GHGs, at 5, 3, and 2.5 percent) is used to calculate net present value of GHG benefits for
internal consistency. Refer to Section Chapter 8.5 for more detail.
0 Present value is the total, aggregated amount that a series of monetized costs or benefits that occur
over time is worth now (in year 2013 dollar terms), discounting future values, over the lifetime of
each model year vehicle, to calendar year 2015.
d Net benefits reflect the fuel savings plus benefits minus costs.
e The annualized value is the constant annual value through a 30 year lifetime whose summed
present value equals the present value from which it was derived. Annualized SC-GHG values are
calculated using the same rate as that used to determine the SC-GHG value, while all other costs
and benefits are annualized at either 3% or 7%.

-------
Table 3 Summary of Final 2021 Standards Including Average Per Vehicle Costs and Projected Improvement
REGULATORY
SUBCATEGORY
CO2 GRAMS
PER TON-
MILE
FUEL
CONSUMPTION
GALLON PER 1,000
TON-MILE
AVERAGE
INCREMENTAL
COST PER
VEHICLE
RELATIVE TO
PHASE 1 COSTS IN
MODEL YEAR
2021 A
AVERAGE
PERCENT FUEL
CONSUMPTION
AND CO2
IMPROVEMENT IN
MY 2021 RELATIVE
TO MY 2017
Tractors
Class 7 Low Roof Day Cab
105.5
10.36346
$5,134
11%
Class 7 Mid Roof Day Cab
113.2
11.11984
$5,134
11%
Class 7 High Roof Day Cab
113.5
11.14931
$5,240
12%
Class 8 Low Roof Day Cab
80.5
7.90766
$5,228
12%
Class 8 Mid Roof Day Cab
85.4
8.38900
$5,228
12%
Class 8 High Roof Day Cab
85.6
8.40864
$5,317
13%
Class 8 Low Roof Sleeper Cab
72.3
7.10216
$7,181
14%
Class 8 Mid Roof Sleeper Cab
78.0
7.66208
$7,175
14%
Class 8 High Roof Sleeper Cab
75.7
7.43615
$7,276
14%
Class 8
Heavy-Haul
52.4
5.14735
$5,063
8%
Trailers
Long Dry Box Trailer
78.9
7.75049
$1,081
5%
Short Dry Box Trailer
123.7
12.15128
$772
2%
Long Refrigerated Box Trailer
80.6
7.91749
$1,081
5%
Short Refrigerated Box Trailer
127.5
12.52456
$772
2%
Vocational Diesel
LHD Urban
424
41.6503
$1,106
12%
LHD Multi-Purpose
373
36.6405
$1,164
11%
LHD Regional
311
30.5501
$873
7%
MHD Urban
296
29.0766
$1,116
11%
MHD Multi-Purpose
265
26.0314
$1,146
10%
MHD Regional
234
22.9862
$851
6%
HELD Urban
308
30.2554
$1,334
9%
HELD Multi-Purpose
261
25.6385
$1,625
9%
HELD Regional
205
20.1375
$2,562
7%
Vocational Gasoline
LHD Urban
461
51.8735
$1,106
8%
LHD Multi-Purpose
407
45.7972
$1,164
8%
LHD Regional
335
37.6955
$873
6%
MHD Urban
328
36.9078
$1,116
7%
MHD Multi-Purpose
293
32.9695
$1,146
7%
MHD Regional
261
29.3687
$851
5%
Note:
a Engine costs are included in average vehicle costs. These costs are based on our projected market adoption rates
of various technologies and these costs include indirect costs via markups along with learning impacts. For a
description of the markups and learning impacts considered in this analysis and how it impacts technology costs for
other years, refer to Chapter 2 of the RIA (see RIA 2.11).

-------
Table 4 Summary of Final 2024 Standards Including Average Per Vehicle Costs and Projected Improvement
REGULATORY
SUBCATEGORY
CO2
GRAMS
PER TON-
MILE
FUEL
CONSUMPTION
GALLON PER
1,000 TON-MILE
AVERAGE
INCREMENTAL
COST PER VEHICLE
RELATIVE TO
PHASE 1 COSTS IN
MODEL YEAR 2024 A
AVERAGE PERCENT
FUEL CONSUMPTION
AND CO2
IMPROVEMENT IN
MY 2024 RELATIVE
TO MY 2017
Tractors
Class 7 Low Roof Day Cab
99.8
9.80354
$8,037
16%
Class 7 Mid Roof Day Cab
107.1
10.52063
$8,037
16%
Class 7 High Roof Day Cab
106.6
10.47151
$8,210
18%
Class 8 Low Roof Day Cab
76.2
7.48527
$8,201
17%
Class 8 Mid Roof Day Cab
80.9
7.94695
$8,201
16%
Class 8 High Roof Day Cab
80.4
7.89784
$8,358
18%
Class 8 Low Roof Sleeper Cab
68.0
6.67976
$11,100
19%
Class 8 Mid Roof Sleeper Cab
73.5
7.22004
$11,100
19%
Class 8 High Roof Sleeper Cab
70.7
6.94499
$11,306
19%
Class 8
Heavy-Haul
50.2
4.93124
$7,937
12%
Trailers
Long Dry Box Trailer
77.2
7.58350
$1,204
7%
Short Dry Box Trailer
120.9
11.87623
$1,171
4%
Long Refrigerated Box Trailer
78.9
7.75049
$1,204
7%
Short Refrigerated Box Trailer
124.7
12.24951
$1,171
4%
Vocational Diesel
LHD Urban
385
37.8193
$1,959
20%
LHD Multi-Purpose
344
33.7917
$2,018
18%
LHD Regional
296
29.0766
$1,272
11%
MHD Urban
271
26.6208
$2,082
18%
MHD Multi-Purpose
246
24.1650
$2,110
16%
MHD Regional
221
21.7092
$1,274
11%
HELD Urban
283
27.7996
$2,932
16%
HELD Multi-Purpose
242
23.7721
$3,813
16%
HELD Regional
194
19.0570
$4,009
12%
Vocational Gasoline
LHD Urban
432
48.6103
$1,959
13%
LHD Multi-Purpose
385
43.3217
$2,018
9%
LHD Regional
324
36.4577
$1,272
12%
MHD Urban
310
34.8824
$2,082
11%
MHD Multi-Purpose
279
31.3942
$2,110
9%
MHD Regional
251
28.2435
$1,274
13%
Note:
a Engine costs are included in average vehicle costs. These costs are based on our projected market adoption rates
of various technologies and these costs include indirect costs via markups along with learning impacts. For a
description of the markups and learning impacts considered in this analysis and how it impacts technology costs for
other years, refer to Chapter 2 of the RIA (see RIA 2.11).

-------
Table 5 Summary of Final 2027 Standards Including Average Per Vehicle Costs and Projected Improvement
REGULATORY
CO2 GRAMS
FUEL
AVERAGE
AVERAGE PERCENT
SUBCATEGORY
PER TON-
CONSUMPTION
INCREMENTAL COST
FUEL

MILE (FOR
GALLON PER 1,000
PER VEHICLE
CONSUMPTION

HDPUV,
TON-MILE (FOR
RELATIVE TO PHASE
AND CO2

GRAMS PER
HDPUV,
1 COSTS IN MODEL
IMPROVEMENT IN

MILE)
GALLONS PER 100
MILES)
YEAR 2027 A
MY 2027 RELATIVE
TO MY 2017
Tractors
Class 7 Low Roof Day Cab
96.2
9.44990
$10,235
19%
Class 7 Mid Roof Day Cab
103.4
10.15717
$10,235
19%
Class 7 High Roof Day Cab
100.0
9.82318
$10,298
21%
Class 8 Low Roof Day Cab
73.4
7.21022
$10,439
20%
Class 8 Mid Roof Day Cab
78.0
7.66208
$10,439
19%
Class 8 High Roof Day Cab
75.7
7.43615
$10,483
22%
Class 8 Low Roof Sleeper Cab
64.1
6.29666
$13,535
24%
Class 8 Mid Roof Sleeper Cab
69.6
6.83694
$13,574
23%
Class 8 High Roof Sleeper Cab
64.3
6.31631
$13,749
25%
Class 8
Heavy-Haul
48.3
4.74460
$9,986
15%
Trailers
Long Dry Box Trailer
75.7
7.43615
$1,370
9%
Short Dry Box Trailer
119.4
11.72888
$1,204
6%
Long Refrigerated Box Trailer
77.4
7.60314
$1,370
9%
Short Refrigerated Box Trailer
123.2
12.10216
$1,204
5%
Vocational Diesel
LHD Urban
367
36.0511
$2,533
24%
LHD Multi-Purpose
330
32.4165
$2,571
21%
LHD Regional
291
28.5855
$1,486
13%
MHD Urban
258
25.3438
$2,727
22%
MHD Multi-Purpose
235
23.0845
$2,771
20%
MHD Regional
218
21.4145
$1,500
12%
HHD Urban
269
26.4244
$4,151
20%
HHD Multi-Purpose
230
22.5933
$5,025
20%
HHD Regional
189
18.5658
$5,670
14%
Vocational Gasoline
LHD Urban
413
46.4724
$2,533
18%
LHD Multi-Purpose
372
41.8589
$2,571
16%
LHD Regional
319
35.8951
$1,486
11%
MHD Urban
297
33.4196
$2,727
16%
MHD Multi-Purpose
268
30.1564
$2,771
15%
MHD Regional
247
27.7934
$1,500
10%
Class 2b and 3 HD Pickups and Vansb
HD Pickup and Van
460
4.88
$1,486
17%
Notes:
" Engine costs are included in average vehicle costs. These costs are based on our projected market adoption rates
of various technologies and these costs include indirect costs via markups along with learning impacts. For a
description of the markups and learning impacts considered in this analysis and how it impacts technology costs for
other years, refer to Chapter 2 of the RIA (see RIA 2.11).
b For HD pickups and vans, Table 5 shows results for MY2029, assuming continuation of MY2027 standard.

-------
Table 6 Summary of Final 2021 and 2024 Custom Chassis Vocational Standards Including Average Per
Vehicle Costs and Projected Improvement
REGULATORY
SUBCATEGORY
CO2 GRAMS PER
TON-MILE
FUEL
CONSUMPTION
GALLON PER 1,000
TON-MILE
AVERAGE
INCREMENTAL
COST PER
VEHICLE
RELATIVE TO
PHASE 1 COSTS IN
MODEL YEAR
2021 A
AVERAGE
PERCENT FUEL
CONSUMPTION
AND CO2
IMPROVEMENT IN
MY 2021 RELATIVE
TO MY 2017
Vocational Custom Chassis
Coach Bus
210
20.6287
900
7%
Motor Home
228
22.3969
600
6%
School Bus
291
28.5855
800
10%
Transit
300
29.4695
1000
7%
Refuse
313
30.7466
700
4%
Mixer
319
31.3360
300
3%
Emergency
324
31.8271
400
1%
Note:
a Engine costs are included in average vehicle costs. These costs are based on our projected market adoption rates
of various technologies and these costs include indirect costs via markups along with learning impacts. For a
description of the markups and learning impacts considered in this analysis and how it impacts technology costs for
other years, refer to Chapter 2 of the RIA (see RIA 2.11).
Table 7 Summary of Final 2027 Custom Chassis Vocational Standards Including Average Per Vehicle Costs
and Projected Improvement
REGULATORY
SUBCATEGORY
CO2 GRAMS PER
TON-MILE
FUEL
CONSUMPTION
GALLON PER 1,000
TON-MILE
AVERAGE
INCREMENTAL
COST PER
VEHICLE
RELATIVE TO
PHASE 1 COSTS IN
MODEL YEAR
2027 A
AVERAGE
PERCENT FUEL
CONSUMPTION
AND CO2
IMPROVEMENT IN
MY 2027 RELATIVE
TO MY 2017
Vocational Custom Chassis
Coach Bus
205
20.1375
1400
11%
Motor Home
226
22.2004
900
9%
School Bus
271
26.6208
1300
18%
Transit
286
28.0943
1800
14%
Refuse
298
29.2731
1300
12%
Mixer
316
31.0413
600
7%
Emergency
319
31.3360
600
6%
Note:
a Engine costs are included in average vehicle costs. These costs are based on our projected market adoption rates
of various technologies and these costs include indirect costs via markups along with learning impacts. For a
description of the markups and learning impacts considered in this analysis and how it impacts technology costs for
other years, refer to Chapter 2 of the RIA (see RIA 2.11).

-------
This Regulatory Impact Analysis (RIA) provides detailed supporting documentation to
EPA and NHTSA joint rules under each of their respective statutory authorities. Because there
are slightly different requirements and flexibilities in the two authorizing statutes, this RIA
provides documentation for the primary joint provisions as well as for provisions specific to each
agency.
This RIA is generally organized to provide overall background information,
methodologies, and data inputs, followed by results of the various technical and economic
analyses. A summary of each chapter of the RIA follows.
Chapter 1: Industry Characterization. In order to assess the impacts of greenhouse gas
(GHG) and fuel consumption regulations upon the affected industries, it is important to
understand the nature of the industries impacted by the regulations. This chapter provides
market information for the trailer industry, as well as the variety of ownership patterns, for
background purposes. It also provides information on the vocational vehicle industry.
Chapter 2: Technology and Cost. This chapter presents details of the vehicle and
engine technologies and technology packages for reducing greenhouse gas emissions and fuel
consumption. These technologies and technology packages represent potential ways that the
industry could meet the CO2 and fuel consumption stringency levels, and they provide the basis
for the technology costs and effectiveness analyses.
Chapter 3: Test Procedures. Laboratory procedures to physically test engines, vehicles,
and components are a crucial aspect of the heavy-duty vehicle GHG and fuel consumption
program. The rulemaking will establish some new test procedures for both engine and vehicle
compliance and will revise existing procedures. This chapter describes the relevant test
procedures, including methodologies for assessing engine emission performance, the effects of
aerodynamics and tire rolling resistance, as well as procedures for chassis dynamometer testing
and their associated drive cycles.
Chapter 4: Vehicle Simulation Model. An important aspect of a regulatory program is
its ability to accurately estimate the potential environmental benefits of heavy-duty truck
technologies through testing and analysis. Most large truck manufacturers employ various
computer simulation methods to estimate truck efficiency for purposes of developing and
refining their products. Each method has advantages and disadvantages. This section will focus
on the use of a type truck simulation modeling that the agencies have developed specifically for
assessing tailpipe GHG emissions and fuel consumption for purposes of this rulemaking. The
agencies will revise the existing simulation model — the "Greenhouse gas Emissions Model
(GEM)" — as the primary tool to certify vocational vehicles, combination tractor, and
combination trailers, Class 2b through Class 8 heavy-duty vehicles that are not heavy-duty
pickups or vans) and discuss the model in this chapter.
Chapter 5: Impacts on Emissions and Fuel Consumption. This program estimates
anticipated impacts from the CO2 emission and fuel efficiency standards. The agencies quantify
fuel use and emissions from the GHGs carbon dioxide (CO2), methane (CH4), nitrous oxide
(N2O) and hydrofluorocarbons (HFCs). In addition to reducing the emissions of greenhouse
gases and fuel consumption, this program will also influence the emissions of "criteria" air

-------
pollutants, including carbon monoxide (CO), fine particulate matter (PM2.5) 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), as
described further in Chapter 5.
The agencies used EPA's Motor Vehicle Emission Simulator (MOVES2014a) to estimate
downstream (tailpipe) emission impacts for combination tractors and vocational vehicles, and a
spreadsheet model based on emission factors the "GREET" model to estimate upstream (fuel
production and distribution) emission changes resulting from the decreased fuel. For HD
pickups and vans, the agencies used DOT's CAFE model to estimate manufacturer responses to
these standards. NHTSA used the CAFE model to estimate emission impacts, and EPA used the
MOVES model to calculate emission impacts using CAFE model technology penetration outputs
as an input. Based on these analyses, the agencies estimate that this program will lead to 199.2
million metric tons (MMT) of CO2 equivalent (CO2EQ) of annual GHG reduction and 14.9
billion gallons of fuel savings in the year 2050, as discussed in more detail in Chapter 5.
Chapter 6: Health and Environmental Impacts. This chapter discusses 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 will not be
directly regulated by the standards, but the standards will affect emissions of these pollutants and
precursors. Reductions in these pollutants are the co-benefits of the rulemaking (that is, benefits
in addition to the benefits of reduced GHGs). This section discusses current and projected
concentrations of non-GHG pollutants as well as the air quality modeling methodology and
modeled projected impacts of this rule. This chapter also discusses GHG-related impacts, such
as changes in atmospheric CO2 concentrations, global mean temperature, sea level rise, and
ocean pH associated with the program's GHG emissions reductions.
Chapter 7: Vehicle-Related Costs of the Program. In this chapter, the agencies present
our estimate of the costs associated with the program. The presentation summarizes the costs
associated with new technology expected to be added to meet the GHG and fuel consumption
standards, including hardware costs to comply with the air conditioning (A/C) leakage program.
The analysis discussed in Chapter 7 provides our best estimates of incremental costs on a per
truck basis and on an annual total basis. We also present the fuel savings and maintenance costs
in this chapter, along with a detailed payback analysis for various vehicle segments.
Chapter 8: EPA's Economic and Other Impacts Analysis. This chapter provides EPA's
description of the net benefits of the HD National Program. To reach these conclusions, the
chapter discusses each of the following aspects of the analyses of benefits:
Rebound Effect. The VMT rebound effect refers to the fraction of fuel savings expected
to result from an increase in fuel efficiency that is offset by additional vehicle use.
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.

-------
Monetized GHG Impacts: The agencies estimate the monetized benefits of GHG
reductions by assigning a dollar value to reductions in GHG emissions using recent estimates of
the social cost of greenhouse gasses (SC-GHG). The SC-GHG is an estimate of the monetized
damages associated with an incremental increase in greenhouse gas emissions in a given year.
Other Impacts: There are other impacts associated with the GHG emissions and fuel
efficiency standards. Lower fuel consumption will, presumably, result in fewer gallons being
refilled and, thus, less time spent refueling. The increase in vehicle-miles driven due to a
positive rebound effect may also increase the societal costs associated with traffic congestion,
crashes, and noise. However, if drivers drive those additional rebound miles, there must be a
value to them which we estimate as the value of increased travel. The agencies also discuss the
impacts of safety standards and voluntary safety improvements on vehicle weight.
Chapter 8 also presents a summary of the total costs, total benefits, and net benefits
expected under the program.
Chapter 9: NHTSA and EPA considered the potential safety impact of technologies that
improve HD vehicle fuel efficiency and GHG emissions as part of the assessment of regulatory
alternatives. This chapter discusses the literature and research considered by the agencies, which
included two National Academies of Science reports, an analysis of safety effects of HD pickups
and vans using estimates from the DOT report on the effect of mass reduction and vehicle size
on safety, and agency-sponsored safety testing and research.
Chapter 10: NHTSA CAFE Model. This chapter describes NHTSA's CAFE modeling
system. The agencies used DOT's CAFE model to estimate manufacturer responses to these
standards for HD pickups and vans, and NHTSA also used the CAFE model to estimate emission
impacts for this sector.
Chapter 11: Results of Preferred and Alternative Standards. The heavy-duty truck
segment is very complex. The sector consists of a diverse group of impacted parties, including
engine manufacturers, chassis manufacturers, truck manufacturers, trailer manufacturers, truck
fleet owners and the public. The agencies have largely designed this program to maximize the
environmental and fuel savings benefits, taking into account the unique and varied nature of the
regulated industries. In developing this program, we considered a number of alternatives that
could have resulted in fewer or potentially greater GHG and fuel consumption reductions.
Chapter 9 Section summarizes the alternatives we considered.
Chapter 12: Small Business Flexibility Analysis. This chapter describes the agencies'
analysis of the small business impacts due to the joint program.
Chapter 13: Natural Gas Vehicles and Engines. This chapter describes EPA's lifecycle
analysis for natural gas used by the heavy-duty truck sector.

-------
Chapter 1: Industry Characterization
1.1	Introduction
The fuel consumption and CO2 emissions standards described in the Preamble of this
FRM will be applicable to three currently-regulated categories of heavy-duty vehicles: (1)
Combination Tractors; (2) Heavy-duty Pickup Trucks and Vans; and (3) Vocational Vehicles, as
well as spark-ignition and compression-ignition heavy-duty engines. The industry
characterization for these sectors can be found in the RIA for the HD Phase 1 rulemaking.1 With
this rulemaking, the agencies will be setting standards for combination trailers for the first time.
Also with this rulemaking, the agencies are setting standards that apply for small businesses for
the first time, as well as offering separate standards for vocational custom chassis. The
characterization laid out in this chapter focuses on trailers and vocational custom chassis,
whereas Chapter 12 of this RIA highlights impacts related to small businesses.
1.2	Trailers
A trailer is a vehicle designed to haul cargo while being pulled by another powered motor
vehicle. The most common configuration of large freight trucks consists of a Class 7 or 8 tractor
hauling one or more trailers. Vehicles in these configurations are called "combination tractor-
trailers" or simply "tractor-trailers." A trailer may be constructed to rest upon the tractor that
tows it, or be constructed so part of its weight rests on an auxiliary front axle called a "converter
dolly" between two or more trailers. Trailers are attached to tractors by a coupling pin (or king
pin) on the front of the trailer and a horseshoe-shaped coupling device called a fifth wheel on the
rear of the towing vehicle or on the converter dolly. A tractor can also pull international
shipping or domestic containers mounted on open-frame chassis, which when driven together on
the road function as trailers.
The Truck Trailer Manufacturers Association, an industry trade group primarily for
manufacturers of Class 7 and 8 truck trailers, offers publications of recommended practices,
technical bulletins and manuals that cover many aspects of trailer manufacture, and serves as a
liaison between the industry and government agencies.2 To date, federal regulations for the
trailer industry are limited to those issued by the Department of Transportation (See 49 CFR).
These regulations govern trailer dimensions and weight, as well as trailer safety requirements
(e.g., lights, reflective materials, bumpers, etc.). In addition, DOT requires that each trailer, like
other on-road vehicles, must have a Vehicle Identification Number (VIN).3 The VIN is
displayed on a label that is permanently-affixed to the trailer. It is required to contain the
manufacturer identification, make and type of vehicle, model year, type of trailer, body type,
length, and axle configuration. Trailer manufactures are responsible for reporting each trailer's
VIN information to NHTSA prior to the sale of the trailer.
1.2.1 Trailer Types
Class 7 and 8 tractors haul a diverse range of trailer types. The most common trailer type
is the box trailer, which is enclosed and can haul most types of mixed freight. The general
rectangular shape of these trailers allows operators to maximize freight volume within the

-------
regulated dimensional limits, since the majority of freight shipped by truck cubes-out (is volume-
limited) before it grosses-out (is weight-limited). Despite considerable improvements in
suspension, material, safety, durability, and other advancements, the basic shape of the box
trailer has not changed much over the past decades, although its dimensions have increased
incrementally from what used to be the industry's standard length of 40' to today's standard 53'
long van trailer. Today, box vans are commonly found in lengths of 28', 48', and 53'and widths
of 102" or 96." The 28' vans ("pups") are often driven in tandem and connected by a dolly.
Current length restrictions for the total combination tractor-trailer vehicle limit tandem operation
to 28' trailers. However, some members of the trucking industry are pushing to increase the
length limits to allow trailers as long as 33' to be pulled in tandem, and arguing that these "less
than truckload" (LTL) operations could increase capacity per truckload, reduce the number of
trucks on the road, reduce the fuel consumption and emissions of these tractor-trailers, and
remain within the current weight limits.4,5
Trailers are often highly customized for each order. The general structure of the box
trailer type is common and consists of vertical support posts in the interior of the trailer covered
by a smooth exterior surface. However the exterior of the trailer may be constructed of
aluminum or a range of composite materials. Historically, floors were constructed of wood,
however many trailer customers are requesting aluminum floors to reduce weight. Semi-trailer
axles are commonly a dual tandem configuration, but can also be single, spread tandem (i.e., two
axles separated to maximize axle loads), tridem (i.e., three axles equally spaced), tri axles (i.e.,
three axles consisting of a tandem and a third axle that may be liftable), or multi-axles to
distribute very heavy loads. Axles can be fixed in place, or allowed to slide to adjust weight
distribution. Doors are commonly located at the rear of the trailer. The most common door is
the side-by-side configuration, in which each door opens outward. Roll-up doors, which are
more costly, allow truck drivers to pull up to loading docks without first stopping to open the
doors. Roll-up doors are common on trailers with temperature-sensitive freight. Additional
variations in trailers include side-access doors, or use the underside of the trailer for belly boxes
or to store on-demand items such as ladders or spare tires.
The most common box trailer is the standard dry van, which transports cargo that does
not require special environmental conditions. In addition to the standard rectangular shape, dry
vans come in several specialty variants, such as drop floor, expandable, and curtain-side.
Another type of specialty box trailer is the refrigerated van trailer (reefer). This is an enclosed,
insulated trailer that hauls temperature sensitive freight, with a transportation refrigeration unit
(TRU) or heating unit mounted in the front of the trailer powered by a small (9-36 hp) diesel
engine. Figure 1-1 shows an example of the standard dry and refrigerated van.

-------
Adapted from http://www.wbmcguire.com/links/Guides/TruckTrailerGuide.pdf
Figure 1-1 Example of Dry and Refrigerated Van
Many other trailer types are uniquely designed to transport a specific type of freight.
Platform trailers carry cargo that may not be easily contained within or loaded and unloaded into
a box trailer, such as large, non-uniform equipment or machine components. Platforms come in
different configurations including standard flatbed, gooseneck, and drop deck. Tank trailers are
pressure-tight enclosures designed to carry liquids, gases or bulk, dry solids and semi-solids.
Tank trailers are generally constructed of steel or aluminum. The plumbing for intake and
discharge of the contents could be located below the tank or at the rear. There are also a number
of other specialized trailers such as grain (with and without hoppers), dump (frameless, framed,
bottom dump, demolition), automobile hauler (open or enclosed), livestock trailers (belly or
straight), construction and heavy-hauling trailers (tilt bed, hydraulic).
A sizable fraction of U.S. freight is transported in large, steel containers both
internationally via ocean-going vessels and domestically via rail cars. Containers are constructed
with steel sidewalls and external support beams, which results in a corrugated exterior. These
containers haul mixed freight and are designed with similar dimensions to box trailers. Ocean-
going international shipping containers are typically 20-feet or 40-feet in length. Domestic
containers, which often travel by rail, are 53-feet in length. Transport of these containers from
ports or rail to their final destination requires the container to be loaded on a specialty piece of
equipment called a chassis. The chassis, which is attached to the fifth wheel of a Class 7 or 8
tractor, consists of a frame, axles, suspension, brakes and wheel assemblies, as well as lamps,
bumpers and other required safety components. Fixed chassis vary in length according to the
type of container that will be attached, though some chassis adjust to accommodate different
sizes. When the chassis and container are assembled the unit serves the same function as a road
trailer.6 However, under customs regulations, the container itself is not considered part of a road
vehicle.7
ACT Research compiles factory shipment information from a Trailer Industry Control
Group that represents 80 percent of the U.S. trailer industry. Figure 1-2 shows the distribution of
trailers sold in the U.S. based on ACT Research's 2013 factory shipment data. The most
common type of trailer in use today is the dry van trailer, followed by the refrigerated van.
Together, these box vans make up greater than 70 percent of the industry. Trailer Body

-------
Builders' annual trailer output report estimates there were over 240,000 trailers sold in North
America 2013.
ACT Research 2013 Factory Shipments
4% 1 %
¦	Dry Van
¦	Refrigerated Van
¦	Platform
¦	HeavyLowbed
¦	MediumLowbed
¦	Dump
¦TankLiquid
¦Tank Bulk
Grain
Other Trailer
Chassis
Figure 1-2 ACT Research's 2013 U.S. factory shipments
1.2.2 Trailer Manufacturers
The diverse van, platform, tank and specialty trailers are produced by a large number of
trailer manufacturers. EPA estimates there are 178 trailer manufacturers. Trailers are far less
mechanically complex than the tractors that haul them, and much of trailer manufacturing is
done by hand. This relatively low barrier to entry for trailer manufacturing accounts, in part, for
the large number of trailer manufacturers. Figure 1-3 shows the production distribution of the
industry for the top 28 companies.8 While the percentages and ranking vary slightly year-to-
year, the top five manufacturers consistently produce over 70 percent of the manufacturing
output of the industry.
Trailer Body Builders 2015 North American Truck Trailer Output Report
(339,948 Total Trailers)
¦	Wabash National Corporation
¦	Great Dane Limited Partnership
¦	Hyundai Translead
¦	Utility Trailer Manufacturing
¦	Vanguard National Trailer/CIMC
¦	Stoughton Trailers LLC*
¦	Manac
¦	Fontaine Trailer Company
¦	Wilson Trailer Company
¦	MAC Trailer Manufacturing
¦	Heil Trailer International, Co.
s Strick Corporation*
Pitts Enterprises*
Timpte Inc.*
Reitnouer Inc.*
Next 13 Companies (None > 1.0%) **
Small business according to SBA definition of <1000 employees
* 9 of 13 are small businesses
Figure 1-3 2015 Trailer Output Report from Trailer Body Builders

-------
Table 1-1 illustrates the varying revenue among trailer manufacturers and further
distinguishes the very different roles in that market played by small and large manufacturers.
The revenue numbers were obtained from Hoovers online company database.9 Over 80 percent
of trailer manufacturers meet the Small Business Administration's (SB A) definition of a small
business (i.e., less than 1,000 employees), yet these manufacturers make up less than 25 percent
of the overall revenue from the industry. In fact, a majority of the small business trailer
manufacturers make less than $10 million in revenue per year.
Table 1-1 Summary of 2014 Trailer Industry Revenue by Business Size
REVENUE RANGE
BUSINESS SIZE
All Sizes
Large
Small3
> 1000M
3
3
0
$500M - $999M
2
2
0
$400M - $499M
1
1
0
$300M - $399M
3
3
0
$200M - $299M
5
4
1
$100M - $199M
3
1
2
$50M - $99M
14
6
8
$40M - $49M
22
2
20
$15M - $19M
8
0
8
$10M - $14M
17
3
14
$5M - $9M
35
4
31
< $5M
65
2
63




Total Companies
178
31
147
Total Revenue ($M)
10841
8543
2298
Average Revenue ($M)
61
276
16




Box Trailer Mfrs
13
8
5
Non-Box Trailer Mfrs
173
29
144
Note:
a The Small Business Administration (SB A) defines a trailer
manufacturer as a "small business" if it has fewer than 1,000
employees
The trailer industry was particularly hard hit by the recent recession. Trailer
manufacturers saw deep declines in new trailer sales of 46 percent in 2009; some trailer
manufacturers saw sales drop as much as 71 percent. This followed overall trailer industry
declines of over 30 percent in 2008. The 30 largest trailer manufacturers saw sales decline 72
percent from 282,750 in 2006, to only 78,526 in 2009. Several trailer manufacturers shut down
entire production facilities and a few went out of business altogether. Trailer production has
steadily grown across the industry since 2010 and, although historic production peaks have not
been repeated to date, it has now returned to levels close to those seen in the mid-2000s. Figure
1-4 shows the ACT Research's annual factory shipments, which illustrates the unsteady
production over the past 17 years. Trailer Body Builders' annual trailer output report estimates
there were over 240,000 trailers sold in North America in 2013. Output increased to 292,000 in
2014 and to nearly 340,000 in 2015 (very close to the current record from 1999).

-------
ACT Research Annual Factory Shipments
400000 -|
350000 -
300000 -
250000 -
200000 -
150000
5 ~ 100000
W
c
QJ -
fc Q)
Q. Q.
I/) "3
^ 0)
5 TO
u £
(0
50000
CD N- CO O)
0)0)0)0)
O) O) O) O)
O (M CO lO CD
oooooooo
oooooooo
00 O) o
o O T— -rr
O O O o o o
(NfM(M(NCM(NCM(N(N(N(NtNCNjf\|
Figure 1-4 Annual Factory Shipments Tracked by ACT Research
1.2.3 Trailer Use
In order to determine the appropriate tractor type for each trailer, the agencies
investigated "primary trip length" results from the Vehicle Inventory and Use Survey database
to determine the distribution of trailers in short- and long-haul applications.10 Using a primary
trip length of 500 miles or less to represent short-haul use, the agencies found that, of the
reported vehicles, over 50 percent of the 53-feet and longer dry vans were used in long-haul and
over 80 percent of the shorter vans were used in short-haul applications. Over 70 percent of the
reported 53-feet and longer refrigerated vans were long-haul trailers, with 65 percent of the
shorter refrigerated vans used in short-haul applications. The survey found that non-box trailers
are most frequently used for short-haul. Figure 1-5 summarizes these findings.
U.S. Census 2002 Vehicle Inventory and Use Survey
Tractor-Trailer Primary Trip Length
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
53'+ Dry Van 53'+ Reefer <53' Dry Van <53' Reefer
Flatbed, Tank {Dry Bulk) Tank (Liquid,
Platform,	Gases)
Curtainside, etc
¦ Short-Haul (<500 mi) ¦Long-Haul
Figure 1-5 2002 Vehicle Inventory and Use Survey Considering Primary Trip Length for Tractor-Trailers

-------
Truck drivers and trucking fleets frequently do not control all or even any of the trailers
that they haul. Trailers can be owned by freight customers, large equipment leasing companies,
third party logistics companies, and even other trucking companies. Containers on chassis,
which function as trailers, are rarely owned by truck operators. Rather, they are owned or leased
by ocean-going shipping companies, port authorities or others. This distinction between who
hauls the freight and who owns the equipment in which it is hauled means that truck owners and
operators have limited ability to be selective about the trailers they carry, and very little incentive
or ability to take steps to reduce the fuel use of trailers that they neither own or control.
For refrigerated trailers, the story is slightly different. These trailers are used more
intensely and accumulate more annual miles than other trailers. Over time, refrigerated trailers
can also develop problems that interfere with their ability to keep freight temperature-controlled.
For example, the insulating material inside a refrigerated trailer's walls can gradually lose its
thermal capabilities due to aging or damage from forklift punctures. The door seals on a
refrigerated trailer can also become damaged or loose with age, which greatly affects the
insulation characteristics of the trailer, similar to how the door seal on a home refrigerator can
reduce the efficiency of that appliance. As a result of age-related problems and more intense
usage, refrigerated trailers tend to have shorter procurement cycles than dry van trailers, which
means a faster turnover rate, although still not nearly as fast as for trucks in their first use.
Tractor-trailers are often used in conjunction with other modes of transportation (e.g.,
shipping and rail) to move goods across the country, known as intermodal shipping. Intermodal
traffic typically begins with containers carried on ships, and then they are loaded onto railcars,
and finally transported to their end destination via truck. Trucks that are used in intermodal
applications are of two primary types. Trailer-on-flatcar (TOFC) involves lifting the entire
trailer or the container attached to its chassis onto the railcar. In container-on-flatcar (COFC)
applications, the container is removed from the chassis and placed directly on the railcar. The
use of TOFCs allows for faster transition from rail to truck, but is more difficult to stack on a
vessel; therefore the use of COFCs has been increasing steadily. Both applications are used
throughout the U.S. with the largest usage found on routes between West Coast ports and
Chicago, and between Chicago and New York.
1.2.4 Trailer Fleet Size Relative to the Tractor Fleet
In 2013, over 800,000 trailers were owned by for-hire fleets and almost 300,000 were
owned by private fleets. Trailers that are purchased by fleets are typically kept much longer than
are the tractors, so trucks and trailers have different purchasing cycles. Also, many trailers are
owned by shippers or by leasing companies, not by the trucking fleets. Because of the
disconnect between owners and operators, the trailer owners may not benefit directly from fuel
consumption and GHG emission reductions.
The industry generally recognizes that the ratio of the number of dry van trailers in the
fleet relative to the number of tractors is typically three-to-one.11 Typically at any one time, two
trailers are parked while one is being transported. Certain private fleets may have ratios as high
as six-to-one and owner-operators may have a single trailer for their tractor. The ratio of
refrigerated vans to tractors is closer to two-to-one. This is partly due to the fact that it is more
expensive to purchase and operate refrigerated vans compared to dry vans. Specialty trailers,

-------
such as tanks and flatbeds are often attached to a single trailer throughout much of their life.
This characteristic of the trailer fleet impacts the cost effectiveness of trailer technologies. The
annual savings achieved due to these technologies are proportional to the number of miles
traveled in a year and the analysis for many of the trailers must account for some amount of
inactivity, which will reduce the benefits.
1.3 Vocational Vehicles: Custom Chassis
Based on public comments, information on entities who have certified, and stakeholder
outreach, we have deepened our understanding of the vocational vehicle market, including the
nature of specialization vs diversification among vocational vehicle manufacturers. We have
identified seven vocations as shown in Table 1-2, for which there are manufacturers who are not
diversified in their products competing for sales with diversified manufacturers. We are calling
these custom chassis in this rulemaking.
Table 1-2 Diversification of Vocational Chassis Manufacturers"
Vehicle Type
Number of Single-type
Chassis Manufacturers
Number of Multiple-type
Chassis Manufacturers
Coach (Intercity) Bus
2
3
Motor Home
3
8
School Bus
1
2
Transit Bus
4
4
Refuse Truck
1
6
Cement Mixer
2
7
Emergency Vehicle
6
7
Note:
a Includes U.S.-made vehicles and those imported for sale in the U.S.
The diversity of vocational vehicles also includes applications such as terminal tractors,
street sweepers, concrete pumpers, asphalt blasters, aircraft deicers, sewer cleaners, mobile
medical clinics, bookmobiles, and mobile command centers. Most of these are produced by
manufacturers of the vehicles listed in Table 1-2, while some are produced by small, specialized
companies.
In terms of total production volume, Table 1-3 summarizes what we know about the sales
of the seven custom chassis vehicle types. Of the other miscellaneous vehicles, the ones
produced in the highest volume are the terminal tractors, at about 6,000 per year (including those
certified with nonroad engines), with typical annual miles of less than 10,000 miles per year.12

-------
Table 1-3 Custom Chassis Population Estimates
APPLICATION TYPE
PERCENT OF NEW MY 2018
VOCATIONAL POPULATION
AVERAGE VMT IN
FIRST YEAR
Coach (Intercity) Bus
1%
85,000
Motor Home
13%
2,000
School Bus
10%
14,000
Transit Bus
1%
64,000
Refuse Truck
3%
34,000
Cement Mixer b
1%
20,000
Emergency Vehicle 0
1%
6,000
Notes:
1:1 Source: MOVES 2014 for all except mixer and emergency.A
b Source for cement mixer is UCS13
0 Source for emergency is ICCT (2009)14 and FAMA (2004)15
A Vehicle populations are estimated using MOVES2014. More information on projecting populations in MOVES is
available in the following report: USEPA (2015). "Population and Activity of On-road Vehicles in MOVES2014 -
Draft Report" Docket No. EPA-HQ-OAR-2014-0827.

-------
References
1	U.S. EPA and NHTSA, 2011. Regulatory Impact Analysis for the Final Rulemaking to Establish Greenhouse Gas
Emissions Standards and Fuel Efficiency Standards for Medium and Heavy-Duty Engines and Vehicles. EPA-420-
R-l 1-901. Available at: http://www3.epa.gov/otaq/climate/documents/420rll901.pdf.
2	http://www.ttmanet.org.
3	49 CFR 565.
4	"Both Sides of Truck Weight, Size Gear Up for Next Battle." Szakonyi, Mark. February 25, 2014. Available at:
http://www.joc.com/regulation-policy/transportation-policy/us-transportation-policy. Accessed: August 18, 2014.
5	"Wabash Shows What a 33-Foot Pup Would Look Like." Berg, Tom. Heavy-Duty Trucking TruckingInfo.com.
March 31, 2014. Available at: www.truckinginfo.com/blog/trailer-talk/story/2014/03/wabash-shows-what-a-33-
foot-pup-would-look-like.aspx. Accessed: September 23, 2014.
6	Per 46 CFR § 340.2.
7	19 CFR 115.3.
8	Trailer-BodyBuliders. North American Trailer Output Report, 2015 Trailer Production Figures Table. Available
online at: http://trailer-bodybuilders.com/trailer-output/2015-trailer-production-figures-table.
9	Dun & Bradstreet. Hoover's Inc. Online Company Database. Available at: http://www.hoovers.com.
10	U.S. Census Bureau. 2002 Economic Census - Vehicle Inventory and Use Survey. 2002. Available at:
https://www.census.gov/prod/ec02/ec02tv-us.pdf.
11	TIAX. LLC. "Assessment of Fuel Economy Technologies for Medium- and Heavy-Duty Vehicles," Final Report
to the National Academy of Sciences, November 19, 2009. Page 4-49.
12	See Charged Magazine, 2012 article, https://chargedevs.com/features/find-your-ninche-balqon-corporation-
targets-short-haul-drayage-tractors/, accessed April 2016.
13	National Ready Mixed Association Fleet Benchmarking and Costs Survey,
http://www.nxtbook.eom/naylor/NRCQ/NRCQ0315/index.php#/22, fromUCS Custom Chassis Recommendations,
May 2016.
14ICCT, May 2009, "Heavy-Duty Vehicle Market Analysis: Vehicle Characteristics & Fuel Use, Manufacturer
Market Shares."
15 Fire Apparatus Manufacturer's Association, Fire Apparatus Duty Cycle White Paper, August 2004, available at
http://www.deepriverct.us/firehousestudy/reports/Apparatus-Duty-Cycle.pdf.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Chapter 2: Technology and Cost
2.1 Overview of Technologies
In discussing the potential for CO2 emission and fuel consumption reductions, it can be
helpful to think of the work flow through the system. The initial work input is fuel. Each gallon
of fuel has the potential to produce some amount of work and will produce a set amount of CO2
(about 22 pounds (10 kg) of CO2 per gallon of diesel fuel). The engine converts the chemical
energy in the fuel to useable work to move the truck. Any reductions in work demanded of the
engine by the vehicle or improvements in engine fuel conversion efficiency will lead directly to
CO2 emission and fuel consumption reductions.
Current diesel engines are around 40 percent efficient over a range of operating
conditions depending on engine sizes and applications, while gasoline engine efficiency is much
lower than that of diesel engines. This means that approximately one-third of the fuel's chemical
energy is converted to useful work and roughly two-thirds is lost to friction, gas exchange, and
waste heat in the coolant and exhaust. In turn, the truck uses this work delivered by the engine to
overcome overall vehicle-related losses such as aerodynamic drag, tire rolling resistance, friction
in the vehicle driveline, and to provide auxiliary power for components such as air conditioning
and lights. Lastly, the vehicle's operation, such as vehicle speed and idle time, affects the
amount of total energy required to complete its activity. While it may be intuitive to look first to
the engine for CO2 emission and fuel consumption reductions given that only about one-third of
the fuel is converted to useable work, it is important to realize that any improvement in vehicle
efficiency proportionally reduces both the work demanded and the energy wasted.
Technology is one pathway to improve heavy-duty truck GHG emissions and fuel
consumption. Near-term solutions exist, such as those being deployed by SmartWay partners in
heavy-duty truck long haul applications. Other solutions are currently under development and
being implemented in the light-duty vehicle segment, especially in the large pickup sector where
some of the technologies apply to the heavy-duty pickup trucks covered under this rulemaking.
Long-term solutions are currently under development to improve efficiencies and cost-
effectiveness. While there is not a "silver bullet" that will significantly eliminate GHG
emissions from heavy-duty trucks like the catalytic converter has for criteria pollutant emissions,
significant GHG and fuel consumption reductions can be achieved through a combination of
engine, vehicle system, and operational technologies.
The following sections will discuss technologies in relation to each of the regulatory
categories - Heavy-Duty Pickup Trucks and Vans, Heavy-Duty Engines, Class 7 and 8
Combination Tractors, Trailers, and Class 2b-8 Vocational Vehicles. In each of these sections,
information on technological approaches, costs, and percent improvements is provided.
Depending on the segment, the vehicle-level technologies available for consideration may
include idle reduction, improved tire rolling resistance, improved transmissions, improved axles,
weight reduction, improved accessories, and aerodynamic technologies. Depending on the
segment, the engine-level technologies available for consideration may include friction
reduction, variable valve timing, cylinder deactivation, turbocharging, downsizing, combustion
optimization, aftertreatment optimization, and waste heat recovery. The agencies are not

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
projecting that all of the technologies discussed in these sections would be used for compliance
with the engine and vehicle standards, for reasons that are also discussed in each section.
Nevertheless, the potential for there to be technologies other than those which form the basis for
the compliance pathway set forth by the agencies, or which can be used in different combinations
or penetration rates than that projected compliance pathway, is an important consideration in
assessing the feasibility of the standards. Summaries of all of the technologies, along with the
corresponding costs, fuel consumption and GHG emissions improvement percentages are
provided in this chapter. This chapter also describes the agencies' basis for determining
penetration rates for the various technologies for each of the respective regulatory subcategories.
Summaries of engine technologies, effectiveness, and costs are provided in Chapters 2.2, 2.3,
2.6, and 2.7. A summary of engine and vehicle technologies, effectiveness, and costs for HD
pickup trucks and vans is provided in Chapter 2.5. A summary of technologies, effectiveness,
and costs for tractors is provided in Chapter 2.8. A summary of technologies, effectiveness, and
costs for vocational vehicles is provided in Chapter 2.9. A summary of technologies,
effectiveness, and costs for trailers is provided in Chapter 2.10. A detailed analysis of
technology costs is found in Chapters, 2.11 and 2.12.
EPA and NHTSA collected information on the cost and effectiveness of fuel
consumption and CO2 emission reducing technologies from several sources. The primary
sources of information were the 2010 National Academy of Sciences report on Technologies and
Approaches to Reducing the Fuel Consumption of Medium- and Heavy-Duty Vehicles (NAS)1,
TIAX's assessment of technologies to support the NAS panel report (TIAX)2, EPA's Heavy-
Duty Lumped Parameter Model3, the analysis conducted by NESCCAF, ICCT, Southwest
Research Institute and TIAX for reducing fuel consumption of heavy-duty long haul combination
tractors (NESCCAF/ICCT)4, and the technology cost analysis conducted by ICF for EPA (ICF).5
In addition, the agencies relied on NHTSA's technology assessment report under contract with
SwRI and Tetra Tech.6'7'8 We also held many meetings with engine and vehicle OEMs and
received information from comment to the notice of proposed rulemaking that further informed
our decision making process. In addition, the agencies used the vehicle simulation model (the
Greenhouse gas Emissions Model or GEM) to quantify the effectiveness of various technologies
on CO2 emission and fuel consumption reductions in terms of vehicle performance. These
values were used, in turn, to calculate standard stringency of all standards where GEM is used in
determining ultimate compliance. Thus, in all instances where GEM is used for compliance, it
was also used in determining standard stringency. The simulation tool is described in RIA
Chapter 4 in more detail.
2.2 Technology Principles - SI Engines
The engine technology principles described in this chapter for SI and CI engines are
typically described as applying for gasoline and diesel-fueled engines, respectively. Even so,
these technology principles generally also apply for engines powered by other fuels, including
natural gas. In Section II of the Preamble to these rules, the agencies describe regulatory
provisions that differ between SI and CI engines. Technologies related to closed crankcases for
natural gas engines are described below in Chapter 2.11 and in the Preamble Section XI.B.2.d.
The agencies describe technologies and test procedures related to minimizing evaporative
emissions from natural gas fuel systems in Chapter 2.11 as well as in Section XI.B.2.f of the
Preamble to these rules. The agencies' approach in this document is to first describe the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
principles of how technologies can work for an engine, without specifying the type of vehicle
into which it will be installed, or the test cycle over which it will be certified. Later, in Chapter
2.5, the agencies describe a subset of these technologies as they apply specifically to complete
HD pickup trucks and vans over their applicable operation and test cycles. In Chapter 2.6, the
agencies describe a subset of these technologies as they apply to SI engines intended for
vocational vehicles. The effectiveness values described in this section are ranges that cover SI
and CI engines in general and will differ between vocational vehicles which are engine certified
and HD pickup trucks and vans which are chassis certified. The effectiveness ranges represent
expected levels of effectiveness with appropriate implementation of the technology but actual
effectiveness levels will vary with manufacturer specific design and specifications for the
technologies. These may include considerations for durability or other related constraints. The
agencies did not receive comments disputing the expected technology effectiveness values
reported in the NPRM.
2.2.1	Engine Friction Reduction
In addition to low friction lubricants, manufacturers can reduce friction and improve fuel
consumption by improving the design of engine components and subsystems. Examples include
improvements in low-tension piston rings, piston skirt design, roller cam followers, improved
crankshaft design and bearings, material coatings, material substitution, more optimal thermal
management, and piston and cylinder surface treatments. The 2010 NAS Report, NESCCAF9
and EEA10 reports as well as confidential manufacturer data used in the both the light-duty and
this heavy-duty vehicle rulemaking suggest a range of effectiveness for engine friction reduction
to be between 1 to 3 percent. Reduced friction in bearings, valve trains, and the piston-to-liner
interface would improve efficiency. Any friction reduction must be carefully developed to avoid
issues with durability or performance capability.
2.2.2	Variable Valve Timing
Variable valve timing (VVT) classifies a family of valve-train designs that alter the
timing of the intake valve, exhaust valve, or both, primarily to reduce pumping losses, increase
specific power, and control the level of residual gases in the cylinder. VVT reduces pumping
losses when the engine is lightly loaded by controlling valve timing closer to the optimum
needed to sustain horsepower and torque. VVT can also improve volumetric efficiency at higher
engine speeds and loads. Additionally, VVT can be used to alter (and optimize) the effective
compression ratio where it is advantageous for certain engine operating modes (e.g., in the
Atkinson Cycle).
VVT has now become a widely adopted technology in the light duty fleet and this
technology is readily adaptable to the heavy-duty fleet: in MY 2014, most of all new cars and
light trucks had engines with some method of variable valve timing.11 There are currently many
different types of variable valve timing being utilized by Manufacturers, which have a variety of
different names and methods. The three major types of VVT are listed below.
Each implementation of VVT uses a cam phaser to adjust the camshaft angular position
relative to the crankshaft position, referred to as "camshaft phasing." This phase adjustment
results in changes to the pumping work required by the engine to accomplish the gas exchange

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
process. The majority of current cam phaser applications use hydraulically-actuated units,
powered by engine oil pressure and managed by a solenoid that controls the oil pressure supplied
to the phaser.
2.2.2.1	Coupled Cam Phasing for Overhead Valve (OHV) and Single Overhead
Camshaft (SOHC) Engines
Valvetrains with coupled (or coordinated) cam phasing (CCP) can modify the timing of
both the inlet valves and the exhaust valves an equal amount by varying the phasing of the
camshaft across an engine's range of operating speeds; also known as VVT. For engines
configured as an overhead valve (OHV) or as a single overhead camshaft (SOHC) only one cam
phaser is required per camshaft to achieve CCP.
Based on the heavy-duty Phase 1 vehicle rulemaking, 2015 NHTSA Technology Study,
and previously-received confidential manufacturer data, the agencies estimate the fuel
consumption reduction effectiveness of this technology to be between 1 and 3 percent for heavy-
duty applications across the different test cycles and operational opportunities.
2.2.2.2	Intake Cam Phasing (ICP) for Dual Overhead Camshaft Engines
(DOHC)
Valvetrains with ICP, which is the simplest of the cam phasing technologies, can modify
the timing of the inlet valves by phasing the intake camshaft while the exhaust valve timing
remains fixed. This requires the addition of a cam phaser on each bank of intake valves on the
engine. An in-line 4-cylinder engine has one bank of intake valves, while V-configured engines
have two banks of intake valves.
Some newer Class 2b and 3 market entries are offering dual overhead camshaft (DOHC)
engine designs where two camshafts are used to operate the intake and exhaust valves
independently. Consistent with the heavy-duty 2014-2018 MY vehicle rulemaking and the SwRI
report, the agencies agree with the effectiveness values of 1 to 2 percent reduction in fuel
consumption for heavy-duty applications across the different test cycles and operational
opportunities, for this technology.
2.2.2.3	Dual Cam Phasing (DCP) for Dual Overhead Camshaft Engines
(DOHC)
The most flexible VVT design is dual (independent) cam phasing, where the intake and
exhaust valve opening and closing events are controlled independently. This option allows the
option of controlling valve overlap, which can be used as an internal EGR strategy. At low
engine loads, DCP creates a reduction in pumping losses, resulting in improved fuel
consumption. Increased internal EGR also results in lower engine-out NOx emissions. The
amount by which fuel consumption is improved depends on the residual tolerance of the
combustion system. Additional improvements are observed at idle, where low valve overlap
may result in improved combustion stability, potentially reducing idle fuel consumption. DCP
requires two cam phasers on each bank of the engine.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Some newer Class 2b and 3 market entries are offering dual overhead camshaft (DOHC)
engine designs where two camshafts are used to operate the intake and exhaust valves
independently. Consistent with the light-duty 2012-2016 MY vehicle rulemaking and the SwRI
report, the agencies agree with the effectiveness values of 1 to 3 percent reduction in fuel
consumption for heavy-duty applications across the different test cycles and operational
opportunities for this technology.
2.2.2.4	Variable Valve Lift (VVL)
Controlling the lift of the valves provides a potential for further efficiency improvements.
By optimizing the valve-lift profile for specific engine operating regions, the pumping losses can
be reduced by reducing the amount of throttling required to produce the desired engine power
output. By moving the throttling losses further downstream of the throttle valve, the heat
transfer losses that occur from the throttling process are directed into the fresh charge-air mixture
just prior to compression, delaying the onset of knock-limited combustion processes. Variable
valve lift control can also be used to induce in-cylinder mixture motion, which improves fuel-air
mixing and can result in improved thermodynamic efficiency. Variable valve lift control can
also potentially reduce overall valvetrain friction. At the same time, such systems may also incur
increased parasitic losses associated with their actuation mechanisms. A number of
manufacturers have already implemented VVL into their fleets (Toyota, Honda, and BMW), but
overall this technology is still available for most of the fleet. There are two major classifications
of variable valve lift, described below:
2.2.2.5	Discrete Variable Valve Lift (DVVL)
Discrete variable valve lift (DVVL) systems allow the selection between two or three
discrete cam profiles by means of a hydraulically-actuated mechanical system. By optimizing
the cam profile for specific engine operating regions, the pumping losses can be reduced by
reducing the amount of throttling required to produce the desired engine power output. This
increases the efficiency of the engine. These cam profiles consist of a low and a high-lift lobe,
and may include an inert or blank lobe to incorporate cylinder deactivation (in the case of a 3-
step DVVL system). DVVL is normally applied together with VVT control. DVVL is also
known as Cam Profile Switching (CPS). DVVL is a mature technology in LD applications with
low technical risk.
Based on the light-duty MY 2017-2025 final rule, previously-received confidential
manufacturer data, 2015 NHTSA Technology Study, and report from the Northeast States Center
for a Clean Air Future (NESCCAF), the agencies estimate the fuel consumption reduction
effectiveness of this technology to be between 1 and 3 percent for heavy-duty applications across
the different test cycles and operational opportunities.
2.2.3 Cylinder Deactivation
In conventional spark-ignited engines throttling the airflow controls engine torque output.
At partial loads, efficiency can be improved by using cylinder deactivation instead of throttling.
Cylinder deactivation can improve engine efficiency by disabling or deactivating (usually) half
of the cylinders when the load is less than half of the engine's total torque capability - the valves

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
are kept closed, and no fuel is injected - as a result, the trapped air within the deactivated
cylinders is simply compressed and expanded as an air spring, with reduced friction and heat
losses. The active cylinders combust at almost double the load required if all of the cylinders
were operating. Pumping losses are significantly reduced as long as the engine is operated in
this "part cylinder" mode. Effectiveness improvements scale roughly with engine displacement-
to-vehicle weight ratio: the higher displacement-to-weight vehicles, operating at lower relative
loads for normal driving, have the potential to operate in part-cylinder mode more frequently.
Cylinder deactivation is less effective on heavily-loaded vehicles because they require more
power and spend less time in areas of operation where only partial power is required. The
technology also requires proper integration into the vehicles which is difficult in the vocational
vehicle segment where often the engine is sold to a chassis manufacturer or body builder without
knowing the type of transmission or axle used in the vehicle or the precise duty cycle of the
vehicle. The cylinder deactivation requires fine tuning of the calibration as the engine moves
into and out of deactivation mode to achieve acceptable NVH. Additionally, cylinder
deactivation would be difficult to apply to vehicles with a manual transmission because it
requires careful gear change control. NHTSA and EPA adjusted the 2017-2025 MY light-duty
rule estimates using updated power to weight ratings of heavy-duty trucks and confidential
business information and downwardly adjusted the effectiveness to 0 to 3 percent over average
driving patterns for these vehicles to reflect the differences in drive cycle and operational
opportunities compared to light-duty vehicles.
2.2.4 Stoichiometric Gasoline Direct Injection (SGDI)
Stoichiometric gasoline direct injection (SGDI) engines inject fuel at high pressure
directly into the combustion chamber (rather than into the intake port in port fuel injection).
SGDI requires changes to the injector design, an additional high pressure fuel pump, new fuel
rails to handle the higher fuel pressures, and changes to the cylinder head and piston crown
design. Direct injection of the fuel into the cylinder improves cooling of the air/fuel charge
within the cylinder, which allows for higher compression ratios and increased thermodynamic
efficiency without the onset of combustion knock. Recent injector design advances, improved
electronic engine management systems and the introduction of multiple injection events per
cylinder firing cycle promote better mixing of the air and fuel, enhance combustion rates,
increase residual exhaust gas tolerance, and improve cold start emissions. SGDI engines achieve
higher power density and match well with other technologies, such as boosting and variable
valvetrain designs.
Several manufacturers have recently introduced vehicles with SGDI engines, including
GM and Ford, who have announced their plans to increase dramatically the number of SGDI
engines in their vehicle portfolios.
Based on the heavy-duty 2014-2018 MY vehicle rulemaking, the 2015 NHTSA
Technology Study, and previously-received confidential manufacturer data, the agencies
estimate the fuel consumption reduction effectiveness of SGDI to be between 1 and 2 percent for
heavy-duty applications across the different test cycles and operational opportunities.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.2.5 Turbocharging and Downsizing (TRBDS)
The specific power of a naturally aspirated engine is primarily limited by the rate at
which the engine is able to draw air into the combustion chambers. Turbocharging and
supercharging (grouped together here as boosting) are two methods to increase the intake
manifold pressure and cylinder charge-air mass above naturally aspirated levels. Boosting
increases the airflow into the engine, thus increasing the specific power level, and with it the
ability to reduce engine displacement while maintaining performance. This effectively reduces
the pumping losses at lighter loads in comparison to a larger, naturally aspirated engine.
Almost every major manufacturer offering both light-duty and heavy-duty vehicles,
including 2b/3s, currently markets vehicles in their light-duty offerings with some form of
boosting. Only one manufacturer, Ford, has allowed the light-duty derived boosted engine to
migrate into its 2b/3 van offering. The ability to use a smaller boosted engine is currently limited
to applications where operational duty cycles are more consistent with light-duty vehicles of
similar utility like full size pick-ups and MDPVs. The Ford 2b/3 van has similar capability as
the light-duty pick-up from which the boosted engine is borrowed. In applications that require
high payload or towing capacity that substantially exceeds the light-duty ranges of towing
capacity, manufacturers have chosen to maintain the larger displacement non-boosted engines
because of the boosted engine's loss of effectiveness when performing towing. In their
comments, AAPC illustrated this issue showing that downsized and boosted engines actually
perform worse from a brake specific fuel consumption perspective when encountering high
loads, such as towing, than a traditional non-boosted engine of more historical displacements.
Class 4 and higher vocational vehicles have not employed any form of boosted and downsized
engines because of this penalty. In our projected compliance pathways for pickups and vans, the
agencies are projecting use of a smaller boosted engine only where suited to a 2b/3 vehicle's
duty cycles - reflecting current industry practice. This approach properly targets GHG and fuel
consumption reductions to the expected vehicle duty cycles and provides a balance based on the
consumer's requirements of their work vehicle.
While boosting has been a common practice for increasing performance for several
decades in light-duty vehicles, turbocharging has considerable potential to improve fuel economy
and reduce CO2 emissions when the engine displacement is also reduced. Specific power levels
for a boosted engine often exceed 100 hp/L, compared to average naturally aspirated engine
power densities of roughly 70 hp/L. As a result, engines can be downsized roughly 30 percent or
higher while maintaining similar peak output levels. However, as just discussed above, the
effectiveness of boosted and downsized engines is a function of duty cycle and may not be
appropriate for some applications encountering regular high loads such as towing. In the last
decade, improvements to turbocharger turbine and compressor design have improved their
reliability and performance across the entire heavy-duty engine operating range. New variable
geometry turbines and ball-bearing center cartridges allow faster turbocharger spool-up (virtually
eliminating the once-common "turbo lag") while maintaining high flow rates for increased boost
at high engine speeds. Low speed torque output has been dramatically improved for modern
turbocharged engines. However, even with turbocharger improvements, maximum engine
torque at very low engine speed conditions, for example launch from standstill, is increased less
than at mid and high engine speed conditions. The potential to downsize engines may be less on
vehicles with low displacement to vehicle mass ratios, for example, a very small displacement

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
engine in a vehicle with significant curb weight, cargo weight or towing, in order to provide
adequate acceleration from standstill, particularly up grades or at high altitudes.
Use of gasoline direct injection (GDI) systems with turbocharged engines and charge air
cooling also reduces the fuel octane requirements for knock limited combustion and allows the
use of higher compression ratios. Ford's "EcoBoost" downsized, turbocharged GDI engines
introduced on MY 2010 light-duty vehicles allow the replacement of V8 engines with V6
engines with improved 0-60 mph acceleration and with fuel economy improvements of up to 12
percent as documented in their technical paper.12
Recently published data with advanced spray-guided injection systems and more
aggressive engine downsizing targeted towards reduced fuel consumption and CO2 emissions
reductions indicate that the potential for reducing CO2 emissions for turbocharged, downsized
GDI engines may be as much as 15 to 30 percent relative to port-fuel-injected engines.
Confidential manufacturer data suggest an incremental range of fuel consumption and CO2
emission reduction of 4.8 to 7.5 percent for turbocharging and downsizing. Other publicly-
available sources suggest a fuel consumption and CO2 emission reduction of 8 to 13 percent
compared to current-production naturally-aspirated engines without friction reduction or other
fuel economy technologies: a joint technical paper by Bosch and Ricardo suggesting fuel
economy gain of 8 to 10 percent for downsizing from a 5.7 liter port injection V8 to a 3.6 liter
V6 with direct injection using a wall-guided direct injection system;13 a Renault report
suggesting a 11.9 percent NEDC fuel consumption gain for downsizing from a 1.4 liter port
injection in-line 4-cylinder engine to a 1.0 liter in-line 4-cylinder engine, also with wall-guided
direct injection;14 and a Robert Bosch paper suggesting a 13 percent NEDC gain for downsizing
to a turbocharged DI engine, again with wall-guided injection.15 These reported fuel economy
benefits show a wide range depending on the SGDI technology employed and the use of these
technologies are directly applicable to heavy-duty SI engines.
The agencies reviewed estimates from the 2017-2025 final light-duty rule, the TSD, and
existing public literature. The previous estimate from the MYs 2017-2025 suggested a 12 to 14
percent effectiveness improvement, which included low friction lubricant (level one), engine
friction reduction (level one), DCP, DVVL and SGDI, over baseline fixed-valve engines, similar
to the estimate for Ford's EcoBoost engine, which is already in production in light-duty and .
Additionally, the agencies analyzed Ricardo vehicle simulation data and the 2015 NHTSA
Technology Study for various turbocharged engine packages.
2.2.6 Engine Down Speeding
In general, engine down speeding has been determined to reduce frictional losses and also
reduce the need for component temperature protection in SI engines. Component protection
occurs at higher engine speeds and loads where components such as exhaust valves, exhaust
manifolds, catalysts and other components in the exhaust system reach temperatures where
materials may require cooling to prevent damage or reduced durability and accelerated
deterioration. The SI engine has various methods of accomplishing this protection requirement
including using additional fuel enrichment to act as a coolant in the exhaust. Other methods to
reduce exhaust component temperatures include reducing engine output such as torque
governing through variable valve timing, limiting boost in boosted engines or simply reducing

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
air flow into the engine by commanding the electronic throttle to a smaller percentage opening
thereby reducing available air volume.
In the case of chassis certified pick-ups and vans, down speeding is generally achieved by
managing the transmission gear selection in electronically controlled automatic transmissions. It
is largely contained in the transmission technology description in Chapter 2.5 below. There is
typically no incentive to implement additional strategies for limiting engine speed as described
above as they are not quantified in the test cycles and may require a reduction in advertised rated
engine power which can become a competitive disadvantage.
Vocational vehicles which use SI engines certified to GHG and criteria emissions over
the FTP engine dyno cycle can capture the benefits of down speeding more favorably. Since
FTP engine certification is based on a test method that first quantifies the total available engine
power from idle to the electronically governed engine top speed or rev limiter, the opportunity
exists to shift the entire engine operation down to lower engine speeds where frictional losses are
lower and need for temperature protection is reduced. This strategy will generally require the
engine manufacturer to reduce peak power and engine speed rating of the engine. This strategy
has not been used in past SI engine certifications so little information exists about its
effectiveness but the expected range of effectiveness is 0 to 4 percent depending on the
aggressiveness of the down speeding.
2.3 Technology Principles - CI Engines
In this section, technology principles for CI engines will be discussed. Although most
technologies discussed here, with the exception of engine downsizing, down speeding, and WHR
with Rankine cycle technology were considered by the agencies as potentially available for
compliance with the Phase 1 engine standards, the level of improvement and complexity are
different for Phase 2. It should be mentioned that the technologies discussed here are for
compression ignition diesel engines and are not interchangeable with technologies used for spark
ignition engines. See the spark ignition engine discussion in Chapter 2.2 Technology Principles
- SI Engines.
2.3.1 Low Temperature Exhaust Gas Recirculation
Most LHDD, MHDD, and HHDD engines sold in the U.S. market today use cooled EGR,
in which part of the exhaust gas is routed through a cooler (rejecting energy to the engine
coolant) before being returned to the engine intake manifold. EGR is a technology employed to
reduce peak combustion temperatures and thus NOx. Low-temperature EGR uses a larger or
secondary EGR cooler to achieve lower intake charge temperatures, which tend to further reduce
NOx formation. For a given NOx requirement, low-temperature EGR can allow changes such as
more advanced injection timing that would increase engine efficiency slightly more than one
percent. Because low-temperature EGR reduces the engine's exhaust temperature, it has not
been considered as part of a technology package that also includes exhaust energy recovery
systems such as turbocompounding or a bottoming cycle.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.3.2	Combustion System Optimization
Improvements in the fuel injection system allow more flexible fuel injection capability
with higher injection pressure and can improve engine fuel efficiency, while maintaining the
same emission level. Combustion system optimization, featuring piston bowl, injector tip and
the number of holes, in conjunction with the advanced fuel injection system, is able to further
improve engine performance and fuel efficiency. Manufacturers have been working to improve
engines in these areas for some time. At this point, all engine manufacturers have substantial
development efforts underway that we project will be translated into production in the near
future. Some examples include the combustion development programs conducted by
Cummins16, Detroit Diesel17, and Navistar18 funded by Department of Energy as part of the
SuperTruck program. These manufacturers found that improvement due to combustion alone
during this program was 1 to 2 percent. While their findings are still more focused on the
research end of development, specifically targeting one optimal operating point, the results of
these research programs do support the possibility that some of the technologies they are
developing could be applied to production engines in the 2027 time frame. The agencies have
determined that it is feasible that fuel consumption and CO2 emissions could be reduced by as
much as 1.0 percent in the agencies' certification cycles in the 2027 time frame through the use
of these technologies.
Some technologies were evaluated but not included in the agencies' technical feasibility
analysis for the Phase 2 regulation since the agencies do not anticipate these technologies will be
commercially available by 2027. For example, alternative combustion processes such as
homogeneous charge compression ignition (HCCI), premixed charge compression ignition
(PCCI), low-temperature combustion (LTI), and reactivity controlled compression ignition
(RCCI) technologies were not included in the agencies' feasibility analysis for Phase 2. While
these technologies show good indicated thermal efficiency, fuel savings over the entire range of
engine operation is still a major challenge. At the current level of development it is not clear that
the technologies will be in commercial production by 2027. This, however, does not preclude
the use of these technologies for compliance should manufacturers develop and commercialize
these alternative combustion or other approaches.
2.3.3	Model Based Control
Another important area of potential improvement is advanced engine control
incorporating model based calibration to reduce losses of control during transient operation.
Improvements in computing power and speed would make it possible to use much more
sophisticated algorithms that are more predictive than today's controls. Because such controls
are only beneficial during transient operation, they would reduce emissions over the Federal Test
Procedure (FTP) cycle, but not over the Supplemental Emission Test (SET) cycle. Detroit Diesel
introduced the next generation model based control concept, achieving 4 percent thermal
efficiency improvement while simultaneously reducing emissions in transient operations in their
earlier report.19 More recently, this model based control technology was put into their one of
vehicles for final demonstration under DOE's SuperTruck program.20 Their model based
concept features a series of real time optimizers with multiple inputs and multiple outputs. This
controller contains many physical based models for engine and aftertreatment. It produces fully
transient engine performance and emissions predictions in a real-time manner. Although we do

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
not project that this control concept would be in MY 2017 production, real time model control
could be in production during the Phase 2 time frame, thus significantly improving engine fuel
economy.
2.3.4 Turbocharging System
Many advanced turbocharger technologies can be added into production in the time frame
between MYs 2021 and 2027 and some of them are already in production, such as mechanical or
electric turbocompound, more efficiency variable geometry turbine, and Detroit Diesel's
patented asymmetric turbocharger. A turbocompound system extracts energy from the exhaust
to provide additional power. Mechanical turbocompounding includes a power turbine located
downstream of the turbine which in turn is connected to the crankshaft to supply additional
power. On-highway demonstrations of this technology began in the early 1980s. It has been
first used in heavy duty production by Detroit Diesel for their DDI5 and DDI6 engines. That
company claims a 3 to 5 percent fuel consumption reduction due to the system.21 Results are
duty cycle dependent, and require significant time at high load to see a fuel efficiency
improvement. Light load factor vehicles can expect little or no benefit. Volvo reports two to
four percent fuel consumption improvement in line haul applications, which would be likely in
production even before 2020.22
Electric turbo-compound is another potential technology that can improve engine brake
efficiency. Efficiencies are attained through better vehicle integration and lower backpressure
impacts. Since the electric power turbine speed is no longer linked to crankshaft speed, this
allows more efficient operation of the turbine. Navistar reports on the order of a 1 to 1.6 percent
efficiency improvement over mechanical turbocompound systems at 0.5 to 0.7 gm/hp-hr engine-
out NOx levels.23'24 This concept, however, does not work well with lower engine out NOx as
indicated in the report, as zero benefit is reported at 0.3 to 0.4 gm/hp-hr engine-out NOx, due to
lower exhaust gas temperatures. Navistar reports a 1.6 percent fuel efficiency improvement,
again as compared to a mechanical turbocompound system.
Two-stage turbocharger technology has been used in production by Navistar and other
manufacturers. Ford's new developed 6.7L diesel engine features a twin-compressor
turbocharger. Higher boost with wider range of operations and higher efficiency can further
enhance engine performance, thus fuel economy. It is expected that this type of technology will
continue to be improved by better matching with system and developing higher compressor and
turbine efficiency.
Furthermore, improved turbocharger efficiency when combined with turbocompounding
was shown in the SwRI study to reduce fuel consumption while maintaining criteria emissions
limits. Findings show that there is limited scope for improved turbocharger efficiency on
engines which do not use turbocompound, because an increase in turbocharger efficiency would
result in reduced or eliminated EGR flow which in turn would cause the engine to exceed NOx
emissions requirements.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.3.5	Engine Breathing System
Various high efficiency air handling (air and exhaust transport) processes could be
produced for heavy duty applications in the Phase 2 time frame. To maximize the efficiency of
such processes, induction systems may be improved by manufacturing more efficiently designed
flow paths (including those associated with air cleaners, chambers, conduit, mass air flow
sensors and intake manifolds) and by designing such systems for improved thermal control.
Improved turbocharging and air handling systems must include higher efficiency EGR systems
and intercoolers that reduce frictional pressure loss while maximizing the ability to thermally
control induction air and EGR. EGR systems that often rely upon an adverse pressure gradient
(exhaust manifold pressures greater than intake manifold pressures) must be reconsidered and
their adverse pressure gradients minimized. "Hybrid" EGR strategies which rely upon pressure
gradients and EGR pumps may provide pathways for improvement. Other components that offer
opportunities for improved flow efficiency include cylinder heads, ports and exhaust manifolds
to further reduce pumping losses. Cummins reports 1.4 percent through optimization.25 Detroit
Diesel projects a 2 percent fuel efficiency improvement through air handling system
development.26 Navistar predicts almost 4 percent through a combination of variable intake
valve closing timing (IVC), turbocharger efficiency and match improvements. A few plots in
this reference show another 4 percent, but these are not explained.
Variable air breathing systems such as variable valve actuation may provide additional
gains at different loads and speeds. The primary gain in diesel engines is achieved by varying
the EVO event versus engine speed and load, in conjunction with turbocharger optimization to
minimize blowdown losses. Navistar reports a 1.25 percent fuel consumption improvement.23
Again, all these reference points are referred to a single optimal point from the DOE SuperTruck
program.
2.3.6	Engine Parasitic and Friction Reduction
Engine parasitic and friction reduction is another key technical area that can be further
improved in the 2020 to 2027 time frame. Reduced friction in bearings, valve trains, and the
piston-to-liner interface can improve efficiency. Friction reduction opportunities in the engine
valve train and at its roller/tappet interfaces exist for several production engines. The piston at
its skirt/cylinder wall interface, wrist pin and oil ring/cylinder wall interface offers opportunities
for friction reduction. Use of more advanced lubricating oil that will be available in the future
will also play a key role in reducing friction. Any friction reduction must be carefully developed
to avoid issues with durability or performance capability. Lube oil and water pumps as well are
another area where efficiency improvements will occur. Navistar identifies a combined
improvement of up to 2 percent through reduced bearing friction, reduced piston and ring
friction, and unspecified lube oil pump improvements.27 In their 2012 paper they report 5.5
percent improvement through a combination of friction reduction and both lube and cooling
system improvements.23 In this same presentation they specified 0.45 percent demonstrated
through water pump improvements and 0.3 percent through lube pump improvements. The total
number of 5.5 percent remains optimistic, even for a single optimal test point. Cummins reports
a combined number of 3 percent.25. Detroit Diesel reports a combined number of 2 percent, with
0.5 percent coming from improved water pump efficiency. 26 Navistar shows a 0.9 percent
benefit for a variable speed water pump and variable displacement oil pump; 0.5 percent for

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
piston/ring/liner friction reduction; and 0.6 percent for bearing friction reduction23. In addition,
Federal-Mogul recently announced new piston ring coatings that can lead to a 20 percent
reduction in engine friction, and, in looking to the future, sees an opportunity to reduce friction
by an additional 30 percent, which is equivalent to a 1.2 percent reduction in brake specific fuel
consumption at road load conditions.28 It should be noted that water pump improvements
include both pump efficiency improvement and variable speed or on/off controls. Lube pump
improvements are primarily achieved using variable displacement pumps and may also include
efficiency improvement. All of the results shown in this paragraph are demonstrated through
DOE's SuperTruck program under a single optimal operating point, which has not been changed
since the proposal.
In addition, SwRI's reports show that if the exact certification cycles, weighting and
vehicle weights are used, the friction reduction in the Phase 2 timeframe is in the range of 1.47
percent compared to a 2018 baseline engine.7
2.3.7	Integrated Aftertreatment System
All manufacturers now use diesel particulate filters (DPF) to reduce particulate matter
(PM) and SCR to reduce NOx emissions, and these types of technologies are likely to be used
for compliance with criteria pollutant standards for many years to come. There are three areas
considered to improve integrated aftertreament systems, which result in a reduction of fuel
consumption. The first is better combustion system optimization through increased
aftertreatment efficiency. The second is reduced backpressure through further development of
the devices themselves. The third is reduced ammonia slip out of SCR during transient
operation, thus reducing net urea consumption. Navistar reports a 7 to 8 percent improvement in
efficiency projected through a combination of higher cylinder pressure, injection optimization,
and engine/aftertreatment optimization.23 Cummins reports a 0.5 percent improvement through
improved aftertreatment flow (catalyst size optimization and improved NOx surface
utilization)25 Detroit Diesel projects a 2 percent fuel efficiency improvement through reduced
use of EGR, thinner wall DPF, improved SCR cell density, and catalyst material optimization.26
2.3.8	Engine Downsizing and Down Speeding
Engine downsizing can be more effective if it is combined with down speeding which
leads to increased vehicle efficiency through lower power demand. This lower power demand
shifts the vehicle operating points to lower load zones, which moves the engine away from some
of the optimum operation points. In order to compensate for this loss, down speeding allows the
engine to move back into the optimum operating points resulting in reduced fuel consumption.
Increasing power density by reducing the engine size allows the vehicle operating points to move
back to the optimum operating points, thus further improving fuel economy. Both Detroit Diesel
and Volvo demonstrate the same methodology for proper implementation of downsizing 29'30
Detroit Diesel also shows that engine downsizing can result in friction reduction due to a
reduction in engine surface area when compared to a bigger bore engine.26
Engine down speeding can also be an effective fuel efficiency technology even when
used alone (i.e. not in combination with engine downsizing), especially when a vehicle uses a
fast axle ratio. Down speeding, in this situation, can allow the engine to operate in a lower speed

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
zone that is closer to or just in the middle of engine sweet spot, which is typically in the speed
range of 1100-1200 rpm for a heavy duty engine. In order to take advantage of a fast or low axle
ratio, the engine must be optimized toward the low speed zone by either generating higher peak
torque in the lower speed zone or shifting the entire rating speed into a lower rating, or a
combination thereof. The engine air handling and combustion system, as a result of these
changes, must be re-optimized to accommodate a typical higher peak cylinder pressure rise.
Depending on how the engine system is optimized, the overall engine fuel consumption can be
improved. However, from an engine certification standard point, such as the 13-mode SET
cycle, down speeding is always accompanied by moving mode speeds to a lower speed zone,
which usually take advantage of the sweet spot, thus making the engine more efficient in terms
of the certification cycle. On the other hand, from a vehicle operating standard point, the benefit
of down speeding is primarily realized through the use of a lower axle ratio, allowing the engine
to operate in an optimal zone.
2.3.9 Waste Heat Recovery
Organic Rankine Cycle waste heat recovery (WHR) systems have been studied for many
years. The agencies' overall assessment of WHR as a fuel saving technology is that it offers
great promise in the long term. However, it would take several years to develop, and initially, it
would be viable primarily in line-haul applications. The agencies recognize the many challenges
that would need to be overcome, but believe with enough time and development effort, this can
be done. We have received a large number of comments from the both the NPRM and NOD A
that yield two differing opinions. Most vehicle and engine manufacturers, with one exception,
objected to the purportedly aggressive technology penetration rate reflected in the proposed
engine standards. They argued that the WHR systems in the literature and utilized in the DOE
SuperTruck program are still in the research and development stage, and these systems are still a
long way off with respect to reaching production. Their voiced concern is that bringing this
technology to market before it is ready could lead to high warranty costs and reliability issues,
leading to significant down time for vehicles or fleets, possibly even beyond 2027. One engine
manufacturer, however, indicated that WHR systems could be used in a production setting as
early as the MY 2021 to 2027 time frame because their WHR system is approaching the
prototype stage of development, with projected small market penetration starting in 2021.
The basic approach of a WHR system is to use engine exhaust waste heat from multiple
sources to evaporate a working fluid in a heat exchanger. This evaporated fluid is then passed
through a turbine or equivalent expander to create mechanical or electrical power. The working
fluid is then condensed back to the fluid in the fluid reservoir tank and returned back to the flow
circuit via a pump to restart the cycle. With support of the Department of Energy, three major
engine and vehicle manufacturers have developed WHR systems under the SuperTruck program.
Cummins' WHR system is based on an organic Rankine cycle using refrigerant as the working
fluid.31'32 Their system recovers heat from the EGR cooler, as well as from the exhaust gas
downstream of the aftertreatment system. It converts that heat to power through a mechanical
gear train coupled to the engine's output shaft. Some iterations of their system also sought gains
from low-temperature coolant and lubricant heat rejection via a parallel loop. The system
includes a recuperator that transfers post-turbine energy back into the working fluid loop prior to
the condenser. This recuperator reduces condenser heat rejection requirements and improves
overall system efficiency. Volvo has developed a similar system to Cummins' with variations in

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
terms of hardware components, including the use of ethanol as the working fluid instead of a
refrigerant.33 Daimler, on the other hand, has developed a different type of ethanol based system
to recover heat from the exhaust gas using an electrical generator to provide power to charge a
high-voltage battery that is primarily used to drive a hybrid system.
Pre-prototype WHR systems have been shown to be very efficient under optimized
conditions. In demonstrations where operation occurred at a single optimal engine operating
point, Cummins reported potential efficiency gains from WHR on the order of 2.8 percent from
the baseline engine without WHR31, Volvo reported around 2.5 percent33, and Daimler reported
2.3 percent.29 It is important to note that all of these WHR systems are still in the pre-prototype
stage of research and development. Despite the promising performance of pre-prototype WHR
systems, the cost and complexity of these packages from Cummins, Volvo and Daimler remain
high. The agencies believe that manufacturers will continue to make improvements to these
systems over time, just as they have for other advanced technologies that initially had high cost
and complexity at a comparable stage of development.
WHR technology also poses issues with respect to package size and transient response.
The agencies believe that WHR will be less effective in urban traffic and will most likely be
applied to line haul vehicles. Our projected technology paths for compliance, and projected
technology penetration rates, reflect this assumption.
WHR may offer the benefit of replacing the EGR cooler and decrease cooling system
heat rejection requirements by converting some heat into work. To the extent that WHR systems
use exhaust heat, they may increase the overall cooling system heat rejection requirement, thus
increasing radiator size, which can have a negative impact on cooling fan power needs, as well as
on vehicle aerodynamics. Significant challenges could arise if the space under a vehicle's hood
happens to be tight, leaving little or no room for a larger radiator, thus necessitating a redesign of
the vehicle's front face, sacrificing potential aerodynamic improvements. This issue becomes
more challenging for truck cooling systems that are currently at cooling capacity design limits.
Current WHR systems are heavy, estimated to be on the order of 300-500 lbs depending
on system design. Without time to optimize designs, any attempt to reduce weight by simply
reducing the size of the key components, such as boilers and condensers, would likely have an
adverse impact on the system efficiency. Given enough lead time, the agencies believe
manufacturers might be able to improve materials and designs to reduce overall system weight
without compromising efficiency.
Manufacturers have not yet arrived at a consensus on which working fluid(s) will be used
in WHR systems to balance concerns regarding performance, global warming potential (GWP),
and safety. Current working fluids have a high GWP (conventional refrigerant), are expensive
(low GWP refrigerant), are hazardous (ammonia, etc.), are flammable (ethanol/methanol), or can
freeze (water). One of the challenges is determining how to seal the working fluid properly
under vacuum conditions with high temperature to avoid safety issues for flammable/hazardous
working fluids. Addressing leaks would also be an important issue with respect to greenhouse
gas emission for a high GWP working fluid. Because of these challenges, choice of working
fluid will be an important factor for system safety, efficiency, and overall production viability.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Other key challenges facing WHR systems are their reliability, durability, and market
acceptance. Durability concerns that have been raised include: boiler fouling and cracking
associated with high thermal gradients, thermal shock, condenser fouling, as well as sensor and
actuator durability under harsh temperature and pressure conditions. It can be reasonably
estimated that the current WHR systems under development by major engine manufacturers
consist of at least two hundred parts including expanders, boilers, condensers, and fluid pumps,
together with many fasteners, wiring cables, sensors, actuators, and piping. Determining overall
system efficacy and reliability involves rigorous testing in support of comprehensive Failure
Modes and Effects Analysis (FEMA). These parts, as well as the entire WHR system as a whole,
must undergo severe winter and summer tests. Multiple trucks equipped with the same WHR
system must be run on the road, accumulating millions of miles. During these tests, all failures
must be recorded, associated with specific failure modes or error codes, and the root cause of the
failure must be determined. Warranty costs for each failure mode based on component cost and
labor must be assigned. Due to the large number of components, some of the failure modes
might not be identified during the road tests even with multiple occurrences. It would be a high
risk for any manufacturer to put their new technology into the market without careful system
validation via on-the-road tests. Similarly, owners and operators might be unlikely to risk early
adoption of such a complex technology if premature deployment leads to potential down time,
along with its associated cost.
Based on the literature and preceding discussion, WHR technology can be characterized
as being in the technology demonstration stage for purposes such as the DOE SuperTruck
program. It should be clear that the demonstration defined by DOE SuperTruck program means
that the demonstrated truck with the technologies developed under the DOE program can be
successfully run through a pre-specific routes, and it doesn't mean that technologies used in the
truck reach any matured stage or prototype stage, regardless of cost. Although a few trucks with
WHR technology have been tested on the road,33'34'35 many of the components used in the trucks
and product-acceptable packaging are still years away from production. Figure 2-1 shows a
generic form of the product process flow. As can be seen from this figure, it could take 5-15
years from the applied research/development stage to arrive at the prototype stage depending on
the complexity of the technology. WHR is now in that prototype stage. During the prototype
stage, all prototype components must be available and extensive engine and vehicle tests with
WHR must be conducted. The production start-up phase would follow. After that, significant
efforts must be made to advance the system from a prototype to a commercial product, which
typically takes about five years for complex systems like WHR. During this approximate five-
year period, multiple vehicles will go through weather condition tests, long lead-time parts and
tools will be identified, and market launch and initial results on operating stability will be
completed. Production designs will be released, all product components should be made
available, production parts on customer fleets and weather road testing will be verified before
finally launching production, and distribution of parts to the vehicle service network for
maintenance and repair will be readied.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
5-15years
3-5years

Prototype component available
Testing complete engine/vehicle
Market
requirements
Product Launch
Verified with production parts
on customer fleets and all
weather road testing
Release long-lead p!xte ahd tool
Complete Market launctvronfc^pt
Initial results on operating st^bilrt'
Technical pro<|ucfvpncept
Fuel economy Mtjmafc^d
Technical risks addressedN
Design Released into production
All product components available
Financial feasibility
Simulation and Testing
Applied R/D
Prototype
Production
Figure 2-1 Product Process Flow
The GHG standards themselves can provide an effective incentive for manufacturers to
reach the commercial product stage earlier than would otherwise occur. They can motivate
manufacturers to shorten the period for advancing from a complicated prototype system to a
commercial product and can also help to ensure market penetration after launching a product.
Nevertheless, in order for WHR to be produced commercially, several things are needed. First, it
is critical to optimize the WHR package volume, cooling capability, and aero drag at typical
cruise speeds on highway since the most significant benefits of WHR technology would be in
line-haul applications. Removal of the exhaust heat exchangers located in the exhaust system
would reduce the total system volume and weight. Working fluids need to be selected with a
reasonably low GWP and high performance potential. In addition, the engine with a WHR
system needs to be continuously tested in a very well equipped engine dynamometer. This allows
to continue optimization in a system level as well as identification of issues associated with
reliability. On top of that, the component bench tests, such as individual components like heat
exchangers, condenser, and expander need to be extensively conducted through a series of
durability and performance test protocols for accumulated thousands of hours, thus identifying
any potential issues associated with reliability. In the meantime, one of the most effective
approaches should be to put a few hundred trucks into fleets for trial in the next several years, so
that a comprehensive FEMA can be thoroughly identified and warranty cost analyses be more
precisely conducted before launching into full volume production. The fleet testing results can
also provide valuable feedback to the engine dynamometer tests, thus continuing optimization of
the component size, weight and performance including working fluid. Manufacturers have
shown in the past that a robust FEMA process can address most problems before a technology is
more widely introduced. Therefore, the lead time appears to be one of the most noticeable
constraints.
We believe that all the issues and hurdles discussed above can be resolved with adequate
lead time. However, it would be challenging to predict high rates of initial market penetration
because of the many uncertainties as stated above. The NACFE report 36analyzes a wide range
of HD fuel efficiency technology adoption rates versus time, and we considered these recent
historic trends as we developed our market adoption rate projections. While more mature

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
technologies such as electronic fuel injection and turbocharging are not presented in this report,
the trends for a number of emerging fuel efficiency technologies are depicted. We note that a
number of charts which are relevant here are presented at the end of this report. Many of these
technologies are those that we are projecting to continue to increase in market adoption during
the Phase 2 timeframe. While there are a number of exceptions, many of the technology
adoption rate curves follow an S-shape: slow initial adoption as shown in Figure 2 of this
report36, then more rapid adoption, and then a leveling off as the market saturates (not always at
100 percent).
This characteristic S-curve is further annotated and expanded in the figure below. There
are two curves in this figure. "Simple" typically means that the technology can be relatively
quickly adopted by the market because of the technology complexity. The example includes the
use of aero fairings on the vehicle side, and turbocharger and fuel injection technologies on the
engine side. "Complex" means that the technology is so complicated that the market will take a
much longer time to adopt. WHR with the Rankine cycle is one of these types (but certainly not
the sole example). The agencies thus view it legitimate to apply this type of S-curve to WHR.
This figure also shows the four typical steps to reach high market penetration, but either
technology needs to go through an S-shape curve because of factors indicated on the left side of
this figure, which would make it difficult to quickly bring the technology into the market with
high market penetration. Taking "fleet consideration" of this figure as an example, the payback
time would be the most sensitive. Reliability, down time, limited credible data, resale values, and
capital investment are many of the other concerns. We believe that WHR adoption behavior can
very well follow the S-shape curve, where we project a steeper rise in market adoption in and
around the 2027 timeframe. We have worked closely with one of the engine manufacturers who
are leading WHR development. With reliable and credible CBI information, we now believe that
our initial estimate for 15 percent market penetration of WHR in MY 2027 was conservative.
Given our averaging, banking and trading program flexibilities and that manufacturers may
choose from a range of other technologies, we believe that manufacturers will be able to meet the
2027 standards, which we based on 25 percent WHR adoption in heavy duty tractor
engines. Again, this illustration is consistent with the findings reported by NACFE.36 For
example, the tire pressure inflation used for trailers follows this type of S-curve took four years
from 1 percent market penetration to 16 percent, and then to 31 percent in another year. One of
the key lessons learned from this report is that if a technology is pushed too hard and too quickly,
the market penetration could be rolled back because of reliability and warranty issues. See 80 FR
40236 noting similar concerns in a general context. Taking idle reduction with diesel APU
engine technology as an example, it quickly reached 15 percent market penetration from 3
percent in one year, and then reached 32 percent in four years, but it quickly dropped back to 13
percent in 3-4 years. This type of behavior could happen to WHR with Rankine cycle
technology if pushed too hard.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
100
sp
c
o
"+¦>
Q.
o
T5
<
Fleet Considerations:
Payback
Reliability/down time concerns
Limited credible data
Resale values
Capital investment and credit
Few early adopters
"Show me" mentality
4. Maximums often less
than 100% due to less than
universal applicability of a
particular technology at
full effectiveness
OEM Considerations:
•	Emissions standards
•	Return of investment
•	Warranty costs
3. Rapid increase in market
adoption once
technology is "proven"
2. Initial increases are modest, "follow-
the-leader"
1. Earliest market penetration from test fleets, on a trial basis
Time
"Simple"
	"Complex"
Figure 2-2 S-shape Market Penetration
2.4 Technology Principles - Class 4 to 8 Vehicles
2.4.1 Aerodynamics
The aerodynamic efficiency of heavy-duty vehicles has gained increasing interest in
recent years as fuel prices, competitive freight markets, and overall environmental awareness has
focused owners and operators on getting as much useful work out of every gallon of diesel fuel
as possible. Up to 25 percent of the fuel consumed by a line-haul tractor traveling at highway
speeds is used to overcome aerodynamic drag forces, making aerodynamic drag a significant
contributor to a Class 7 or 8 tractor's GHG emissions and fuel consumption.37 Because
aerodynamic drag varies by the square of the vehicle speed, small changes in the tractor
aerodynamics can have significant impacts on GHG emissions and fuel efficiency of that vehicle.
With much of their driving at highway speed, the benefits of reduced aerodynamic drag for Class
7 or 8 tractors can be significant.38
The common measure of aerodynamic efficiency is the coefficient of drag (Cd) or drag
area (CdA). The aerodynamic drag force (i.e., the force the vehicle must overcome due to air) is
a function of Cd, the area presented to the wind (i.e., the projected area perpendicular to the
direction of travel or frontal area), and the square of the vehicle speed. Cd values for today's
line-haul fleet typically range from greater than 0.80 for a classic body tractor to approximately
0.58 for tractors that incorporate a full package of widely, commercially available aerodynamic
features on both the tractor and trailer.
While designers of heavy-duty vehicles and aftermarket products try to aerodynamically
streamline heavy-duty vehicles, there are some challenges. Aerodynamic design must meet
practical and safety needs such as providing for physical access and visual inspections of vehicle
equipment. Since weight added to the vehicle can impact its overall fuel efficiency, GHG
emissions and, in limited cases, the amount of freight the vehicle can carry, aerodynamic design

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
and devices must balance the aerodynamic benefit with the contribution to the vehicle weight. In
addition, aerodynamic designs and devices must balance being as light and streamlined as
possible with in-use application durability to withstand the rigors a working freight vehicle
encounters while traveling or loading and unloading.
However, there are some macro and micro scale techniques that can be employed to
reduce vehicle drag such as reducing vehicle size, especially, the frontal area; smoothing the
shape to make it more aerodynamically efficient, thus reducing the Cd; and/or re-directing air to
prevent entry into areas of high drag (e.g., wheel wells) or to maintain smooth air flow in certain
areas of the vehicle. Reducing the size of the vehicle can reduce the frontal area; which reduces
the pressure building up on the lateral surface area exposed to the airflow. Improving the vehicle
shape may include revising the fore components of the vehicle such as rearward canting/raking
or smoothing/rounding the edges of the front end components (e.g., bumper, headlights,
windshield, hood, cab, mirrors) or integrating the components at key interfaces (e.g.,
windshield/glass to sheet metal) to alleviate fore vehicle drag. Finally, redirecting the air to
prevent areas of low pressure and slow moving air; thus eliminating areas where air builds
creating turbulent vortices and increasing drag. Techniques such as blocking gaps in the sheet
metal, ducting of components, shaping or extending sheet metal to reduce flow separation and
turbulence are methods being considered to direct air from areas of high drag (e.g., underbody,
tractor-trailer gap, underbody and/or rear of trailer).
The issue for heavy-duty vehicles is that the cab and/or passenger compartment is
designed for a specific purpose such as accommodating an inline cylinder engine or allowing for
clear visibility given the size of the vehicle. Consequently, a reduction in vehicle size and/or
frontal area may not be realistic for some applications. This also may necessitate an expensive,
ground-up vehicle redesign and, with a tractor model lifecycle of up to 20 years, may mean that a
mid-cycle tractor design is not feasible. In addition, the frontal area is defined by the shape
behind the cab so reducing just the cab frontal area/size reduction may not be effective. Thus,
this approach is something that may occur in a long-term timeframe of 10-20 years from today.
Instead, most heavy-duty tractor manufacturers have explored, or are exploring, the latter
two techniques in the short-term. Compared to previous generation tractors, every high roof
tractor today has a roof fairing directing air over the top of the cab, fuel tank/chassis fairings that
prevent side air from flowing underneath the vehicle, and cab side extenders that prevent flow
from being trapped in the tractor-trailer gap. As a compliance strategy for HD Phase 1, many
manufacturers refined the aerodynamic shape of their front end components and other
components (e.g., curving or further extending side extenders) resulting in efficiency difference
between pre- and post-HD Phase 1, model year tractor aerodynamic performance. Further,
manufacturers have developed new tractor designs that are taking advantage of sealing gaps in
sheet metal to redirect the flow and introducing some hard edges to induce turbulent flow on
certain surfaces to prevent premature flow separation and downstream turbulent flow. For HD
Phase 2, we anticipate manufacturers would continue to apply these techniques across their
models and continue to explore refinements and re-designs in other areas of the tractor.
In addition to tractor improvements, there has been growth in the market for trailer
aerodynamic devices encouraged by our successful SmartWay Partnership and Technology
Verification Program. These devices function similar to components on the tractor by preventing

-------
*** E. O. 12866 Review - Revised - Do Not Cite, Quote, or Release During Review ***
air intrusion into areas of the trailer prone to high aerodynamic drag including the tractor-trailer
gap, the trailer underhody, and the rear of the trailer as shown in Figure 2-3 and Figure 2-4
below.
520
480
440
400
360
320
I 260
0
1	240
Q
200
160
120
80
40
0


	1	1	1	
Bas e 1 i ne-N o-Anerno mete r
	



	



































Slight drag increase along entire vehicle














IT




































































I £


OOi
	1	

Q(
sr i
12 14 16 18
X location fml
Figure 2-3 Progression of Total Drag along a Typical Line-Haul Tractor-Trailer Vehicle
Figure 2-4 Low Pressure Regions Contributing To Aerodynamic Drag Along A Typical Line-Haul Tractor-
Trailer Vehicle
To address this, trailer front/nose devices are being used to round the front end and edges
of the trailer while also reducing the tractor-trailer gap; skirts on the side of the trailer prevent air
entering the underside of the trailer and becoming turbulent on the various underbody structure
components; and trailer aft/rear treatments reduce separation of air flow of the rear edge of the
trailer to reduce the large wake of turbulent air behind the trailer. Based on current SmartWay
Technology Verification, these devices can reduce fuel consumption from 1 to 9 percent,
depending on the technology, and if it is employed indi vidually or in combination.
As a result, we believe there is an opportunity within FID Phase 2 to promote continual
improvement of tractor aerodynamics and capitalize on the potential improvement that
aerodynamic trailer devices can provide for trailers, and for overall combination tractor-trailer
efficiency.

-------
*** E. O. 12866 Review — Revised - Do Not Cite, Quote, or Release During Review ***
2.4.2 Advanced Aerodynamic Concepts
The ITD Phase 2 standards will be fully phased in by the 2027 model year. This
represents a significant amount of time from today's action. As such, it is possible that by the
time the Phase 2 standards are implemented, the state of heavy-duty aerodynamic technology and
performance may have advanced significantly. Thus, there may be a need to have standards to
adequately address future tractor-trailer aerodynamic advances.
Accordingly, we are considering aerodynamic concepts that can achieve aerodynamic
performance beyond that of the aerodynamic-attributed improvements in the ITD Phase 2
standards. There are many approaches applicable to today's tractors and trailers that are not
considered in the HD Phase 2 standards and there is also ongoing advanced research aimed at
creating a completely new design paradigm for tractor-trailer combinations.
The advanced aerodynamic standards would not be required but would rather serve as a
marker for future aerodynamic concepts and/or as a metric for HD Phase 2 advanced/innovative
aerodynamic technologies.
2.4.2.1 Aerodynamic Improvements to Current Tractor-Trailer Combinations
Based on Existing Technology
2.4.2.1.1 Manufacturer Commercial Initiatives
In order to anticipate technology advancement, it is important to benchmark current
technology improvements based on today's tractors and trailers. A number of Class 8 tractor
OEM's have incorporated the technologies requested by their customers to improve fuel
economy and to meet the HD Phase 1 standards. These technologies include side skirts, boat
tails and roof fairings as well as some driver monitoring tools. Recently, Jack Roberts released
an article on the internet titled: "Photo, video: Western Star introduces re-designed on-highway
tractor." 39
Figure 2-5 Pictures of the Western Star Class 8 Tractors
In addition to providing photos and videos of Western Star's redesigned on-highway
tractor, the article describes a multitude of new features that define the new tractor. These
features include:

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
•	A new sweptback four piece bumper with an under bumper valance that contributes
to aerodynamic efficiency.
•	New halogen headlights that are optimized for aerodynamic performance and
excellent visibility.
•	A state-of-the-art visor specifically engineered to work with the impressive slope in
the hood's rear air ramp to direct airflow over the cab without an aerodynamic
penalty.
•	Roof and cab fairings that sweep back for tighter trailer gap and help direct air flow
over and around the trailer.
•	Optional chassis side fairings that reduce drag by up to 6 percent while still providing
easy access to batteries and DEF tank.
•	The Western Star Twin Force dual air intake, which feeds a massive centrally
mounted air filter to improve efficiency.
This example demonstrates that manufacturers are continuing to find ways to improve
tractors and are continually exploring concepts, such as those in used in the SuperTruck
initiative, to improve commercially-available products.
2.4.2.1.2 Supplier Research: SABIC Roof Fairing Technology and
Manufacturing
Developments in aerodynamics have long been assumed to yield advances in vehicle fuel
efficiency. SABIC Innovative Plastics US LLC (SABIC) evaluated a variety of injection
moldable thermoplastic roof fairing designs for a heavy tractor day cab to quantify efficiencies
that could be obtained through advanced aerodynamics. Computational Fluid Dynamic (CFD)
modeling was performed by Exa Corporation, an industry recognized leader in CFD. Multiple
designs exhibited significant reductions in drag compared to a baseline roof fairing (Figure 1 of
Figure 2-6). The baseline represented a top performing roof fairing on the market today. The
best performing SABIC concept (Figure 2-7) achieved a 5.8 percent reduction in drag and fuel
use compared to the baseline. Under the well-established 2:1 relationship between delta drag
and fuel use, the fuel efficiency improved by nearly 3 percent from the baseline design.
The design concept optimized the shape to manage the airflow over the vehicle and
enable reduced drag and increased fuel economy. Air channels - developed for injection
molding processes - limit the air stagnation on the front of the trailer as well as accelerate and
control the direction of the air flow. This innovative concept has been validated using state of
the art CFD methods. On vehicle tests are suggested to validate findings from these studies,
(from Matthew D. Marks, Senior Business Manager, Regulatory Automotive and Mass
Transportation, November 14, 2014).

-------
*** E. O. 12866 Review — Revised - Do Not Cite, Quote, or Release During Review ***
Figure 1: baseline roof fairing	Figure 2: SABIC concept roof fairing
Surface X-Force [dimless]
-0.300	-0150	O.OQO	0.150	0.300
Figure 2-6 Surface X-Force (dimensionless) on Baseline and SABIC Concept Roof Fairing
Aerodynamic (surface) force is the force exerted on a body whenever there is a relative
velocity between the body and the air. These plots represent this force in the direction of the
vehicle travel at highway speeds. Red indicates a 'pushing' of the vehicle rearward, while blue
indicates a 'pulling' of the vehicle forward.
Figure 3: SABIC concept roof Figure 4: SABIC concept roof
fairing showing directed airflow	fairing showing airflow detail
Figure 2-7 SABIC Concept Roof Fairing Operation
2.4.2.1.3 HD Phase 1 Research: External Active Grille Shutter Potential on
Heavy-Duty Tractors
During ITD Phase 1 aerodynamic assessment, we looked at several trends to understand
some of the aerodynamic trends such as removal of tractor chassis fairings and side extenders,
different tractor-trailer gap widths, and different trailer leading edge radii. However, one trend
of particular relevance to advanced aerodynamic improvements for current tractors is the case of
open versus closed grille.
We evaluated the open vs. closed grille trend in the full and reduced scale wind tunnel.
Below in Figure 2-8 is a picture of a 1/8'1' scale tractor model in the reduced scale wind tunnel
with the grille covered with aluminum tape to simulate a fully closed grille.

-------
*** E. O. 12866 Review — Revised - Do Not Cite, Quote, or Release During Review ***
Figure 2-8 Photo of l/8th Scale Model of a Tractor with the Front, External Grille
Covered With Aluminum Tape to Simulate A Closed Grill Configuration
Below in Table 2-1 and Table 2-2 are the results of our open versus closed grille evaluations in
the full and reduced scale wind tunnel separately. The tables provide the deltas for an open grille
CdA minus the closed grille C\iA; where the CdAs have been corrected for blockage, in the case
of the full scale wind tunnel, and normalized for differences in measured frontal area between the
full and reduced scale wind tunnels using a nominal frontal area of 10.4 m2 (111.95 in2). For the
full scale wind tunnel, only one tractor OEM was tested. In contrast, for the reduced scale wind
tunnel, three tractor OEMs were tested.
Table 2-1 Full Scale Wind Tunnel Results for Open verses Close Grille Configurations
TRACTOR
MODEL
DELTA WACdA
r?55MPH
% DELTA CdA VS. OPEN GRILLE
CdA
1
0.03
0.60%
Table 2-2 Reduced Scale Wind Tunnel Results for Open versus Close Grille Configurations
TRACTOR
MODEL
DELTA WACdA
SiSMPH
% DELTA CdA VS. OPEN GRILLE
CdA
A
0.10
1.69%
B
0.12
1.89%
C
0.09
1.45%
Based on the data in these tables, there is a potential wind-average drag improvement of
0.6 percent to 1.45 percent from closing off the external, front grille of the tractor. This indicates
the potential of active grille shutter systems on heavy-duty tractors. These systems are currently
being applied on light duty vehicles behind the external grille to improve aerodynamics.
However, a recent SAE paper determined that the optimal position for active grille shutter
systems was the external grille flush with the vehicle sheet metal.40 This technique could be
implemented on the external grille designs for current-design, heavy-duty tractors as well.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.4.2.1.4 National Research Council of Canada Historical Research on
Improving Heavy-Duty Tractors
The National Research Council of Canada (NRC-Can) performed an assessment of the
drag effect of various tractor components41 and found the following in Table 2-3.
Table 2-3 Reduced Scale Wind Tunnel Results for Open versus Close Grille Configurations
COMPONENT
DELTA
WACdA
OEM Side Mirrors
-0.156
OEM Fender Mirrors
-0.098
Wheel Covers (Tractor and Trailer)
0.020
Tractor Drive Axle Wrap-Around Splash Guards
0.049
Roof Fairing Rear-Edge Filler
0.137
Based on this table, there is the potential to improve tractor aerodynamics by 0.206
WACdA) with the addition of wheel covers, drive axle wrap around splash guards, and roof
fairing rear edge filler, and up to 0.460 if the OEM side and fender mirrors are replaced with a
camera system, as suggested by the study, and combined with the wheel covers, drive axle wrap
around splash guards, and roof fairing rear edge filler. Therefore, considering the current wind-
average drag performance of current heavy-duty tractors, this study demonstrates the possibility
to improve tractors an additional ~1 percent with some simple changes.
2.4.2.2 Aerodynamic Improvements to Current Tractor-Trailer Combinations
Based on Complete Vehicle Redesign
This section contains summaries of ongoing work from various DOE efforts as well as
individual efforts such as Airflow Truck Company to develop improved aerodynamic Class 8
vehicles. In addition to aerodynamics, there are other technologies such as driver awareness and
ability to drive for maximum fuel economy with increased aerodynamics. Overall it is expected
that the research being performed over the next year or two will reveal drastic improvements in
CdA and fuel economy. DOE's Lawrence Livermore National Laboratory is also looking at the
aerodynamics of tankers.
2.4.2.2.1 Collaborative, Government-Industry Advanced Aerodynamic
Research: SuperTruck Program
DOE's SuperTruck project is one of several initiatives which is a public-private initiative
to stimulate innovation in the trucking industry through sponsorship from government agencies,
companies, national laboratories and universities. DOE's Vehicle Technologies Program
provided matching funds to the program. Four programs basically involved all major vehicle
and engine manufacturers were awarded by DOE under SuperTruck program. They are
Cummins, Daimler, Navistar, and Volvo. Cummins was teamed up with Peterbilt on the vehicle
side of the program.
The goal of the SuperTruck Initiative was to achieve 50 percent freight efficiency
improvement with 30 percent from vehicle and 20 percent from engine compared to a 2009

-------
*** E. O. 12866 Review — Revised - Do Not Cite, Quote, or Release During Review ***
vehicle. This means that it require development of a tractor that could meet or exceed 10 mpg -
where tractors at this point are averaging between 5.5 and 6.5 mpg. Advances in engines,
aerodynamics and more helped the tractor project increase its fuel economy. The SuperTruck
objectives included development and demonstration of a highly efficient and clean diesel engine,
an advanced waste heat recovery system, an aerodynamic tractor and trailer combination and a
lithium ion battery auxiliary power unit, to reduce engine idling.
Significant progress has been made since the initiation of this program in 2010. Two programs
are particularly worth noting. They are the Cummins-Peterbilt and Daimler programs. The
Cummins-Peterbilt SuperTruck project team was the first to report and demonstrate a
SuperTruck vehicle withl0.7mpg. Details of the SuperTruck are given in four videos on the
todaystrucking.com website.42 Aerodynamic features of the tractor include the following:
airflow into the engine compartment (through the front bumper, through the radiator and under
the vehicle), less clearance between the road and the bottom of the tractor (rubber skirt under
steps), close gaps on tractor/trailer (between hood and bumper, etc.), minimized gap between the
trailer and tractor with a ball and socket design, full trailer skirt, roof fairing, smaller mirrors,
minimized gap between wheels and wheel wells, wheel covers, boat tail, air foil on rear bumper
design, single wide tires, and perforated mud flaps that allow air to bypass through them and
reduce drag. A picture of this truck based on a Peterbilt tractor is shown in Figure 2-9 below.
Eaton Corp, also part of the Cummins-Peterbilt SuperTruck project team, contributed
technologies including the design, development and prototyping of an advanced automated
transmission that facilitated reduced engine-operating speeds. Cummins and Eaton jointly
designed shift schedules and other features to yield further improved fuel efficiency.
Even with the additi on of these aerodynamic features, overall the tractor mass was
reduced by over 1,300 lbs. The article states that the CFD analysis of the tractor showed a 50
percent reduction in drag and with a 2:1 drag reduction the aero improvements resulted in a 25
percent improvement in fuel economy. In the 300 mile test course shown on the video, it was
stated that the tractor achieved 10.7-11.1 mpg.
	
)>
~T 1 hm

su

• I <

Figure 2-9 Peterbilt SuperTruck Concept (Picture from: http://www.peterbilt.com/about/media/2014/396/)
This effort represents the first step in the evolution of improving the aerodynamic
efficiency of tractor-trailer by radically redesigning today's tractor-trailer combination, as a
wholly integrated system rather each component, tractor and trailer, independently.

-------
*** E. O. 12866 Review — Revised - Do Not Cite, Quote, or Release During Review ***
Just one year later after Cummins' demonstration, Daimler demonstrated their own
SuperTruck vehicle with 12.2 mpg as showed in Figure 2-10.
Figure 2-10 Daimler SuperTruck Vehicle (picture from: http://freightlinersupertmck.eom/#main)
The key enabling technology on the aero side in this vehicle includes, but is not limited to, full
tractor aero with cab/sleeper, underbody, drive wheel fairing, mirror cam, steer wheel, and full
side extender. In addition, this vehicle also includes a 50 percent BTE DDI 1 Engine with WHR,
predictive hybrid controller, predictive engine controller, new final drive active oil management
with high efficiency gear oil, lightweight aluminum frame and cross members, ultra-light weight
air suspension, advanced load shift with 6x2 axle, solar reflective paint, and enhanced Trailer
aerodynamics. More detailed features on this Daimler truck can be seen in their DOE report34.
2.4.2.2.2 Government Sponsored Advanced Aerodynamic Research: Lawrence
Livermore National Laboratory
Lawrence Livermore National Laboratory's (LLNL) Kambiz Salari presented
information at the 2014 DOE Annual Merit Review on "DOE's Effort to Improve Heavy Duty
Vehicle Fuel Efficiency through Improved Aerodynamics." A joint project with Wabash,
Navistar, Michelin, Safeway, Frito Lay, Praxair, Freight Wing Inc., ATDynamics, Kentucky
Trailer and Spirit was funded in 2013 and 2014. The objective was to develop a new integrated
tractor-trailer design from ground up by first, designing the first generation of an integrated
tractor-trailer geometry called Generic Speed Form one (GSF1) and second, performing wind
tunnel tests of selected aero devices for tractor-trailers and tankers to improve fuel efficiency.
The goal was to reduce aerodynamic drag of Class 8 tractor-trailers by approximately 25 percent
leading to a 10-15 percent increase in fuel efficiency at 65 mph. In addition, the group
developed an aerodynamic tractor-trailer prototype designed to achieve 50 percent reduced
aerodynamic drag as shown in Figure 2-11. This effort represents the next generation of tractors
and trailers: a completely redesigned, fully integrated, optimized shape for the tractor-trailer
combination.

-------
*** E. O. 12866 Review — Revised - Do Not Cite, Quote, or Release During Review ***
Tractor-trailer integration is the next step in
achieving a radical improvement in fuel economy
> 50% aerodynamic drag
reduction compared to
heavy vehicles on the
road today
Lawrence Uvwmwe National LabcrHay
Figure 2-11 Pictures Showing Future Heavy-Duty Tractor Trailer Concept to Achieve >50 percent
Aerodynamic Improvement for Class 8 Line Haul Heavy-Duty Vehicles
2.4.2.2.3 Independent Advanced Aerodynamic Research: Airflow Truck
Company Bullet Truck Concept
In addition to the work being performed by the OEMs and consortiums mentioned above,
there are also independent commercial initiatives underway to radically redesign the tractor-
trailer combination similar to the concept by Lawrence Livermore National Laboratories
discussed above.
The Class 8 tractor and trailer modifications in Figure 2-12 were designed, built, and
tested in 2012 by Mr. Robert Sliwa of the Airflow Truck Company. Mr. Sliwa built his first
aerodynamic tractor in the 1983 when he was an owner-operator. After that, Mr. Sliwa became
interested in computers and used his computer background along with his truck driver and race
car driver experience to create the Bullet Truck. His current design is described at
www.airflowtruck.com and his tractor design modifications are similar in appearance to the
bullet looking trains used in Europe. The tractor uses a 1999 engine and the test was conducted
in a manner in which the tractor was driven at 55 mph by an experienced driver throughout its
test while loaded at 65,000 lbs from Newington, Connecticut to Tracy, California.
The website shows that the vehicle achieved 13.4 mpg during this trip that included
traveling through the Rocky Mountains. CFD analyses of the design after the vehicle was built
found a modest decrease in CdA, thus giving credence to the design work under the hood (most
of which are outlined at airflowtruck.com) and driving techniques. Several new technologies
were developed during this work which included retractable tractor steps, all electric air
conditioning, crankshaft mounted cooling fan, computer-controlled fan hub, waterless engine
coolant, reduced engine parasitic losses, full tractor and trailer side skirts, 4 axle ATIS, and an
engine feedback information display.

-------
*** E. O. 12866 Review — Revised - Do Not Cite, Quote, or Release During Review ***
Figure 2-12 Figure of the Bullet Truck by Airflow Truck Company9
AirFlow has designed and is currently building the third prototype (Proof of Concept) of
a "Hyper Fuel Mileage, Ultra Low GHG Emissions, and readable Class 8 heavy duty truck"
called the StarShip. The StarShip is a Class 8 heavy duty truck tractor that will be mated with a
new 2016 Strick 53' dry van trailer, which is typical of an over-the-road freight hauling trailer.
AirFlow has further modified the stock trailer to be much more aerodynamic than when it left the
Strick factory. There is also a full array of solar panels on the trailer roof. This solar array will
charge batteries mounted on the tractor during the day to enable to provide electric Fleat,
Ventilation, and Air Conditioning (HVAC) to the cab for driver comfort while traveling down
the roadway, and when the driver is engaged in federally mandated rest and safety breaks.
Utilizing a proprietary all-electric FIVAC system will allow the StarShip to reduce GFIG
emissions and increase fuel efficiency by completely removing the diesel engine-driven air
conditioning compressor, and its associated engine parasitic efficiency losses. It will also allow
the StarShip to automatically and periodically turn off its diesel engine belt-driven 300 amp
alternator, further saving fuel and further reducing GHG emissions. These aerodynamic, solar,
and hybridized component improvements will further reduce GHG vehicle emissions and vastly
increase fuel efficiency.
The latest proof of concept vehicle, the StarShip, is due to be completed in Q3 2016 and
will begin its local and regional road testing then. The design of the StarShip continues to be
refined. The StarShip utilizes an experimental 2017 EPA low-emissions certified, six cylinder,
400 horsepower diesel-fueled Cummins engine to power the vehicle. The engine is certified to
produce air pollutants and GHG emissions in an amount significantly below the current 2013
standard. Future versions of the StarShip model may include a hybrid (diesel engine/electric
motor) and/or a purely electric propulsion unit, powered only with an onboard battery bank,
similar to a Tesla automobile.

-------
*** E. O. 12866 Review — Revised - Do Not Cite, Quote, or Release During Review ***
iductng fo Airflow iecirSK®, iH» wwcf'i moir innovolbv nuck In '!» ckas. Wlh
Figure 2-13 StarShip Advertisement by Shell Uotella and Airflow Truck Company
2.4.3 Tires
2.4.3.1 Improved Rolling Resistance
Research indicates that a tire's contribution to overall vehicle fuel efficiency is
approximately proportional to the vehicle weight on it.43 Energy loss associated with tires is
mainly due to deformation of the tires under the load of the vehicle, known as hysteresis, but
smaller losses result from aerodynamic drag, and other friction forces between the tire and road
surface and the tire and wheel rim. Collectively the forces that result in energy loss from the
tires are referred to as rolling resistance. Tires with higher rolling resistance lose more energy,
thus using more fuel and producing more CO2 emissions in operation, while tires with lower
rolling resistance lose less energy, and use less fuel, producing less CO2 emissions in operation.
A tire's rolling resistance is a factor considered in the design of the tire, and is affected by
the tread and casing compound materials, the architecture of the casing, tread design, and the tire
manufacturing process. It is estimated that 35 to 50 percent of a tire's rolling resistance is from
the tread and the other 50 to 65 percent is from the casing.43 Tire inflation can also impact
rolling resistance in that under-inflated tires can result in increased deformation and contact with
the road surface. In addition to the effect on CO2 emissions and fuel consumption, these design
and use characteristics of tires also influence durability, traction (both wet and dry grip), vehicle
handling, ride comfort, and noise. Tires that have higher rolling resistance are likely designed to
address one or more of these other tire attributes.
EPA's SmartWay program identified test methods and established criteria to designate
certain tires as having lower rolling resistance (LRR) for use in the program's emissions tracking
system, verification program, and SmartWay vehicle specifications. To measure a tire's
efficiency, the vertical load supported by the tire must be considered, because rolling resistance
is a function of the load on a tire. EPA uses a tire's rolling resistance coefficient (CRR) to
characterize LRR tires. CRR is measured using the ISO 28850 test method (see 40 CFR

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
1037.520(c),) and reported as the rolling resistance force over vertical load (kg/metric ton).
Differences in rolling resistance of up to 50 percent have been identified for tires designed to
equip the same vehicle.44
LRR tires are commercially available from most tire manufacturers and can be applied to
vehicles in all MD/HD classes. According to an energy audit conducted by Argonne National
Lab, tires were shown to be the second largest contributor to energy losses for a Class 6 delivery
truck at 50 percent load and speeds up to 35 mph (a typical average speed of urban delivery
vehicles).45 For Class 8 tractor-trailers, the share of vehicle energy required to overcome rolling
resistance is estimated at nearly 13 percent.46
NHTSA, EPA, and ARB met with stakeholders from the tire industry (Bridgestone,
Continental, Cooper, Goodyear, and Michelin) in 2014 to discuss the next generation of LRR
tires for the Phase 2 timeframe for all segments of Class 2b-8 vehicles, including trailers.
Manufacturers discussed forecasts for rolling resistance levels and production availability in the
Phase 2 timeframe, as well as their plans for improving rolling resistance performance while
maintaining other performance parameters such as traction, handling, wear, mass reduction,
retreadability, and structural durability.
The meetings included specific discussions of the impacts of the current generation of
LRR tires on vehicle stopping distance and handling. Manufacturers indicated no known safety
disbenefit in the current on-road fleet from use of LRR tires. While the next generation of tires
may require some tradeoffs in wear performance and costs over the next 10 years to achieve
better tire rolling resistance performance, manufacturers said they will not trade off safety for
performance. They also emphasized that keeping tires inflated (through proper maintenance or
automatic systems) was the best way to assure long term fuel efficiency and safety during
vehicle operation.
2.4.3.2 Wide Base Singles
Low rolling resistance tires can be offered for dual assembly tires and as wide base
singles (WBS). Wide base singles are primarily intended for combination tractor-trailers, but
some vocational vehicles are able to accommodate them. In the early years of this technology,
some states and local governments restricted use of WBS, but many of these restrictions have
since been lifted. As of December 2010, NACFE reports that there is virtual acceptance in North
America with only a few provinces in Canada that disallow or require special permitting for the use
of wide base tires.47 A wide base single is a larger tire with a lower profile. The common wide
base single sizes include 385/65R22.5, 425/65R22.5, 445/65R22.5, 435/50R22.5 and
445/50R22.5. Generally, a wide base single tire has less sidewall flexing compared to a dual
assembly and therefore less hysteresis occurs. Compared to a dual tire assembly, wide base
singles also produce less aerodynamic resistance or drag. Wide base singles can contribute to
improving a vehicle's fuel efficiency through design as a low rolling resistance tire and/or
through vehicle weight reduction.
According to one study, the use of fuel efficient wide base singles can reduce rolling
resistance by 3.7 to 4.9 percent compared to the most equivalent dual tire.48 An EPA study with
a tractor-trailer demonstrated an improvement in fuel consumption of 6 percent at 55 mph on the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
highway, 13 percent at 65 mph on the highway and 10 percent on a suburban loop49 using wide
base singles on the drive and trailer axles. EPA attributed the fuel consumption improvement to
the reduction in rolling resistance and vehicle weight reduction from using wide base singles. In
2008 the Department of Energy (DOE) compared the effect of different combinations of tires on
the fuel efficiency of Class 8 tractors. The data collected based on field testing indicates that
tractors equipped with wide base singles on the drive axle experience better fuel efficiency than
tractors equipped with dual tires, independent of the type of tire on the trailer.50 This study in
particular indicated a 6.2 percent improvement in fuel efficiency from wide base singles.
There is also a weight savings associated with wide base singles compared to dual tires.
Wide base singles can reduce a tractor and trailer's weight by as much as 1,000 lbs. when
combined with aluminum wheels. Bulk haulers of gasoline and other liquids recognize the
immediate advantage in carrying capacity provided by the reduction in the weight of tires and
have led the transportation industry in retrofitting their tractors and trailers.51
New generation wide base singles, which were first introduced in 2000, are designed to
replace a set of dual tires on the drive and/or trailer positions. They are designed to be
interchangeable with the dual tires without any change to the vehicle52. If the vehicle does not
have hub-piloted wheels, there may be a need to retrofit axle components.51'53 In addition to
consideration of hub/bearing/axle, other axle-end components may be affected by use of wide
base singles. To assure successful operation, suitable components should be fitted as
recommended by the vehicle manufacturer.54
Current wide base singles are wider than earlier models and legal in all 50 states for a 5-
axle, 80,000 GVWR truck 48 Wide base singles meet the "inch-width" requirements nationwide,
but are restricted in certain states up to 17,500 lbs. on a single axle at 500 lbs/inch width limit,
and are not allowed on single axle positions on certain double and triple combination vehicles52.
An inch-width law regulates the maximum load that a tire can carry as a function of the tire
width. Typically wide base singles are optimized for highway operation and not for city or
on/off highway operation. However, newer wide base singles are being designed for better scrub
resistance, which would allow an expansion of their use. The current market share of wide base
singles in combination tractor applications is 5 percent and the potential market is all
combination tractors.48 New generation wide base singles represent an estimated 0.5 percent of
the 17.5 million tires sold each year in the U.S.52
2.4.3.3 Tire Pressure Systems
Proper tire inflation is critical to maintaining proper stress distribution in the tire, which
reduces heat loss and rolling resistance. Tires with reduced inflation pressure exhibit more
sidewall flexing and tread shearing, resulting in greater rolling resistance than a tire operating at
its optimal inflation pressure. Bridgestone tested the effect of inflation pressure and found a 2
percent variation in fuel consumption over a 40 psi range.43 Tractor-trailers operating with all
tires under-inflated by 10 psi have been shown to increase fuel consumed by up to 1 percent.55
Tires can gradually lose pressure from small punctures, leaky valves or simply diffusion through
the tire casing. Changes in ambient temperature can also have an effect on tire pressure. Trailers
that remain unused for long periods of time between hauls may experience any of these

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
conditions. To achieve the intended fuel efficiency benefits of low rolling resistance tires, it is
critical that tires are maintained at the proper inflation pressure.
Although most truck fleets understand the importance of keeping tires properly inflated,
it is likely that a substantial proportion of trucks on the road have one or more underinflated tires.
An industry survey conducted in 2002 at two truck stops found that fewer than half of the tires
checked were within 5 pounds per square inch (psi) of their recommended inflation pressure.
Twenty-two percent of the vehicles checked had at least one tire underinflated by at least 20 psi,
and 4 percent of the vehicles were running with at least one flat tire, defined as a tire
underinflated by 50 psi or more. The survey also found mismatches in tire pressure exceeding 5
percent for dual tires on axle ends.56
A commercial vehicle tire condition study conducted by the Federal Motor Carrier Safety
Administration (FMCSA) in 2003 found similar indicators of poor tire inflation pressure
maintenance in commercial fleets. The FMCSA concluded that only 44 percent of all tires on
commercial vehicles were inflated within 5 psi of the recommended pressure, while over 7
percent of all tires in operation on commercial vehicles were underinflated by at least 20 psi. It
was also determined that the rates of tires used in dual assemblies that differed in pressure by
more than 5 psi was approximately 20 percent for tractor duals and 25 percent for trailer duals.
Finally, the FMCSA concluded that there were significant differences in tire inflation
maintenance practices between private and for-hire fleets, smaller and larger fleets, and local bus
and motor coach fleets.57
If drivers or fleets are not diligent about checking and attending to under-inflated tires,
the trailer may have much higher rolling resistance and much higher CO2 emissions and fuel
consumption. Proper tire inflation pressure can be maintained with a rigorous tire inspection and
maintenance program and EPA provides information on proper tire inflation pressure through its
SmartWay program.58 Tire pressure monitoring (TPM) and automatic tire inflation (ATI)
systems are designed to address under-inflated tires. Both systems alert drivers if a tire's
pressure drops below its set point. TPM systems monitor the tires and require user-interaction to
reinflate to the appropriate pressure. Yet unless the vehicle experiences a catastrophic tire
failure, simply alerting the driver that a tire's pressure is low may not necessarily result in action
to correct the problem. A driver may continue driving to their final destination before addressing
the tires, resulting in many miles of driving with improperly inflated tires. Current ATI systems
take advantage of trailers' air brake systems to supply air back into the tires (continuously or on
demand) until a selected pressure is achieved. In the event of a slow leak, ATI systems have the
added benefit of maintaining enough pressure to allow the driver to get to a safe stopping area.59
Estimates of the benefits of ATI systems vary depending on the base level of
maintenance already performed by the driver or fleet, as well as the number of miles the trailer
travels. Vehicles that are well maintained or that travel fewer miles would experience less
benefits from ATI systems compared to vehicles that log many miles or have a history of driving
with poorly inflated tires. The agencies believe ATI systems can provide a CO2 and fuel
consumption benefit to most tractors and trailers. Drivers and fleets that diligently maintain their
tires will spend less time and money to inspect each tire knowing that they are properly inflated.
Vehicles that have lower annual VMT due to long periods between uses would be less

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
susceptible to low tire pressures when they resume activity. Vehicles with high annual VMT
would experience the fuel savings associated with consistent tire pressures.
2.4.3.4 Retreaded Tires
The tread life of a tire is a measure of durability and some tires are designed specifically
for greater durability. Commercial vehicle tires are designed to be retreaded, a process in which
a new tread is bonded to the tire casing. The original tread of a tire will last anywhere from
100,000 miles to over 300,000 miles, depending on vehicle operation, original tread depth, tire
axle position, and proper tire maintenance. Retreading can extend the tire's useful life by
100,000 miles or more.60 In 2005, the Tire Industry Association estimated that approximately
17.6 million retreaded truck tires were sold in North America61.
All of the top commercial vehicle tire manufacturers are involved in tire retread
manufacturing. Bridgestone Bandag Tire Solutions accounts for 42 percent of the domestic
retreaded vehicle tire market with its Bandag retread products; Goodyear Tire and Rubber
Company accounts for 28 percent, mostly through its Wingfoot Commercial Tire Systems;
Michelin Retread Technologies Incorporated, with Megamile, Oliver, and Michelin retread
products, accounts for 23 percent. Other tire companies like Continental and independent retread
suppliers like Marangoni Tread North America (which also produces the Continental
"ContiTread" retread product) make up the remaining 7 percent.62 The retreading industry itself
consists of hundreds of retreaders who sell and service retreaded tires, often (but not always)
using machinery and practices identified with one of the major retread producers. There are
about 800 retread plants in North America.63 The top 100 retreaders in the U.S. retread 47,473
truck tires per day.
To maintain the quality of the casing and increase the likelihood of retreading, a tire
should be retreaded before the tread depth is reduced to its legal limit. At any time, steer tires
must have a tread depth of at least 4/32 of an inch and other tires, including drive tires and trailer
tires, must have a tread depth of at least 2/32 of an inch (49 CFR 393.75). Trucking fleets often
retread tires before tire treads reach this minimum depth in order to preserve the integrity of the
tire casing for retreading. If the casing remains in good condition, a truck tire can be safely
retreaded multiple times. Heavy truck tires in line haul operation can be retread 2 to 3 times and
medium-duty truck tires in urban use can be retread 5 or more times.64 To accommodate this
practice, many commercial vehicle tire manufacturers warranty their casings for up to five years,
excluding damage from road hazards or improper maintenance.
To protect the casing, a steer tire is generally retreaded once the tread is worn down to
6/32 of an inch and a drive tire is retreaded once the tread is worn down to 8/32 of an inch.65
Tires used on Class 8 vehicles are retreaded as many as three times.
Both the casing and the tread contribute to a tire's rolling resistance. It is estimated that
35 to 50 percent of a tire's rolling resistance is the result of the tread. Differences in drive tire
rolling resistance of up to 50 percent for the same casing with various tread compounds have
been demonstrated. For example, a fuel efficient tread (as defined by the manufacturer) was
added to two different casings resulting in an average increase in rolling resistance of 48 percent.
When a nonfuel efficient tread (also defined by the manufacturer) was added to the same casings,

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
the rolling resistance increased by 125 percent on average. This characterizes the effect of the
tread on the rolling resistance of a tire.
Because tires can be retreaded multiple times, changes in the casing due to wear, damage
and material aging may impact rolling resistance to a greater degree than would occur in an
original tire. Additionally, as evidenced above, if a tread compound different than the original
tread is used, a retreaded tire can have higher or lower rolling resistance than the original tire.
Since the agencies have no way of knowing whether the rolling resistance of retreaded tires will
be higher or lower than the rolling resistance of the original tires, we similarly have no way of
knowing whether low rolling resistance tire benefits will continue to accrue for a vehicle's entire
lifetime.
2.4.4 Transmissions
Transmissions are a significant vehicle component. They are part of the drivetrain, which
also includes axles and tires. Ways to improve transmissions include electronic controls, shift
strategy, gear efficiency, and gear ratios. The relative importance of having an efficient
transmission increases when vehicles operate in conditions with a higher shift density. Each
shift represents an opportunity to lose speed or power that would have to be regained after the
shift is completed. Further, each shift engages gears that have their own inherent inefficiencies.
Optimization of vehicle gearing to engine performance through selection of transmission
gear ratios, final drive gear ratios and tire size can play a significant role in reducing fuel
consumption and GHGs. Optimization of gear selection versus vehicle and engine speed
accomplished through driver training or automated transmission gear selection can provide
additional reductions. The 2010 NAS report found that the opportunities to reduce fuel
consumption in heavy-duty vehicles due to transmission and driveline technologies in the 2015
time frame ranged between 2 and 8 percent.66
The design goal is for the transmission to deliver the needed power to the vehicle while
maintaining engine operation within the engine's "sweet spot" for most efficient operation.
Truck and chassis manufacturers today offer a wide range of tire sizes, final gear ratios and
transmission choices so that owners can work with application engineers to specify an optimal
combination given the intended vehicle service class and other performance needs.
2.4.4.1 Optimizing Number of Gears and Gear Ratios
Manufacturers of light and medium heavy-duty vehicles can choose to replace 6-speed
transmissions with 8-speed or more automatic transmissions. Additional ratios allow for further
optimization of engine operation over a wider range of conditions, but this is subject to
diminishing returns as the number of speeds increases. As additional planetary gear sets are
added (which may be necessary in some cases to achieve the higher number of ratios), additional
weight and friction are introduced. Also, the additional shifting of such a transmission can be
perceived as bothersome to some consumers, so manufacturers need to develop strategies for
smooth shifts.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The Phase 1 rulemaking projected that 8-speed transmissions could incrementally reduce
fuel consumption by 1 to 3 percent from a baseline 6-speed automatic transmission over some
test cycles. The SwRI report uses 2 to 3 percent fuel consumption reduction when replacing 6-
speed baseline automatic transmissions with improved 8-speed automatic transmissions. Chapter
2.9 of the RIA outlines the agencies' updated analysis that takes into account public comments
on the proposal.
2.4.4.2	Gear Efficiencies
As described elsewhere for axles and engines, the efficiency of gears can be improved by
reducing friction and minimizing mechanical losses. This can be done by reducing the friction
between the two gears in contact. This friction is reduced mainly by improving the surface finish
of the gears. The other way of doing is by reducing the amount of distance the gear faces are
sliding against each other.
2.4.4.3	Shift Strategies
Calibrating the transmission shift schedule to upshift earlier and quicker, and to lock up
or partially lock up the torque converter under a broader range of operating conditions can
reduce fuel consumption and CO2 emissions. However, this operation can result in a perceptible
degradation in noise, vibration, and harshness. The degree to which NVH can be degraded
before it becomes noticeable to the driver is strongly influenced by characteristics of the vehicle,
and although it is somewhat subjective, it always places a limit on how much fuel consumption
can be improved by transmission control changes.
During operation, an automatic transmission's controller manages the operation of the
transmission by scheduling the upshift or downshift, and locking or allowing the torque
converter to slip based on a preprogrammed shift schedule. The shift schedule contains a
number of lookup table functions, which define the shift points and torque converter lockup
based on vehicle speed and throttle position, and other parameters such as temperature.
Aggressive shift logic can be employed in such a way as to maximize fuel efficiency by
modifying the shift schedule to upshift earlier and inhibit downshifts under some conditions,
which reduces engine pumping losses and engine friction. The application of this technology
does require a manufacturer to confirm that drivability, durability, and NVH are not significantly
degraded.
A torque converter is a fluid coupling located between the engine and transmission in
vehicles with automatic transmissions and continuously-variable transmissions (CVT). This
fluid coupling allows for slip so the engine can run while the vehicle is idling in gear (as at a stop
light), provides for smoothness of the powertrain, and also provides for torque multiplication
during acceleration, and especially launch. During light acceleration and cruising, the inherent
slip in a torque converter causes increased fuel consumption, so modern automatic transmissions
utilize a clutch in the torque converter to lock it and prevent this slippage. Fuel consumption can
be further reduced by locking up the torque converter at lower vehicle speeds, provided there is
sufficient power to propel the vehicle, and noise and vibration are not excessive. If the torque
converter cannot be fully locked up for maximum efficiency, a partial lockup strategy can be
employed to reduce slippage. Early torque converter lockup is applicable to all vehicle types

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
with automatic transmissions. Some torque converters would require upgraded clutch materials
to withstand additional loading and the slipping conditions during partial lock-up. As with
aggressive shift logic, confirmation of acceptable drivability, performance, durability and NVH
characteristics would be required to successfully implement this technology.
2.4.4.4 Architectures
The manual transmission architecture has traditionally been considered the most efficient
architecture since it did not experience the losses inherent in a torque converter required on a
traditional automatic transmission (a traditional automatic transmission being a transmission with
fully automated shifting and using a hydraulic lock-up torque converter for smooth vehicle
launching from a stop). However, this traditional understanding has been called into question as
advances in electronics and computer processing power allow for more efficiency from a manual
transmission architecture with fully automated shifting. The two primary manual transmission
architectures employing automated shifting are the automated manual transmission (AMT) and
the dual-clutch transmission (DCT). When implemented well, these mechanically more efficient
designs could inherently provide better fuel efficiency and lower greenhouse gas emissions than
conventional torque converter automatic transmission designs and, potentially, even fully manual
transmissions. These transmissions offer the inherently lower losses of a manual transmission
with the efficiency and shift quality advantages of electronic controls. The lower losses stem
from the elimination of the conventional lock-up torque converter, and a greatly reduced need for
high pressure hydraulic circuits to hold clutches to maintain gear ratios (in automatic
transmissions).
2.4.4.4.1	AMT
An AMT is mechanically similar to a conventional manual transmission, but shifting and
launch functions are automatically controlled by electronics. The term AMT generally refers to
a single clutch design (differentiating it from a dual-clutch transmission, or dual-clutch AMT,
described below) which is essentially a manual transmission with automated clutch and shifting.
Because of shift quality issues with single-clutch designs, dual-clutch designs are more common
in light-duty applications where driver acceptance is of primary importance. In the HD sector,
shift quality remains important but is less so when compared to light-duty. As a result, the
single-clutch AMT architecture can be an attractive technology for HD vehicles.
2.4.4.4.2	DCT
A DCT uses separate clutches (and separate gear shafts) for the even-numbered and the
odd-numbered gears. In this way, the next expected gear is pre-selected thereby allowing for
faster and smoother shifting. For example, in a 6 speed DCT, if the vehicle is accelerating in
third gear, the shaft with gears one, three and five has gear three engaged and is transmitting
power to the wheels. The shaft with gears two, four, and six is idle but has gear four engaged.
When a shift is required, the controller disengages the odd-gear clutch while simultaneously
engaging the even-gear clutch, thus making a smooth shift. If, on the other hand, the driver
slows the vehicle instead of continuing to accelerate, the transmission would have to change to
second gear on the idling shaft to anticipate a downshift. This shift can be made quickly on the
idling shaft since there is no torque being transferred on it.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
There are variations of the DCT design, with some having wet clutches and some dry
clutches, and more recent versions that incorporate a torque converter similar to but smaller than
the torque converter of a traditional automatic transmission. The wet clutch designs offer a
higher torque capacity that comes from the use of a hydraulic system that cools the clutches.
Wet clutch systems are also less efficient than dry clutch systems due to the losses associated
with the hydraulic pumping. They also are more costly due to the hydraulics.
2.4.4.5 Hybrid Powertrain Systems
The industry is currently developing many variations of hybrid powertrain systems. The
fully integrated hybrids developed to date have seen fuel consumption and CO2 emissions
reductions between 20 and 50 percent in the field where they are used in high kinetic intensity
applications. However, there are still some key issues that are restricting the penetration of
hybrids, including overall system cost, battery technology, and lack of cost-effective electrified
accessories.
A hybrid vehicle is a vehicle that combines two significant sources of propulsion energy,
where one uses a consumable fuel (like diesel), and one is rechargeable (during operation, or by
another energy source). Hybrid technology is well established in the U.S. light-duty market,
some manufacturers have been producing heavy-duty hybrid models for many years, and others
are looking to develop hybrid models in future years.
Hybrids reduce fuel consumption through three major mechanisms:
The internal combustion engine can be optimized (through downsizing, modifying
the operating cycle, or other control techniques) to operate at or near its most
efficient point more of the time. Power loss from engine downsizing can be
mitigated by employing power assist from the secondary power source.
Some of the energy normally lost as heat while braking can be captured and
stored in the energy storage system for later use.
The engine is turned off when it is not needed, such as when the vehicle is
coasting or when stopped.
Hybrid vehicles utilize some combination of these three mechanisms to reduce fuel
consumption and CO2 emissions. The effectiveness of fuel consumption and CO2 reduction
depends on the utilization of the above mechanisms and how aggressively they are pursued. One
area where this variation is particularly prevalent is in the choice of engine size and its effect on
balancing fuel economy and performance. Some manufacturers choose not to downsize the
engine when applying hybrid technologies. In these cases, performance is vastly improved,
while fuel efficiency improves significantly less than if the engine were downsized to maintain
the same performance as the conventional version. The non-downsizing approach is used for
vehicles where towing and/or hauling are an integral part of their performance requirements. In
these cases, if the engine is downsized, the battery can be quickly drained during a long hill
climb with a heavy load, leaving only a downsized engine to carry the entire load. Because
towing capability is currently a heavily-marketed HD pickup truck attribute, manufacturers are

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
hesitant to offer a truck with a downsized engine that can lead to a significantly diminished
towing performance when the battery state of charge level is low, and therefore engines are
traditionally not downsized for these vehicles. In assessing the cost of hybrid technology for
heavy duty vehicles, the agencies have assumed that engines will not be downsized.
Strong hybrid technology utilizes an axial electric motor connected to the transmission
input shaft and connected to the engine crankshaft through a clutch. The axial motor is a
motor/generator that can provide sufficient torque for launch assist, all electric operation, and the
ability to recover significant levels of braking energy.
A hybrid drive unit is complex and consists of discrete components such as the electric
traction motor, transmission, generator, inverter, controller and cooling devices. Certain types of
drive units may work better than others for specific vehicle applications or performance
requirements. Several types of motors and generators have been developed for hybrid-electric
drive systems, many of which merit further evaluation and development on specific applications.
Series HEVs typically have larger motors with higher power ratings because the motor alone
propels the vehicle, which may be applicable to Class 3-5 applications. In parallel hybrids, the
power plant and the motor combine to propel the vehicle. Motor and engine torque are usually
blended through couplings, planetary gear sets and clutch/brake units. The same mechanical
components that make parallel heavy-duty hybrid drive units possible can be designed into series
hybrid drive units to decrease the size of the electric motor(s) and power electronics.
An electrical energy storage system is needed to capture energy from the generator, to
store energy captured during vehicle braking events, and to return energy when the driver
demands power. This technology has seen a tremendous amount of improvement over the last
decade and recent years. Advanced battery technologies and other types of energy storage are
emerging to give the vehicle its needed performance and efficiency gains while still providing a
product with long life. The focus on the more promising energy storage technologies such as
nickel metal-hydride (NiMH) and lithium technology batteries along with ultra-capacitors for the
heavy-duty fleet should yield interesting results after further research and applications in the
light-duty fleet.
Heavy-duty hybrid vehicles also use regenerative braking for improved fuel economy,
emissions, brake heat, and wear. A conventional heavy vehicle relies on friction brakes at the
wheels, sometimes combined with an optional engine retarder or driveline retarder to reduce
vehicle speed. During normal braking, the vehicle's kinetic energy is wasted when it is
converted to heat by the friction brakes. The conventional brake configuration has large
components, heavy brake heat sinks, and high temperatures at the wheels during braking, audible
brake squeal, and consumable components requiring maintenance and replacement. Hybrid
electric systems recover some of the vehicle's kinetic energy through regenerative braking,
where kinetic energy is captured and directed to the energy storage system. The remaining
kinetic energy is dissipated through conventional wheel brakes or in a driveline or transmission
retarder. Regenerative braking in a hybrid electric vehicle can require integration with the
vehicle's foundation (friction) braking system to maximize performance and safety.
Today's systems function by simultaneously using the regenerative features and the
friction braking system, allowing only some of the kinetic energy to be saved for later use.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Optimizing the integration of the regenerative braking system with the foundation brakes would
increase the benefits and is a focus for continued work. This type of hybrid regenerative braking
system improves fuel economy, GHG emissions, brake heat, and wear.
In a hydraulic hybrid system, deceleration energy is taken from the drivetrain by an inline
hydraulic pump/motor unit by pumping hydraulic fluid into high pressure cylinders. The fluid,
while not compressible, pushes against a membrane in the cylinder that compresses an inert gas
to 5,000 PSI or more when fully charged. Upon acceleration, the energy stored in the
pressurized tank pushes hydraulic fluid back into the drivetrain pump/motor unit, allowing it to
motor into the drivetrain and assist the vehicle's engine with the acceleration event. This heavy-
duty vehicle hybrid approach has been demonstrated successfully, producing good results on a
number of commercial and military trucks.
Despite the significant future potential for hybrids discussed above, there are no simple
solutions applicable for each heavy-duty hybrid application due to the large vocational vehicle
fleet variation. A choice must be made relative to the requirements and priorities for the
application. Challenges in motor subsystems such as gear reductions and cooling systems must
be considered when comparing the specific power, power density, and cost of the motor
assemblies. High speed motors can significantly reduce weight and size, but they require speed
reduction gear sets that can offset some of the weight savings, reduce reliability and add cost and
complexity. Air-cooled motors are simpler and generally less expensive than liquid cooled
motors, but they are larger and heavier, and they require access to ambient air, which can carry
dirt, water, and other contaminants. Liquid-cooled motors are generally smaller and lighter for a
given power rating, but they may require more complex cooling systems that can be avoided
with air-cooled versions. Various coolant options, including water, water-glycol, and oil, are
available for liquid-cooled motors but must be further researched for long term durability.
Electric motors, power electronics, electrical safety, regenerative braking, and power-plant
control optimization have been identified as the most critical technologies requiring further
research to enable the development of higher efficiency hybrid electric propulsion systems.
2.4.5 Axles
2.4.5.1 Axle Efficiency
Axle efficiency is improved by reducing generally two categories of losses; mechanical
losses and spin losses.
Mechanical losses can be reduced by reducing the friction between the two gears in
contact. This friction is reduced mainly by improving the surface finish of the gears. The other
way of doing this by reducing the amount of distance the gear faces are sliding against each
other. Generally speaking frictional losses are proportional to the torque on the axle not a
function of rotational speed of the axle.
Spin losses on the other hand are a function of speed and not torque. One of the main
ways to reduce the spin losses of the axle is by using a lower viscosity lubricant. Some high-
performance lower viscosity formulations have been designed to have superior performance at
high operating temperatures, and may have extended change intervals.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
A study conducted by researchers at Shell Global Solutions on a Mercedes Benz OM
460LA heavy-duty diesel engine run under the World Harmonized Transient Cycle (WHTC) and
World Harmonized Stationary Cycle (WHSC), used a combination of a SAE 5W-30 engine oil,
SAE 75W-80 gearbox oil and SAE 75W-90 axle oil. The combination yielded average fuel
economy improvements of 1.8 percent over the WHTC and 1.1 percent over the WHSC, relative
to a SAE 15W-40 engine oil, SAE 80W gearbox and SAE 90 axle oil [VT-27], The baseline
lubricants represent current mainstream products, and the new lubricants were top-tier
formulations focusing on modified viscometric effects. Using the WHSC cycle, significant
variations in the individual lubricant contribution under different speed and load conditions
within the cycle were identified. Additionally, an average fuel economy improvement of 1.8
percent was observed using medium-duty trucks under a range of typical European driving
conditions in a controlled field trial.67
Spin losses can also be reduced by lowering the volume of lubricant in the sump. This
reduces the surface area of the gears that are churning through the lubricant. One of the main
challenges of doing this is making sure that there is still adequate coverage of lubricant on the
gears and bearings as well as adequate circulation so that the lubricant temperature does not rise
too high and accelerate the aging of the lubricant.
If a manufacturer wishes to demonstrate a benefit specific to any technology that
improves axle efficiency, an axle efficiency test can be performed and input into GEM. See RIA
Chapter 3 for a description of the test procedure for axle efficiency.
2.4.5.2	Gear Ratio
Combining with transmission ratio, selection of the axle ratio can play a significant role
in vehicle performance. For an on-highway tractor, the axle ratio must be selected in such a way
that the engine can constantly run inside the sweet spot, where the engine efficiency is optimal
for a typical constant cruise speed like 65 miles per hour. Although many vehicles on the road
already use a fast axle ratio as low as 2.64:1 with the direct drive of transmission, which moves
the engine speed in the range of 1200 rpm or even lower, most vehicles still use higher or slower
axle ratio, which puts the engine speed in the range of 1300-1400 rpm. In order to take
advantage of optimal engine speed, which is typically in the range of 1100-1150 rpm for HHD
diesel vehicles, it is expected that a faster axle ratio lower than 2.64:1 would be widely used in
2018 and beyond for tractors. Furthermore, in order to enhance vehicle performance, many axle
manufacturers are developing dual speed axles, allowing vehicles to switch to a higher axle ratio
during transient driving conditions, such as city traffic. On the vocational side, the ability to start
a heavy vehicle, climb hills, and operate smoothly at low speed is strongly influenced by axle
ratio, and therefore, one can see a large variation of axle ratios depending on the application.
2.4.5.3	Tandem Drive Axle Improvements
Manufacturers are developing technologies to enable heavy trucks with two rear drive
axles to be driven solely by the lead rear axle either permanently or on a part time basis.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.4.5.3.1	6x2
Most tractors and heavy heavy-duty vocational vehicles today have three axles - a steer
axle and two rear drive axles, which is commonly referred to as a 6x4 configuration.
Manufacturers offer 6x2 tractors that include one rear drive axle and one rear non-driving axle.
The 6x2 tractors offer three distinct benefits. First, the non-driving rear axle does not have
internal friction and therefore reduces the overall parasitic losses in the drivetrain. In addition,
the 6x2 configuration typically weighs approximately 300 to 400 pounds less than a 6x4
configuration.68 Finally, the 6x2 typically costs less or is cost neutral when compared to a 6x4
tractor. Sources cite the effectiveness of 6x2 axles at between 1 and 3 percent.69 Similarly, with
the increased use of double and triple trailers, which reduce the weight on the tractor axles when
compared to a single trailer, manufacturers offer 4x2 axle configurations. The 4x2 axle
configuration would have as good as or better fuel efficiency performance than a 6x2.
2.4.5.3.2	Enhanced 6x2
One of the drawbacks of 6x2 axle is lack of traction, specifically during the winter
condition and high grade road when the road is slippery. In order to overcome this deficiency,
some axle manufacturers offer products that perform similar to the 6x4 configurations.
SMARTandem offered by Meritor is just one of the examples.70 In this system, the axle runs
6x2 for most time. Once the conditions that require more traction are experienced, the vehicle
activates the system to add more loads into one the powered axle, thus instantly increasing
traction. This system offers weight savings in the range of 300 to 400 lbs, as well as 2 percent
fuel saving when compared to a conventional 6x4 axle.
2.4.5.3.3	Part Time 6x2 Axle
Based on confidential stakeholder discussions, the agencies anticipate that the axle
market may offer, in the time frame of Phase 2, a Class 8 version of the type of axle disconnect
that today allows 4x4 operators of HD pickup trucks to automatically disconnect or reconnect the
front axle depending on needs for traction in varying driving conditions. The Class 8 version
would likely function for the two tandem drive axles in a similar manner as the HD pickup trucks
do for the front axle. The switching could be automated or user-commanded. In these cases, the
axle actuator housing, sometimes called the axle disconnect housing, is part of the differential
that houses the gears and shift fork required to lock two axles together. The axle actuator works
together with the transfer case to send torque to all four wheel-ends. Recently, Dana Holding
Corporation has developed an axle system that switches between the two modes based on driving
conditions to maximize driveline efficiency.71 When high traction is required, the system
operates in 6x4 mode. When 6x4 tractive effort is not required, the system operates in 6x2
mode. It is reported that this type of system can offer a benefit of 2.5 percent.
In the 4x4 example, the transfer case connects the input from the transmission to the rear
and front driveshafts. The axle actuator housing is found on the differential. In the 4x4 example,
a shift fork inside the axle actuator housing slides a locking collar over two gears locking both
driver and passenger side axles together. In some 4x4 vehicles, those with automatic 4WD, this
process occurs automatically. In others, with selective 4WD, the driver can choose to engage
4WD or RWD with a switch. These have slightly different axle actuator housings and have

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
actuator solenoids mounted to them.72 These systems would not provide the weight reduction
benefit of the permanent 6x2 configuration, and may offer less fuel savings, especially with
operator-switchable systems.
2.4.6 Weight Reduction
Mass reduction is a technology that can be used in a manufacturer's strategy to meet the
Phase 2 standards (although the agencies are not predicating the standards on use of downsizing).
Vehicle mass reduction (also referred to as "light-weighting"), decreases fuel consumption and
GHG emissions by reducing the energy demand needed to overcome inertia forces, and rolling
resistance. Reduced mass in heavy-duty vehicles can benefit fuel efficiency and CO2 emissions
in two ways. If a truck is running at its gross vehicle weight limit with high density freight, more
freight can be carried on each trip, increasing the truck's ton-miles per gallon. If the vehicle is
carrying lower density freight and is below the GVWR (or GCW) limit, the total vehicle mass is
decreased, reducing rolling resistance and the power required to accelerate or climb grades.
Many vehicle components are typically made of heavier material, such as traditional steel.
Manufacturers have worked with mass reduction technologies for many years and a lot of these
technologies have been used in production vehicles. The weight savings achieved by adopting
mass reduction technologies offset weight gains due to increased vehicle size, larger powertrains,
and increased feature content (sound insulation, entertainment systems, improved climate
control, etc.). Generally, an empty truck contributes to about one-third of the total vehicle
weight. Every 10 percent drop in vehicle weight reduces fuel use about 5 percent.73
Although many gains have been made to reduce vehicle mass, many of the features being
added to modern tractors to benefit fuel efficiency, such as additional aerodynamic features or
idle reduction systems, have the effect of increasing vehicle weight, causing mass to stay
relatively constant. Material and manufacturing technologies can also play a significant role in
vehicle safety by reducing vehicle weight, and in the improved performance of vehicle passive
and active safety systems. Hybrid powertrains, fuel cells and auxiliary power would not only
present complex packaging and weight issues, they would further increase the need for
reductions in the weight of the body, chassis, and powertrain components in order to maintain
vehicle functionality.
Manufacturers may employ a systematic approach to mass reduction, where the net mass
reduction is the addition of a direct component or system mass reduction, also referred to as
primary mass reduction, plus the additional mass reduction taken from indirect ancillary systems
and components, also referred to as secondary mass reduction or mass compounding.
Mass reduction can be achieved through a number of approaches, even while maintaining
other vehicle functionalities. As summarized by NAS in its 2011 light duty vehicle report, there
are two key strategies for primary mass reduction: 1) substituting lighter materials for heavier
materials; and 2) changing the design to use less material.74

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.4.6.1 Material Substitution
Substitution of a material used in an assembly or a component for one with lower density
and/or higher strength includes replacing a common material such as mild steel with higher-
strength and advanced steels, aluminum, magnesium, and composite materials. In practice,
material substitution tends to be quite specific to the manufacturer and situation. Some materials
work better than others for particular vehicle components, and unless strength is matched, some
substituted components may need to be more numerous (i.e. two brackets instead of one).
Further, one choices of material may lead a manufacturer to invest more heavily in adjusting its
manufacturing process to its properties, thus possibly impeding its ability to consider other
materials. The agencies recognize that like any type of mass reduction, material substitution has
to be conducted not only with consideration to maintaining equivalent component strength, but
also to maintaining all the other attributes of that component, system or vehicle, such as
crashworthiness, durability, and noise, vibration and harshness (NVH).
One example that combines material substitution with component-elimination is the use
of wide-based single tires and aluminum rims to replace traditional dual tires and rims,
eliminating eight steel rims and eight tires from a tractor. Using aluminum, metal alloys, metal
matrix composites, and other lightweight components where appropriate can reduce empty
vehicle weight (known as "tare weight"), improve fuel efficiency, and reduce greenhouse gas
emissions. In addition, in weight-sensitive applications, lightweight components can allow more
cargo and increased productivity. A report by the National Commission on Energy Policy
estimates that a fuel economy gain of 5.0 percent on certain applications could be achieved by
vehicle mass reduction further illustrating the fuel economy gains possible on heavy-duty
applications.75 A report for the U.S. DOT estimated potential reductions in modal GHG
emissions are 4.6 percent, though it also found that current light-weight materials are costly and
are application- and vehicle-specific with need for further research and development for
advanced materials.76
The principal barriers to overcome in reducing the weight of heavy vehicles are
associated with the cost of lightweight materials, the difficulties in forming and manufacturing
lightweight materials and structures, the cost of tooling for use in the manufacture of relatively
low-volume vehicles (when compared to automotive production volumes), and ultimately, the
extreme durability requirements of heavy vehicles. While light-duty vehicles may have a life
span requirement of several hundred thousand miles, typical heavy-duty commercial vehicles
must last over 1 million miles with minimum maintenance, and often are used in secondary
applications for many more years. This requires high strength, lightweight materials that provide
resistance to fatigue, corrosion, and can be economically repaired. Additionally, because of the
limited production volumes and the high levels of customization in the heavy-duty market,
tooling and manufacturing technologies that are used by the light-duty automotive industry are
often uneconomical for heavy vehicle manufacturers. Lightweight materials such as aluminum,
titanium and carbon fiber composites provide the opportunity for significant weight reductions,
but their material cost and challenging forming and manufacturing requirements make it difficult
for them to compete with low-cost steels. In addition, although mass reduction is currently
occurring on both vocational vehicles and line haul tractor-trailers, the addition of other systems
for fuel economy, performance or comfort increases the vehicle mass offsetting the mass
reduction that has already occurred, thus it is not captured in the overall vehicle mass

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
measurement (e.g. 500 lbs for WHR). Most vehicle manufacturers offer lightweight tractor
models that are 1,000 or more pounds lighter than comparable models. Lighter-weight models
combine different weight-saving options that may include:77
•	Cast aluminum alloy wheels can save up to 40 pounds each for total savings of 400
pounds
•	Aluminum axle hubs can save over 120 pounds compared to ductile iron or steel
•	Centrifuge brake drums can save nearly 100 pounds compared to standard brake drums
•	Aluminum clutch housing can save 50 pounds compared to iron clutch housing
•	Composite front axle leaf springs can save 70 pounds compared to steel springs
•	Aluminum cab frames can save hundreds of pounds compared to standard steel frames
2.4.6.2 Synergistic Effects - Reduced Power Demand
Manufacturers employ a systematic approach to mass reduction, where the net mass
reduction is the addition of a direct component or system mass reduction plus the additional mass
reduction that can be taken from indirect ancillary systems and components, as a result of full
vehicle optimization, effectively compounding or obtaining a secondary mass reduction from a
primary mass reduction. The strategy of using less material compared to the baseline component
or system can be achieved by optimizing the design and structure of vehicle components,
systems and vehicle structure. Vehicle manufacturers have long used these continually-
improving CAE tools to optimize vehicle designs. For example, the Future Steel Vehicle (FSV)
project sponsored by World Auto Steel used three levels of optimization: topology optimization,
low fidelity 3G (Geometry, Grade, and Gauge) optimization, and subsystem optimization, to
achieve 30 percent mass reduction in the body structure of a vehicle with a mild steel unibody
structure.78 Using less material can also be achieved through improving the manufacturing
process, such as by using improved joining technologies and parts consolidation. This method is
often used in combination with applying new materials.
If vehicle mass is reduced sufficiently through application of the two primary strategies
of using less material and material substitution described above, secondary mass reduction
options may become available. Secondary mass reduction is enabled when the load requirements
of a component are reduced as a result of primary mass reduction. If the primary mass reduction
reaches a sufficient level, a manufacturer may use a smaller, lighter, and potentially more
efficient powertrain while maintaining vehicle performance. If a powertrain is downsized, a
portion of the mass reduction may be attributed to the reduced torque requirement that results
from the lower vehicle mass. The lower torque requirement enables a reduction in engine
displacement, changes to transmission torque converter and gear ratios, and changes to the final
drive gear ratio. The reduced powertrain torque may enable the downsizing and/or mass
reduction of powertrain components and accompanying reduced rotating mass (e.g., for
transmission, driveshafts/halfshafts, wheels, and tires) without sacrificing powertrain durability.
However, there may be trade-offs, as it is possible that use of a downsized engine may require a
transmission with more gears. The combined mass reductions of the engine, drivetrain, and body
would reduce stresses on the suspension components, steering components, wheels, tires, and
brakes, which can allow further reductions in the mass of these subsystems. Reducing the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
unsprung masses such as the brakes, control arms, wheels, and tires further reduce stresses in the
suspension mounting points, which would allow for further optimization and potential mass
reduction.
One example of a synergistic effect is rotational inertia. Reducing the weight of rotating
components provides an enhanced fuel efficiency benefit over reducing the weight of static
components. In theory, as components such as brake rotors, brake drums, wheels, tires,
crankshafts, camshafts, and piston assemblies become lighter, the power consumption to rotate
the masses would be directly proportional to the mass decrease. Using physical properties of a
rotating component such as a wheel, it is relatively straightforward to calculate an equivalent
mass. However, we do not have enough information to derive industry average values for
equivalent mass, nor have we evaluated the best way for GEM to account for this. Using typical
values for a heavy-duty steel wheel compared to a similar-sized aluminum wheel, the agencies
estimate the equivalent mass ratio is in the range of 1.2 to 1.3. That means that by reducing the
mass of a wheel by 20 pounds, the vehicle could theoretically perform as if 26 pounds had been
reduced.
Estimates of the synergistic effects of mass reduction and the compounding effect that
occurs along with it can vary significantly from one report to another. For example, in
discussing its estimate, an Auto-Steel Partnership report states that "These secondary mass
changes can be considerable—estimated at an additional 0.7 to 1.8 times the initial mass
change."79 This means for each one pound reduction in a primary component, up to 1.8 pounds
can be reduced from other structures in the vehicle (i.e., a 180 percent factor). The report also
discusses that a primary variable in the realized secondary weight reduction is whether or not the
powertrain components can be included in the mass reduction effort, with the lower end
estimates being applicable when powertrain elements are unavailable for mass reduction.
However, another report by the Aluminum Association, which primarily focuses on the use of
aluminum as an alternative material for steel, estimated a factor of 64 percent for secondary mass
reduction even though some powertrain elements were considered in the analysis.80 That report
also notes that typical values for this factor vary from 50 to 100 percent. Although there is a
wide variation in stated estimates, synergistic mass reductions do exist, and the effects result in
tangible mass reductions. Mass reductions in a single vehicle component, for example a door
side impact/intrusion system, may actually result in a significantly higher weight savings in the
total vehicle, depending on how well the manufacturer integrates the modification into the
overall vehicle design. Accordingly, care must be taken when reviewing reports on weight
reduction methods and practices to ascertain if compounding effects have been considered or not.
2.4.7 Vehicle Speed Limiter
The power required to move a vehicle increases as the vehicle speed increases.
Travelling at lower speeds provides additional efficiency to the vehicle performance. Most
vehicles today have the ability to electronically control the maximum vehicle speed through the
engine controller. This feature is used today by fleets and owners to provide increased safety
and fuel economy. Currently, these features are designed to be able to be changed by the owner
and/or dealer.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The impact of this feature is dependent on the difference between the governed speed and
the speed that would have been travelled, which is dependent on road type, state speed limits,
traffic congestion, and other factors. The agencies assess the benefit of a vehicle speed limiter
by reducing the maximum drive cycle speed on the 65 mph Cruise mode of the cycle. The
maximum speed of the drive cycle is 65 mph, therefore any vehicle speed limit with a setting
greater than this would show no benefit for purposes of these regulations, but may still show
benefit in the real world in states where the interstate truck speed limit is greater than the
national average of 65.5 mph.
The benefits of this simple technology are widely recognized. The American Trucking
Association (ATA) developed six recommendations to reduce carbon emissions from trucks in
the United States. Their first recommendation is to enact a national truck speed limit of 65 mph
and require that trucks manufactured after 1992 have speed governors set at not greater than 65
mph.81 The SmartWay program includes speed management as one of their key Clean Freight
Strategies and provides information to the public regarding the benefit of lower highway
speeds.82
Some countries have enacted regulations to reduce truck speeds. For example, the United
Kingdom introduced regulations in 2005 which require new trucks used for goods movement to
have a vehicle speed limiter not to exceed 90 km/hr (56 mph).83 The Canadian Provinces of
Ontario and Quebec developed regulations which took effect in January 2009 that requires on-
highway commercial heavy-duty trucks to have speed limiters which limit the truck's speed to
105 km/hr (65 mph).84
Many truck fleets consider speed limiter application a good business practice in their
operations. A Canadian assessment of heavy-duty truck speed limiters estimated that 60 percent
of heavy truck fleets in North America use speed limiters.84 Con Way Freight, Con Way
Truckload, and Wal-Mart currently govern the speeds of their fleets between 62 and 65 mph.85
A potential disbenefit of this technology is the additional time required for goods
movement, or loss of productivity. The elasticity between speed reduction and productivity loss
has not been well defined in industry. The Canadian assessment of speed limiters cited above
found that the fuel savings due to the lower operating speeds outweigh any productivity losses.
A general consensus among the OEMs is that a 1 percent decrease in speed might lower
productivity by approximately 0.2 percent.85
In Phase 1, the agencies did not premise the tractor standards on a technology package
that included VSL. Vehicle speed limiters are a technology recognized in Phase 1 GEM, but
manufacturers are not opting to use the tamper-resistant VSLs as a strategy for complying with
the early years of Phase 1 CO2 emissions and fuel consumption standards.
The impact of VSL set to 55 mph of a typical high roof tractor-trailer is approximately 7
percent for day cabs and 10 percent for sleeper cabs, as shown below in Figure 2-14.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
12.0%
-5 10.0%
CD
c
O
U
~cD
=3
u_
"O
c
rvi
O
U
NP
0s-
3.0%
6.0%
4.0%
2.0%
0.0%
54
Vehicle Speed Limiter Effectiveness
Composite Result
-Sleeper Cab
¦ Day Cab
56	58	60	62
Vehicle Speed Limiter Setting (mph)
64
66
Figure 2-14 Vehicle Speed Limiter Effectiveness in Tractors
2.4.8 Reduced Idling Time
2.4.8.1 Engine Shutdown with Alternate Power Source during Hoteling
Class 8 heavy-duty diesel truck extended engine idling expends significant amounts of
fuel in the United States. Department of Transportation regulations require a certain amount of
rest for a corresponding period of driving hours, as discussed in Chapter 1. Extended idle occurs
when Class 8 long haul drivers rest in the sleeper cab compartment during rest periods as drivers
find it more convenient and economical to rest in the truck cab itself. In many cases it is the only
option available. During this rest period a driver will generally idle the truck in order to provide
heating or cooling or run on-board appliances. During rest periods the truck's main propulsion
engine is running but not engaged in gear and it remains in a stationary position. In some cases
the engine can idle in excess of 10 hours. During this period of time, fuel consumption will
generally average 0.8 gallons per hour.86 Average overnight fuel usage would exceed 8 gallons
in this example. When multiplied by the number of long haul trucks without idle control
technology that operate on national highways on a daily basis, the number of gallons consumed
by extended idling would exceed 3 million gallons per day. Fortunately, a number of
alternatives (idling reduction technologies) are available to alleviate this situation.
2.4.8.1.1 Idle Control Technologies
Idle reduction technologies in general utilize an alternative energy source in place of
operating the main engine. By using these devices the truck driver can obtain needed power for

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
services and appliances without running the engine. A number of these devices attach to the
truck providing heat, air conditioning, or electrical power for microwave ovens, televisions, etc.
The idle control technologies (along with their typical hourly fuel rate) available today
include the following:87
•	Auxiliary Power Unit (APU) powers the truck's heating, cooling, and electrical system.
The fuel use of an APU is typically 0.2 gallons per hour
•	Fuel Operated Heater (FOH) provides heating services to the truck through small diesel
fired heaters. The fuel use is typically 0.04 gallons per hour
•	Battery Air Conditioning Systems (B AC) provides cooling to the truck
•	Automatic Stop/Start Systems powers the truck systems through the battery and starts the
engine to recharge the battery after it reaches a threshold level.
•	Thermal Storage Systems provide cooling to trucks
Another alternative involves electrified parking spaces, with or without modification to
the truck. An electrified parking space system operates independently of the truck's engine and
allows the truck engine to be turned off while it supplies heating, cooling, and electrical power.
These systems provide off-board electrical power to operate either:
1.	A single system electrification which requires no on-board equipment by providing an
independent heating, cooling, and electrical power system, or
2.	A dual system which allows driver to plug in on-board equipment
In the first case, power is provided to stationary equipment that is temporarily attached to
the truck. In the second, the truck is modified to accept power from the electrical grid to operate
on-board truck equipment. The retail price of idle reduction systems varies depending on the
level of sophistication. For example, on-board technologies such as APUs can retail for over
$8,000 while options such as electrified parking spaces require negligible up-front costs for
equipment for the tractor itself, but will accrue fees with usage.88
CO2 emissions and fuel consumption during extended idling are significant contributors
to emissions and fuel consumption from Class 8 sleeper cabs. The federal test procedure does
evaluate idle emissions and fuel consumption as part of the drive cycle and related emissions
measurement. However, long duration extended idle emissions and fuel consumption are not
fully represented during the prescribed test cycle. To address the fact that real-world fuel and
emissions savings can occur with idle reduction technologies that cannot be reflected on the test
cycle, the agencies adopted a GEM input for manufacturers who provide for idle control using an
automatic engine shutdown system (AESS) on the tractor.
The GEM input, calculated as shown in Table 2-5, recognizes the CO2 reductions and
fuel consumption savings attributed to idle control systems and allows vehicle manufacturers
flexibility in product design and performance capabilities. The agencies first determined the fuel
consumption of each idle reduction technology, as noted previously. Due to the range of fuel
consumption of APUs and the precision of the available test information, the agencies are

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
utilizing, as proposed, an APU fuel consumption of 0.2 gal/hr. Then the agencies determined a
split between idling hours using the main engine versus the idle reduction technology. For
example, the baseline idle emission rate was assumed to be determined by 100 percent main
engine idling. For APU and battery APU technologies with a tamper-proof AESS, the agencies
assumed that these technologies would be operating 100 percent of the idling time. For
automatic start/stop systems with a tamper-proof AESS, the agencies determined that the idling
power would come from the battery half of the idling time and the other half would require main
engine idling. For fuel operated heaters with a tamper-proof AESS, the agencies assumed that
800 of the idling hours would involve the use of the fuel operated heater and that the main engine
would idle for the other 1000 hours per year to supply cooling and other needs. For idle
reduction technologies with an adjustable AESS, the agencies discounted the number of hours
operated by the idle reduction technology by 20 percent to account for the fact that it is an
adjustable (non tamper-proof) system. For adjustable AESS without an additional idle reduction
technology, the agencies set the number of main engine operating hours at 25 percent of the total
idle time to also reflect that it is adjustable and that the agencies have less certainty in the
continued use of this in the real world.
MEMA commented that the agencies should assume 2,500 hours of idling per year. The
agencies reviewed this and other studies to quantify idling operation. The 2010 NAS study
assumes between 1,500 and 2,400 idling hours per year.89 Gaines uses 1,800 hours per year.90
Brodrick, et al. assumes 1,818 hours per year (6 hours per day for 303 days per year) based on an
Argonne study and Freightliner fleet customers.91 An EPA technical paper states between 1,500
and 2,400 hours per year.92 Kahn uses 1,830 hours as the baseline extended idle case.93 Based
on the literature, the agencies are finalizing as proposed the use of 1,800 hours per year as
reasonably reflecting the available range of information.
The agencies assumed the average Class 8 sleeper cab travels 125,000 miles per year
(500 miles per day and 250 days per year) and carries 19 tons of payload (the standardized
payload finalized for Class 8 tractors) to calculate the baseline running emissions. For each
technology combination, the sum of the running and idling emissions was calculated and the
percent reduction in CO2 emissions from the main engine idling scenario was calculated. These
percent reduction values are included in 40 CFR 1037.520.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-4 Idle CO2 Emissions per Year for Idle Reduction Technologies

Idle Fuel
Consumption
(gal/hour)
Idle C02
emissions per
hour
IRT Idle
Hours per
Year
Main Engine
Idle Hours per
Year
Idle CO2 Emission per
year (grams)
Baseline
0.8
8144

1800
14,659,200
Tamper-Proof AESS
0.3
3054
1800
0
5,497,200
Tamper-Proof AESS w/
Diesel APU
0.3
3054
1800
0
5,497,200
Tamper-Proof AESS w/
Battery APU
0.02
203.6
1800
0
366,480
Tamper-Proof AESS w/
Automatic Stop-Start
0
0
900
900
7,329,600
Tamper-Proof AESS w/
FOH Cold, Main Engine
Warm
0.04
407.2
800
1000
8,469,760
Adjustable AESS w/
Diesel APU
0.3
3054
1440
360
7,329,600
Adjustable AESS w/
Battery APU
0.02
203.6
1440
360
3,225,024
Adjustable AESS w/
Automatic Stop-Start
0
0
720
1080
8,795,520
Adjustable AESS w/ FOH
Cold, Main Engine Warm
0.04
407.2
640
1160
9,707,648
Adjustable AESS
programmed to 5 minutes
0.3
3054
450
1350
12,368,700

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-5 GEM Input for Idle Reduction Technologies

TYPICAL
G/TON-
MILE
MILES
PER
YEAR
PAYLOAD
(TONS)
GHG
EMISSIONS
DUE TO
RUNNING (g)
GHG
EMISSIONS
DUE TO
RUNNING PLUS
IDLE (g)
% RED.
FROM
BASELINE
Baseline
88
125000
19
209,000,000
223,659,200
0%
Tamper-Proof AESS
88
125000
19
209,000,000
214,497,200
4.1%
Tamper-Proof AESS w/
Diesel APU
88
125000
19
209,000,000
214,497,200
4.1%
Tamper-Proof AESS w/
Battery APU
88
125000
19
209,000,000
209,366,480
6.4%
Tamper-Proof AESS w/
Automatic Stop-Start
88
125000
19
209,000,000
216,329,600
3.3%
Tamper-Proof AESS w/
FOH Cold, Main Engine
Warm
88
125000
19
209,000,000
217,469,760
2.8%
Adjustable AESS w/ Diesel
APU
88
125000
19
209,000,000
216,329,600
3.3%
Adjustable AESS w/
Battery APU
88
125000
19
209,000,000
212,225,024
5.1%
Adjustable AESS w/
Automatic Stop-Start
88
125000
19
209,000,000
217,795,520
2.6%
Adjustable AESS w/ FOH
Cold, Main Engine Warm
88
125000
19
209,000,000
218,707,648
2.2%
Adjustable AESS
programmed to 5 minutes
88
125000
19
209,000,000
221,368,700
1.0%
2.4.8.2 Stop Start
For heavy-duty vehicles to apply engine stop-start technology without a reduction in
vehicle function, some additional vehicle technologies are needed. To some extent this could be
considered similar to a mild hybrid system, but it is not the same as the mild hybrid system
described for HD pickups and vans described below in Chapter 2.5. The agencies are projecting
the presence of a battery sufficient to offer electrified power steering, and some other electrified
accessories. Some systems may replace the conventional alternator with a belt or crank driven
starter/alternator and may add high voltage electrical accessories (which may include electric
power steering and an auxiliary automatic transmission pump). The limited electrical
requirements of these systems allow the use of lead-acid batteries or supercapacitors for energy
storage, or the use of a small lithium-ion battery pack.
The NACFE Idle Reduction Confidence report was written with long haul tractors in
mind; however the section on vehicle electrification discusses inverters and on-vehicle solar

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
energy capture, and offers some insights relevant to vocational vehicle electrification as it
pertains to stop-start systems.94 Inverters and beltless alternators can use DC power stored in
batteries to power on-board electrical devices and re-start engines. One example of a company
that supplies battery-inverter idle reduction systems for vocational vehicles is Vanner.95 There
are also systems available today that are designed to capture solar energy and store this energy
for distribution to electrified accessories and engine re-starting. One example of a company that
supplies on-vehicle solar energy capture for vocational vehicles is eNow.96
2.4.8.3 Neutral Idle
Automatic transmissions historically apply torque to an engine when in gear at zero speed
because of torque converter, such as when stopped at a traffic light. A neutral idle technology
can disengage transmission with torque converter, thus reducing power loss to a minimum.
2.4.9 Air Conditioning
2.4.9.1 Refrigerant Leakage
Hydrofluorocarbon (HFC) refrigerants, which are powerful GHG pollutants, can be
emitted to the atmosphere through component and system leaks during operation, during
maintenance and servicing, and with disposal at the end of the vehicle's life. The current widely-
used refrigerant - R134a, has a much higher global warming potential (GWP) than CO2,
therefore a small leakage of this refrigerant has a much greater global warming impact than a
similar amount of emissions of CO2 or other mobile source GHGs.
Direct emissions of HFC from air conditioning systems can be reduced by minimizing
system leaks. Based on measurements from 300 European light-duty 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.97 This corresponds to a leakage rate of 6.9 percent 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.98 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 percent empty and the fraction recovered at end-of-life was 8.5 percent.
Manufacturers today are complying with the HD Phase 1 program requirements to reduce
A/C leakage emissions by utilizing high-quality, low-leakage air conditioning system
components in the production of new tractors, and HD pickup trucks and vans. Some of the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
components available to manufacturers are low-permeation flexible hoses, multiple o-ring or seal
washer connections, and multiple-lip compressor shaft seals. The availability of low leakage
components in the market is being driven by the air conditioning credit program in the light-duty
GHG rulemaking. The cooperative industry and government Improved Mobile Air Conditioning
(IMAC) program has demonstrated that new-vehicle leakage emissions can be reduced by 50
percent by reducing the number and improving the quality of the components, fittings, seals, and
hoses of the AJC system."
2.4.9.2	System Efficiency
CO2 emissions and fuel consumption are also associated with air conditioner efficiency,
since air conditioners create load on the engine. See 74 FR at 49529. The agencies are adopting
Phase 2 provisions for tractors and vocational vehicles recognizing the opportunity for more
efficient air conditioning systems.
2.4.9.3	Solar Control
Solar control glazing consists of both solar absorbing and solar reflective glazing that can
reduce the temperature inside a vehicle, and therefore reduce the air conditioning requirements.
The reduction in air conditioning load can lead to reductions in fuel consumption and GHG
emissions. CARB's Low Emission Vehicle III Regulations (LEVIII) include a GHG credit for
this technology.100 The Enhanced Protective Glass Automotive Association indicated that new
heavy-duty trucks today typically use solar absorbing glass.
Solar reflective paints reflect approximately a half of the solar energy by reflecting the
infrared portion of the solar spectrum. A study conducted by National Renewable Energy
Laboratory found benefits to sleeper cab tractors using reflective paint and other thermal control
technologies.101
There are many factors that influence the level of emissions and fuel consumption
reductions due to solar control glazing and solar reflective paint. The fraction of time spent
idling during the daytime hours, the fraction of hours of the day that are sunny, the ambient
temperatures, the wind conditions and/or vehicle speed, the fraction of the vehicles that are
painted colors other than white, and other factors influence the potential impact of these
technologies. Because of the difficulty in assessing the potential emission reductions from solar
control paint and glazing, the agencies did not propose this technology as part of HD Phase 2.
The agencies received some clarifications from ARB on our evaluation of solar technologies and
some CBI from Daimler, but not a sufficient amount of information to evaluate the baseline level
of solar control that exists in the heavy-duty market today, determine the effectiveness of each of
the solar technologies, or to develop a definition of what qualifies as a solar control technology
that could be used in the regulations. Therefore, the agencies would consider solar control to be
a technology that manufacturers may consider pursuing through the off-cycle credit program.
2.4.10 Other Accessory Improvements
Electric power steering (EPS) provides a potential reduction in CO2 emissions and fuel
consumption over hydraulic power steering because of reduced overall accessory loads. This

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
eliminates the parasitic losses associated with belt-driven power steering pumps which
consistently draw load from the engine to pump hydraulic fluid through the steering actuation
systems even when the wheels are not being turned. EPS is an enabler for all vehicle
hybridization technologies since it provides power steering when the engine is off. EPS may be
implemented on most vehicles with a standard 12V system. Some heavier vehicles such as Class
2b and 3 may require a higher voltage system which may add cost and complexity.
The 2017 light-duty final rule estimated a one to two percent effectiveness based on the
2002 NAS report, a Sierra Research report, and confidential manufacturer data. The SwRI report
estimated 0.8 percent to 1 percent effectiveness. The agencies reviewed these SwRI
effectiveness estimates and found them to be accurate, thus they have been retained for this rule.
In addition to the purely hybrid technologies, which decreases the proportion of
propulsion energy coming from the fuel by increasing the proportion of that energy coming from
electricity, there are other steps that can be taken to improve the efficiency of auxiliary functions
(e.g., power-assisted steering or air-conditioning) which also reduce CO2emissions and fuel
consumption. Optimization of the auxiliary functions is collectively referred to as vehicle or
accessory load electrification because they generally use electricity instead of engine power.
These improvements are considered enablers for hybrid systems.
2.4.11 Predictive Cruise Control
Cruise control is commonly used in light-duty and heavy-duty applications to maintain a
vehicle at a set speed. However, cruise control systems with additional intelligence and
predictive control are much more complex but offer opportunities to reduce fuel consumption
and GHG emissions. Many of the heavy-duty manufacturers are developing intelligent cruise
control systems and though they resemble each other in overall function, each manufacturer is
doing it differently.
As an example, an intelligent cruise control system partnered with a source of elevation
information could detect when the vehicle is on a hill and know when it is close to cresting the
hill. During this time, the vehicle may be allowed to temporarily travel at a lower speed to
prevent the need for a transmission downshift, which consumes more fuel because it requires the
engine to increase the rpm and run in a less efficient part of the fuel map. Similarly, predictive
cruise control allows a vehicle to exceed the speed set point by a specified amount so that the
vehicle will start the next hill at a higher speed and reduce the likelihood of needing to downshift
on the next hill.
The amount of reduction in fuel consumption and CO2 emissions depends significantly on the
terrain. Sources estimate that the overall savings is approximately two percent.102
2.5 Technology Application- HD Pickups and Vans
2.5.1 Gasoline Engines
Spark ignited (gasoline) engines used in complete Class 2b and 3 pickups and vans
include engines offered in a manufacturer's light-duty truck counterparts, as well as engines

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
specific to the Class 2b and 3 segment. Based on 2014 MY specifications, these engines
typically range in displacement between 5 and 7 liters, though smaller and larger engines have
also been used in this market. The majority of these engines are a V8 configuration, although the
VI0 configuration is also marketed.
The engine technologies are based on the technologies described in the Light-Duty
Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy Standards
Joint Technical Support Document and in Chapter 2. 3 above.103 (Note, however, that because
this section deals specifically with application to 2b/3 vehicles, the projected effectiveness may
vary from that presented in the generic discussions presented earlier). Some of the references
come from the 2010 NAS Report, Technologies and Approaches to Reducing the Fuel
Consumption of Medium and Heavy-Duty Vehicles. These technologies include engine friction
reduction, cam phasing, cylinder deactivation and stoichiometric gas direct injection. Included
with each technology description is an estimate of the improvement in fuel consumption and
GHGs that is achievable through the use of the technology in heavy-duty pickup trucks and vans
over their applicable operation and test cycles.
The technology effectiveness values are generally described as ranges that represent
expected levels of effectiveness with appropriate implementation of the technology but actual
effectiveness levels will vary with manufacturer-specific design, and with specifications for the
technologies. These may include considerations for durability or other related constraints. The
agencies did not receive comments disputing the expected technology effectiveness values
reported in the NPRM and draft RIA.
2.5.1.1 Low Friction Lubricants
One of the most basic methods of reducing fuel consumption in both gasoline and diesel
engines is the use of lower viscosity engine lubricants. More advanced multi-viscosity engine
oils are available today with improved performance in a wider temperature band and with better
lubricating properties. This can be accomplished by changes to the oil base stock (e.g.,
switching engine lubricants from a Group I base oils to lower-friction, lower viscosity Group III
synthetic) and through changes to lubricant additive packages (e.g., friction modifiers and
viscosity improvers). The use of 5W-30 motor oil is now widespread and auto manufacturers are
introducing the use of even lower viscosity oils, such as 5W-20 and 0W-20, to improve cold-
flow properties and reduce cold start friction. However, in some cases, changes to the
crankshaft, rod and main bearings and changes to the mechanical tolerances of engine
components may be required. In all cases, durability testing would be required to ensure that
durability is not compromised. The shift to lower viscosity and lower friction lubricants would
also improve the effectiveness of valvetrain technologies such as cylinder deactivation, which
rely on a minimum oil temperature (viscosity) for operation.
Based on light-duty 2017-2025 MY vehicle rulemaking, and previously-received
confidential manufacturer data, the agencies have estimated the effectiveness of low friction
lubricants to be between 0 to 3 percent.
We present cost estimates for this technology in Chapter 2.11 of this RIA.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.5.1.2	Engine Friction Reduction
Manufacturers can reduce friction and improve fuel consumption by improving the
design of engine components and subsystems. Approximately 10 percent of the energy
consumed by a vehicle is lost to friction, and just over half is due to frictional losses within the
engine. Examples include improvements in low-tension piston rings, piston skirt design, roller
cam followers, improved crankshaft design and bearings, material coatings, material substitution,
more optimal thermal management, and piston and cylinder surface treatments. Additionally, as
computer-aided modeling software continues to improve, more opportunities for evolutionary
friction reductions may become available.
Estimations of fuel consumption improvements due to reduced engine friction from the
2015 NHTSA Technology Study range from 1 percent to 3 percent. The agencies believe that
this range is accurate.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.1.3	Engine Parasitic Demand Reduction
Manufacturers can reduce mechanical engine loads and improve fuel consumption by
implementing variable-displacement oil pumps, higher-efficiency direct injection fuel pumps,
and variable speed/displacement coolant pumps.
Estimations of fuel consumption improvements due to reduced engine parasitic demand
from the 2015 NHTSA Technology Study range from 1 percent to 2 percent. The agencies
believe that this range is accurate.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.1.4	Variable Valve Timing
Variable valve timing (VVT) classifies a family of valve-train designs that alter the
timing of the intake valve, exhaust valve, or both, primarily to reduce pumping losses, increase
specific power, and control the level of residual gases in the cylinder. VVT reduces pumping
losses when the engine is lightly loaded by controlling valve timing closer to the optimum
needed to sustain horsepower and torque. VVT can also improve volumetric efficiency at higher
engine speeds and loads. Additionally, VVT can be used to alter (and optimize) the effective
compression ratio where it is advantageous for certain engine operating modes (e.g., in the
Atkinson Cycle).
VVT has now become a widely adopted technology in the light duty fleet: in MY 2014,
most of all new cars and light trucks had engines with some method of variable valve timing.104
Manufacturers are currently using many different types of variable valve timing, which have a
variety of different names and methods. Therefore, the degree of further improvement across the
fleet is limited by the level of valvetrain technology already implemented on the vehicles. The
three major types of VVT are listed below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Each of the implementations of VVT uses a cam phaser to adjust the camshaft angular
position relative to the crankshaft position, referred to as "camshaft phasing." The phase
adjustment results in changes to the pumping work required by the engine to accomplish the gas
exchange process. The majority of current cam phaser applications use hydraulically-actuated
units, powered by engine oil pressure and managed by a solenoid that controls the oil pressure
supplied to the phaser.
2.5.1.4.1	Coupled Cam Phasing for Overhead Valve (OHV) and Single
Overhead Camshaft (SOHC) Engines
Valvetrains with coupled (or coordinated) cam phasing (CCP) can modify the timing of
both the inlet valves and the exhaust valves an equal amount by varying the phasing of the
camshaft across an engine's range of operating speeds; also known as VVT. For engines
configured as an overhead valve (OHV) or as a single overhead camshaft (SOHC) only one cam
phaser is required per camshaft to achieve CCP.
Based on the heavy-duty 2014-2018 MY vehicle rulemaking, 2015 NHTSA Technology
Study, and previously-received confidential manufacturer data, the agencies estimate the fuel
consumption reduction effectiveness of this technology to be between 1 and 3 percent.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.1.4.2	Intake Cam Phasing (ICP) for Dual Overhead Camshaft Engines
(DOHC)
Valvetrains with ICP, which is the simplest of the cam phasing technologies, can modify
the timing of the inlet valves by phasing the intake camshaft while the exhaust valve timing
remains fixed. This requires the addition of a cam phaser on each bank of intake valves on the
engine. An in-line 4-cylinder engine has one bank of intake valves, while V-configured engines
have two banks of intake valves.
Some newer Class 2b and 3 market entries are offering dual overhead camshaft (DOHC)
engine designs where two camshafts are used to operate the intake and exhaust valves
independently. Consistent with the heavy-duty 2014-2018 MY vehicle rulemaking and the SwRI
report, the agencies agree with the effectiveness values of 1 to 2 percent reduction in fuel
consumption for this technology.
2.5.1.4.3	Dual Cam Phasing (DCP) for Dual Overhead Camshaft Engines
(DOHC)
The most flexible VVT design is dual (independent) cam phasing, where the intake and
exhaust valve opening and closing events are controlled independently. This option allows the
option of controlling valve overlap, which can be used as an internal EGR strategy. At low
engine loads, DCP creates a reduction in pumping losses, resulting in improved fuel
consumption. Increased internal EGR also results in lower engine-out NOx emissions. The
amount by which fuel consumption is improved depends on the residual tolerance of the
combustion system. Additional improvements are observed at idle, where low valve overlap

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
could result in improved combustion stability, potentially reducing idle fuel consumption. DCP
requires two cam phasers on each bank of the engine.
Some newer Class 2b and 3 market entries are offering dual overhead camshaft (DOHC)
engine designs where two camshafts are used to operate the intake and exhaust valves
independently. Consistent with the light-duty 2012-2016 MY vehicle rulemaking and the SwRI
report, the agencies agree with the effectiveness values of 1 to 3 percent reduction in fuel
consumption for this technology.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.1.5 Variable Valve Lift (VVL)
Controlling the lift of the valves provides a potential for further efficiency improvements.
By optimizing the valve-lift profile for specific engine operating regions, the pumping losses can
be reduced by reducing the amount of throttling required to produce the desired engine power
output. By moving the throttling losses further downstream of the throttle valve, the heat
transfer losses that occur from the throttling process are directed into the fresh charge-air mixture
just prior to compression, delaying the onset of knock-limited combustion processes. Variable
valve lift control can also be used to induce in-cylinder mixture motion, which improves fuel-air
mixing and can result in improved thermodynamic efficiency. Variable valve lift control can
also potentially reduce overall valvetrain friction. At the same time, such systems may also incur
increased parasitic losses associated with their actuation mechanisms. A number of
manufacturers have already implemented VVL into their fleets (Toyota, Honda, and BMW).
There are two major classifications of variable valve lift, described below:
2.5.1.5.1 Discrete Variable Valve Lift (DWL)
Discrete variable valve lift (DVVL) systems allow the selection between two or three
discrete cam profiles by means of a hydraulically-actuated mechanical system. By optimizing
the cam profile for specific engine operating regions, the pumping losses can be reduced by
reducing the amount of throttling required to produce the desired engine power output. This
increases the efficiency of the engine. These cam profiles consist of a low and a high-lift lobe,
and may include an inert or blank lobe to incorporate cylinder deactivation (in the case of a 3-
step DVVL system). DVVL is normally applied together with VVT control. DVVL is also
known as Cam Profile Switching (CPS). DVVL is a mature technology with low technical risk.
Based on the light-duty MY 2017-2025 final rule, previously-received confidential
manufacturer data, 2015 NHTSA Technology Study, and report from the Northeast States Center
for a Clean Air Future (NESCCAF), the agencies estimate the fuel consumption reduction
effectiveness of this technology to be between 1 and 3 percent.
We present cost estimates for this technology in Chapter 2.11 of this RIA.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.5.1.6	Cylinder Deactivation
In conventional spark-ignited engines throttling the airflow controls engine torque output.
At partial loads, efficiency can be improved by using cylinder deactivation instead of throttling.
Cylinder deactivation can improve engine efficiency by disabling or deactivating (usually) half
of the cylinders when the load is less than half of the engine's total torque capability - the valves
are kept closed, and no fuel is injected - as a result, the trapped air within the deactivated
cylinders is simply compressed and expanded as an air spring, with reduced friction and heat
losses. The active cylinders combust at almost double the load required if all of the cylinders
were operating. Pumping losses are significantly reduced as long as the engine is operated in
this "part-cylinder" mode.
Cylinder deactivation control strategy relies on setting maximum manifold absolute
pressures or predicted torque within which it can deactivate the cylinders. Noise and vibration
issues reduce the operating range to which cylinder deactivation is allowed, although
manufacturers are exploring vehicle changes that enable increasing the amount of time that
cylinder deactivation might be suitable. Some manufacturers may choose to adopt active engine
mounts and/or active noise cancellations systems to address Noise Vibration and Harshness
(NVH) concerns and to allow a greater operating range of activation.
Effectiveness improvements scale roughly with engine displacement-to-vehicle weight
ratio: the higher displacement-to-weight vehicles, operating at lower relative loads for normal
driving, have the potential to operate in part-cylinder mode more frequently.
Based on the 2015 NHTSA Technology Study and previously-received confidential
manufacturer data, the agencies estimate the fuel consumption reduction effectiveness of this
technology to be between 0 and 3 percent.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.1.7	Stoichiometric Gasoline Direct Injection
Stoichiometric gasoline direct injection (SGDI) engines inject fuel at high pressure
directly into the combustion chamber (rather than the intake port in port fuel injection). SGDI
requires changes to the injector design, an additional high pressure fuel pump, new fuel rails to
handle the higher fuel pressures, and changes to the cylinder head and piston crown design.
Direct injection of the fuel into the cylinder improves cooling of the air/fuel charge within the
cylinder, which allows for higher compression ratios and increased thermodynamic efficiency
without the onset of combustion knock. Recent injector design advances, improved electronic
engine management systems and the introduction of multiple injection events per cylinder firing
cycle promote better mixing of the air and fuel, enhance combustion rates, increase residual
exhaust gas tolerance and improve cold start emissions. SGDI engines achieve higher power
density and match well with other technologies, such as boosting and variable valvetrain designs.
Several manufacturers have recently introduced vehicles with SGDI engines, including
GM and Ford, who have announced their plans to increase dramatically the number of SGDI
engines in their light-duty portfolios.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Based on the heavy-duty 2014-2018 MY vehicle rulemaking, 2015 NHTSA Technology
Study, and previously-received confidential manufacturer data, the agencies estimate the fuel
consumption reduction effectiveness of SGDI to be between 1 and 2 percent.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.1.8 Turbocharging and Downsizing (TRBDS)
The specific power of a naturally aspirated engine is primarily limited by the rate at
which the engine is able to draw air into the combustion chambers. Turbocharging and
supercharging (grouped together here as boosting) are two methods to increase the intake
manifold pressure and cylinder charge-air mass above naturally aspirated levels. Boosting
increases the airflow into the engine, thus increasing the specific power level, and with it the
ability to reduce engine displacement while maintaining performance. This effectively reduces
the pumping losses at lighter loads in comparison to a larger, naturally aspirated engine.
Almost every major manufacturer currently markets a vehicle with some form of
boosting. While boosting has been a common practice for increasing performance for several
decades, turbocharging has considerable potential to improve fuel economy and reduce CO2
emissions when the engine displacement is also reduced. Specific power levels for a boosted
engine often exceed 100 hp/L, compared to average naturally aspirated engine power densities of
roughly 70 hp/L. As a result, engines can be downsized roughly 30 percent or higher while
maintaining similar peak output levels. In the last decade, improvements to turbocharger turbine
and compressor design have improved their reliability and performance across the entire engine
operating range. New variable geometry turbines and ball-bearing center cartridges allow faster
turbocharger spool-up (virtually eliminating the once-common "turbo lag") while maintaining
high flow rates for increased boost at high engine speeds. Low speed torque output has been
dramatically improved for modern turbocharged engines. However, even with turbocharger
improvements, maximum engine torque at very low engine speed conditions, for example launch
from standstill, is increased less than at mid and high engine speed conditions. The potential to
downsize engines may be less on vehicles with low displacement to vehicle mass ratios for
example a very small displacement engine in a vehicle with significant curb weight, in order to
provide adequate acceleration from standstill, particularly up grades or at high altitudes.
Use of GDI systems with turbocharged engines and charge air cooling also reduces the
fuel octane requirements for knock limited combustion and allows the use of higher compression
ratios. Ford's "EcoBoost" downsized, turbocharged GDI engines introduced on MY 2010
vehicles allow the replacement of V8 engines with V6 engines with improved in 0-60 mph
acceleration and with fuel economy improvements of up to 12 percent.105
Recently published data with advanced spray-guided injection systems and more
aggressive engine downsizing targeted towards reduced fuel consumption and CO2 emissions
reductions indicate that the potential for reducing CO2 emissions for turbocharged, downsized
GDI engines may be as much as 15 to 30 percent relative to port-fuel-injected engines.14'15'16'17'18
Confidential manufacturer data suggests an incremental range of fuel consumption and CO2
emission reduction of 4.8 to 7.5 percent for turbocharging and downsizing. Other publicly-
available sources suggest a fuel consumption and CO2 emission reduction of 8 to 13 percent

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
compared to current-production naturally-aspirated engines without friction reduction or other
fuel economy technologies: a joint technical paper by Bosch and Ricardo suggesting fuel
economy gain of 8 to 10 percent for downsizing from a 5.7 liter port injection V8 to a 3.6 liter
V6 with direct injection using a wall-guided direct injection system;106 a Renault report
suggesting a 11.9 percent NEDC fuel consumption gain for downsizing from a 1.4 liter port
injection in-line 4-cylinder engine to a 1.0 liter in-line 4-cylinder engine, also with wall-guided
direct injection;107 and a Robert Bosch paper suggesting a 13 percent NEDC gain for downsizing
to a turbocharged DI engine, again with wall-guided injection.108 These reported fuel economy
benefits show a wide range depending on the SGDI technology employed.
The agencies reviewed estimates from the LD 2017-2025 final rule, the TSD, and
existing public literature. The previous estimate from the MYs 2017-2025 suggested a 12 to 14
percent effectiveness improvement, which included low friction lubricant (level one), engine
friction reduction (level one), DCP, DVVL and SGDI, over baseline fixed-valve engines, similar
to the estimate for Ford's EcoBoost engine, which is already in production. Additionally, the
agencies analyzed Ricardo vehicle simulation data and the 2015 NHTSA Technology Study for
various turbocharged engine packages. Based on these data, and considering the widespread
nature of the public estimates, the agencies assume that turbocharging and downsizing, would
provide a 16.4 percent effectiveness improvement over naturally aspirated engines as applied to
Class 2b and 3 vehicles.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
Note that for this analysis we determined that this technology path is only applicable to
heavy duty applications that have operating conditions more closely associated with light duty
vehicles. This includes vans designed mainly for cargo volume or modest payloads having
similar GCWR to light duty applications. These vans cannot tow trailers heavier than similar
light duty vehicles and are largely already sharing engines of significantly smaller displacement
and cylinder count compared to heavy duty vehicles designed mainly for trailer towing.
2.5.1.9 Cooled Exhaust-Gas Recirculation
Cooled exhaust gas recirculation or Boosted EGR is a combustion concept that involves
utilizing EGR as a charge diluent for controlling combustion temperatures and cooling the EGR
prior to its introduction to the combustion system. Higher exhaust gas residual levels at part load
conditions reduce pumping losses for increased fuel economy. The additional charge dilution
enabled by cooled EGR reduces the incidence of knocking combustion and obviates the need for
fuel enrichment at high engine power. This allows for higher boost pressure and/or compression
ratio and further reduction in engine displacement and both pumping and friction losses while
maintaining performance. Engines of this type use GDI and both dual cam phasing and discrete
variable valve lift. The EGR systems considered in this rule would use a dual-loop system with
both high and low pressure EGR loops and dual EGR coolers. The engines would also use
single-stage, variable geometry turbocharging with higher intake boost pressure available across
a broader range of engine operation than conventional turbocharged SI engines. Such a system is
estimated to be capable of an additional 3 to 5 percent effectiveness relative to a turbocharged,
downsized GDI engine without cooled-EGR. The agencies have also considered a more

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
advanced version of such a cooled EGR system that employs very high combustion pressures by
using dual stage turbocharging.
2.5.2 Diesel Engines
Diesel engines in this class of vehicle have emission characteristics that present
challenges to meeting federal NOx emissions standards. It is a significant systems-engineering
challenge to maintain the fuel consumption advantage of the diesel engine while meeting U.S.
emissions regulations. Fuel consumption can be negatively impacted by emissions reduction
strategies depending on the combination of strategies employed. Emission compliance strategies
for diesel vehicles sold in the U.S. are expected to include a combination of improvements of
combustion, air handling system, aftertreatment, and advanced system control optimization.
These emission control strategies are being introduced on Tier 2 light-duty diesel vehicles today.
Some of the engine technologies are described in the Light-Duty Vehicle Greenhouse
Gas Emission Standards and Corporate Average Fuel Economy Standards Joint Technical
Support Document.109 Others are from the 2010 NAS Report, Technologies and Approaches to
Reducing the Fuel Consumption of Medium- and Heavy-Duty Vehicles, and the 2015 NHTSA
Technology Study. Several key advances in diesel technology have made it possible to reduce
emissions coming from the engine prior to aftertreatment. These technologies include engine
friction and parasitic loss reduction, improved fuel systems (higher injection pressure and
multiple-injection capability), advanced controls and sensors to optimize combustion and
emissions performance, higher EGR levels and EGR cooling to reduce NOx, and advanced
turbocharging systems.
2.5.2.1	Low Friction Lubricants
Consistent with the discussion above for gasoline engines (see Chapter 2.5.1), the
agencies are expecting some engine changes to accommodate low friction lubricants. Based on
the light-duty 2014-2018 MY HD vehicle rulemaking, and previously-received confidential
manufacturer data, the agencies estimated the effectiveness of low friction lubricants to be
between 0 and 3 percent.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
Based on a survey of the current powertrains being applied to the Class 2b and 3 segment
and the level of powertrain sharing with the light duty vehicle market for these vehicles, the
majority of light heavy duty gasoline engines in the 2014 Class 2b and 3 vehicle models are
utilizing some form of low friction lubricants to achieve power and emission goals, and so this
technology is considered to be in the baseline.
2.5.2.2	Engine Friction Reduction
Reduced friction in bearings, valve trains, and the piston-to-liner interface will improve
efficiency. Friction reduction opportunities in the engine valve train and at its roller/tappet
interfaces exist for several production engines. In virtually all production engines, the piston at
its skirt/cylinder wall interface, wrist pin and oil ring/cylinder wall interface offer opportunities

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
for friction reduction. Use of more advanced oil lubricant that could be available for production
in the future may also eventually play a key role in reducing friction. Mechanical loads can also
be reduced by converting the water, oil, and fuel pumps in the engine from fixed displacement to
variable displacement.
Estimations of fuel consumption improvements due to reduced engine friction from the
2015 NHTSA Technology Study range from 1 percent to 2 percent. The agencies believe that
this range is accurate.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.2.3	Turbocharger Technology
Compact two stage turbochargers can increase the boost level with wider operation range,
thus improving engine thermal efficiency. Ford's new developed 6.7L Scorpion engine features
a twin-compressor turbocharger110. Cummins has also developed its own two stage
turbochargers.111 It is expected that this type of technology will continue to be improved by
better system matching and development of higher compressor and turbine efficiency.
Based on the 2015 NHTSA Technology Study and previously-received confidential
manufacturer data, the agencies estimate the fuel consumption reduction effectiveness of this
technology to be between 2 and 3 percent.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.2.4	Reduction of Parasitic Loads
Accessories that are traditionally gear- or belt-driven by a vehicle's engine can be
optimized and/or converted to electric power. Examples include the engine water pump, oil
pump, fuel injection pump, air compressor, power-steering pump, cooling fans, and the vehicle's
air-conditioning system which can be converted to full electrically driven loads or an electro-
mechanical arrangement that retains some mechanically connected aspects. Optimization and
improved pressure regulation may significantly reduce the parasitic load of the water, air and
fuel pumps. Electrification may result in a reduction in power demand, because electrically-
powered accessories (such as the air compressor or power steering) operate only when needed if
they are electrically powered, but they impose a parasitic demand all the time if they are engine-
driven. In other cases, such as cooling fans or an engine's water pump, electric power allows the
accessory to run at speeds independent of engine speed, which can reduce power consumption.
The 2015 NHTSA Technology Study used a 1 to 2 percent fuel consumption reduction for diesel
engine parasitic improvements.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.2.5	Aftertreatment Improvements
The HD diesel pickup and van segment has largely adopted the SCR type of
aftertreatment system to comply with criteria pollutant emission standards. As the experience
base for SCR expands over the next few years, many improvements in this aftertreatment system

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
such as construction of the catalyst, thermal management, and reductant optimization may result
in a reduction in the amount of fuel consumed by the engine via combustion optimization, taking
advantage of the SCR's capability to reduce higher levels of NOx emitted by the engine.
However, due to uncertainties with these improvements regarding the extent of current
optimization and future criteria emissions obligations, the agencies are not considering
aftertreatment improvements as a fuel-saving technology in the rulemaking analysis for HD
pickups and vans.
2.5.3 Drivetrain
The agencies have also reviewed the transmission technology estimates used in the light-
duty 2012-2016 MY vehicle rulemaking. In doing so, the agencies have considered or
reconsidered all available sources and updated the estimates as appropriate. The section below
describes each of the transmission technologies considered for this rulemaking.
2.5.3.1	Automatic 8-Speed Transmissions
Manufacturers can also choose to replace 6-speed transmissions with transmissions
capable of 8-speeds or more. Additional ratios allow for further optimization of engine operation
over a wider range of conditions, but this is subject to diminishing returns as the number of
speeds increases. As additional gear sets are added (which may be necessary in some cases to
achieve the higher number of ratios), additional weight and friction are introduced. Also, the
additional shifting of such a transmission can be perceived as bothersome to some consumers, so
manufacturers continue to develop strategies for smooth operation.
As discussed in the heavy-duty 2014-2018 MY vehicle rulemaking, taking into account
confidential manufacturer data, we projected that 8-speed transmissions could incrementally
reduce fuel consumption by 1 to 3 percent from a baseline 6-speed automatic transmission. The
SwRI report uses 2 to 3 percent fuel consumption reduction when replacing 6-speed baseline
automatic transmissions with improved 8-speed automatic transmissions.
The agencies reviewed and revised these effectiveness estimates based on usage and
testing methods for Class 2b and 3 vehicles. The agencies estimate the effectiveness for a
conversion from a 6 to 8-speed transmission to be 2.7 percent.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.3.2	High Efficiency Transmission
For this rule, a high efficiency transmission refers to some or all of a suite of incremental
transmission improvement technologies that should be available within the 2019 to 2025
timeframe. The majority of these improvements address mechanical friction within the
transmission. These improvements include but are not limited to: shifting clutch technology
improvements, improved kinematic design, dry sump lubrication systems, more efficient seals,
bearings and clutches (reducing drag), component superfinishing and improved transmission
lubricants.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.5.3.3	Electric Power Steering (EPS)
Electric power steering (EPS) provides a potential reduction in CO2 emissions and fuel
consumption over hydraulic power steering because of reduced overall accessory loads. This
eliminates the parasitic losses associated with belt-driven power steering pumps which
consistently draw load from the engine to pump hydraulic fluid through the steering actuation
systems even when the wheels are not being turned. EPS is an enabler for all vehicle
hybridization technologies since it provides power steering when the engine is off. EPS may be
implemented on most vehicles with a standard 12V system. Some heavier vehicles such as Class
2b and 3 may require a higher voltage system which may add cost and complexity.
The 2017 light-duty final rule estimated a 1 to 2 percent effectiveness based on the 2002
NAS report, a Sierra Research report, and confidential manufacturer data. The SwRI report
estimated 0.8 percent to 1 percent effectiveness. The agencies reviewed these SwRI
effectiveness estimates and found them to be accurate, thus they have been retained for this rule.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.3.4	Improved Accessories
The accessories on an engine, including the alternator, coolant and oil pumps are
traditionally mechanically-driven. A reduction in CO2 emissions and fuel consumption can be
realized by driving them electrically, and only when needed ("on-demand").
Electric water pumps and electric fans can provide better control of engine cooling. For
example, coolant flow from an electric water pump can be reduced and the radiator fan can be
shut off during engine warm-up or cold ambient temperature conditions which would reduce
warm-up time, reduce warm-up fuel enrichment, and reduce parasitic losses.
Indirect benefit may be obtained by reducing the flow from the water pump electrically
during the engine warm-up period, allowing the engine to heat more rapidly and thereby
reducing the fuel enrichment needed during cold operation and warm-up of the engine. Faster oil
warm-up may also result from better management of the coolant warm-up period. Further
benefit may be obtained when electrification is combined with an improved, higher efficiency
engine alternator used to supply power to the electrified accessories.
Intelligent cooling can more easily be applied to vehicles that do not typically carry
heavy payloads, so larger vehicles with towing capacity present a challenge, as these vehicles
have high cooling fan loads.A However, towing vehicles tend to have large cooling system
capacity and flow scaled to required heat rejection levels when under full load situations such as
towing at GCWR in extreme ambient conditions. During almost all other situations, this design
characteristic may result in unnecessary energy usage for coolant pumping and heat rejection to
the radiator.
A In the CAFE model, improved accessories refers solely to improved engine cooling. However, EPA has included
a high efficiency alternator in this category, as well as improvements to the cooling system.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The agencies considered whether to include electric oil pump technology for the
rulemaking. Because it is necessary to operate the oil pump any time the engine is running,
electric oil pump technology has insignificant effect on efficiency. Therefore, the agencies
decided to not include electric oil pump technology.
2.5.3.5	Mild Hybrid (MHEV)
Mild hybrid systems offer idle-stop functionality and a limited level of regenerative
braking and power assist. These systems replace the conventional alternator with a belt or crank
driven starter/alternator and may add high voltage electrical accessories (which may include
electric power steering and an auxiliary automatic transmission pump). The limited electrical
requirements of these systems allow the use of lead-acid batteries or supercapacitors for energy
storage, or the use of a small lithium-ion battery pack.
For the MHEV technology the agencies sized the system using a 7 kW starter/generator
and 8 kWh Li-ion battery pack. The estimates were developed by Argonne National Laboratory
as a supplement to the 2015 NHTSA Technology Study, resulting in an effectiveness range of 4
to 5 percent depending on the vehicle's engine.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.3.6	Strong Hybrid (SHEV)
A hybrid vehicle is a vehicle that combines two significant sources of propulsion energy,
where one uses a consumable fuel (like gasoline), and one is rechargeable (during operation, or
by another energy source). Hybrid technology is well established in the U.S. market and more
manufacturers are adding hybrid models to their lineups. Hybrids reduce fuel consumption
through three major mechanisms:
•	The internal combustion engine can be optimized (through downsizing, modifying the
operating cycle, or other control techniques) to operate at or near its most efficient
point more of the time. Power loss from engine downsizing can be mitigated by
employing power assist from the secondary power source.
•	Some of the energy normally lost as heat while braking can be captured and stored in
the energy storage system for later use.
•	The engine is turned off when it is not needed, such as when the vehicle is coasting or
when stopped.
Hybrid vehicles utilize some combination of the three above mechanisms to reduce fuel
consumption and CO2 emissions. The effectiveness of fuel consumption and CO2 reduction
depends on the utilization of the above mechanisms and how aggressively they are pursued. One
area where this variation is particularly prevalent is in the choice of engine size and its effect on
balancing fuel economy and performance. Some manufacturers choose not to downsize the
engine when applying hybrid technologies. In these cases, performance is vastly improved,
while fuel efficiency improves significantly less than if the engine was downsized to maintain

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
the same performance as the conventional version. The non-downsizing approach is used for
vehicles like trucks where towing and/or hauling are an integral part of their performance
requirements. In these cases, if the engine is downsized, the battery can be quickly drained
during a long hill climb with a heavy load, leaving only a downsized engine to carry the entire
load. Because towing capability is currently a heavily-marketed truck attribute, manufacturers
are hesitant to offer a truck with downsized engine which can lead to a significantly diminished
towing performance when the battery state of charge level is low, and therefore engines are
traditionally not downsized for these vehicles. As noted above, in assessing costs of this
technology, the agencies assumed in all instances that the engine would not be downsized.
Strong Hybrid technology utilizes an axial electric motor connected to the transmission
input shaft and connected to the engine crankshaft through a clutch. The axial motor is a
motor/generator that can provide sufficient torque for launch assist, all electric operation, and the
ability to recover significant levels of braking energy.
For SHEV, the agencies also relied on the study by Argonne National Laboratory to
supplement the 2015 NHTSA Technology Study to determine that the effectiveness of these
systems in terms of CO2 reduction. For the SHEV technology, the agencies sized the system
using a 50 kW starter/generator and a 70 kWh Li-ion battery pack. The estimates resulted in an
effectiveness range of 18 to 22 percent depending on the engine. The estimates assume no
engine downsizing so as to maintain vehicle performance and/or maintain towing and hauling
performance.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.4 Aerodynamics
Aerodynamic drag is an important aspect of the power requirements for Class 2b and 3
trucks. Because aerodynamic drag is a function of the cube of vehicle speed, small changes in
the aerodynamics of a Class 2b and 3 can reduce drag, fuel consumption, and GHG emissions.
Some of the opportunities to reduce aerodynamic drag in Class 2b and 3 vehicles are similar to
those in Class 1 and 2 (i.e., light-duty) vehicles. In general, these transferable features make the
cab shape more aerodynamic by streamlining the airflow over the bumper, grill, windshield,
sides, and roof. Class 2b and 3 vehicles may also borrow from light-duty vehicles certain drag
reducing accessories (e.g., streamlined mirrors, operator steps, and sun visors). The great variety
of applications for Class 2b and 3 trucks result in a wide range of operational speed profiles (i.e.,
in-use drive cycles) and functional requirements (e.g., shuttle buses that must be tall enough for
standing passengers, trucks that must have racks for ladders). This variety makes it challenging
to develop aerodynamic solutions that consider the entire vehicle.
Many factors affect a vehicle's aerodynamic drag and the resulting power required to
move it through the air. While these factors change with air density and the square and cube of
vehicle speed, respectively, the overall drag effect is determined by the product of its frontal area
and drag coefficient. Reductions in these quantities can therefore reduce fuel consumption and
CO2 emissions. Although frontal areas tend to be relatively similar within a vehicle class
(mostly due to market-competitive size requirements), significant variations in drag coefficient
can be observed. Significant changes to a vehicle's aerodynamic performance may need to be

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
implemented during a redesign (e.g., changes in vehicle shape). However, shorter-term
aerodynamic reductions, with a somewhat lower effectiveness, may be achieved through the use
of revised exterior components (typically at a model refresh in mid-cycle) and add-on devices
that are currently being applied. The latter list would include revised front and rear fascias,
modified front air dams and rear valances, addition of rear deck lips and underbody panels, and
lower aerodynamic drag exterior mirrors.
For this rule, the agencies considered two levels of aero improvements. The first level
includes such body features as air dams, tire spats, and perhaps one underbody panel resulting in
a 5 percent aerodynamic drag reduction. The agencies estimated the CO2 and fuel consumption
effectiveness of this first level of aerodynamic drag at 0.75 percent.
The second level which includes the features of level 1 plus additional body features such
as active grille shutters6, rear visors, larger under body panels or low-profile roof racks resulting
in a 10 percent aerodynamic drag reduction. The agencies estimated the CO2 and fuel
consumption effectiveness of this second level of aerodynamic drag at 1.5 percent. We present
cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.5	Tires
Typically, tires used on Class 2b/3 vehicles are not designed specifically for the vehicle.
These tires are designed for broader use and no single parameter is optimized. Similar to
vocational vehicles, the market has not demanded tires with improved rolling resistance thus far;
therefore, manufacturers have not traditionally designed tires with low rolling resistance for
Class 2b/3 vehicles. The agencies believe that a regulatory program that incentivizes the
optimization of tire rolling resistance, traction and durability can bring about GHG emission and
fuel consumption reductions of 1.1 percent from this segment based on a 10 percent reduction in
rolling resistance.
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.5.6	Mass Reduction
Mass reduction is a technology that can be used in a manufacturer's strategy to meet the
Heavy Duty Greenhouse Gas Phase 2 standards (although, as noted, it is not part of the agencies'
projected technology path for either the standards for pickups and vans, or any of the other
standards). Vehicle mass reduction (also referred to as "light-weighting"), decreases fuel
consumption and GHG emissions by reducing the energy demand needed to overcome inertia
forces, and rolling resistance. Automotive companies have worked with mass reduction
technologies for many years and a lot of these technologies have been used in production
vehicles. The weight savings achieved by adopting mass reduction technologies offset weight
gains due to increased vehicle size, larger powertrains, and increased feature content (sound
B For details on how active aerodynamics are considered for off-cycle credits, see the Technical Support Document
for Final Rulemaking for 2017-2025 Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate
Average Fuel Economy, August 2012, Chapter 5.2.2.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
insulation, entertainment systems, improved climate control, panoramic roof, etc.). Sometimes
mass reduction has been used to increase vehicle towing and payload capabilities.
Manufacturers employ a systematic approach to mass reduction, where the net mass
reduction is the addition of a direct component or system mass reduction, also referred to as
primary mass reduction, plus the additional mass reduction taken from indirect ancillary systems
and components, also referred to as secondary mass reduction or mass compounding. There are
more secondary mass reductions achievable for light-duty vehicles compared to heavy-duty
vehicles, which are limited due to the higher towing and payload requirements.
Mass reduction can be achieved through a number of approaches, even while maintaining
other vehicle functionalities. As summarized by NAS in its 2011 light duty vehicle report, there
are two key strategies for primary mass reduction: 1) changing the design to use less material; 2)
substituting lighter materials for heavier materials.112
The first key strategy of using less material compared to the baseline component can be
achieved by optimizing the design and structure of vehicle components, systems and vehicle
structure. Vehicle manufacturers have long used these continually-improving CAE tools to
optimize vehicle designs. For example, the Future Steel Vehicle (FSV) project sponsored by
World Auto Steel used three levels of optimization: topology optimization, low fidelity 3G
(Geometry Grade and Gauge) optimization, and subsystem optimization, to achieve 30 percent
mass reduction in the body structure of a vehicle with a mild steel unibody structure.113 Using
less material can also be achieved through improving the manufacturing process, such as by
using improved joining technologies and parts consolidation. This method is often used in
combination with applying new materials.
The second key strategy to reduce mass of an assembly or component involves the
substitution of lower density and/or higher strength materials. Material substitution includes
replacing materials, such as mild steel, with higher-strength and advanced steels, aluminum,
magnesium, and composite materials. In practice, material substitution tends to be quite specific
to the manufacturer and situation. Some materials work better than others for particular vehicle
components, and a manufacturer may invest more heavily in adjusting to a particular type of
advanced material, thus complicating its ability to consider others. The agencies recognize that
like any type of mass reduction, material substitution has to be conducted not only with
consideration to maintaining equivalent component strength, but also to maintaining all the other
attributes of that component, system or vehicle, such as crashworthiness, durability, and noise,
vibration and harshness (NVH).
If vehicle mass is reduced sufficiently through application of the two primary strategies
of using less material and material substitution described above, secondary mass reduction
options may become available. Secondary mass reduction is enabled when the load requirements
of a component are reduced as a result of primary mass reduction. If the primary mass reduction
reaches a sufficient level, a manufacturer may use a smaller, lighter, and potentially more
efficient powertrain while maintaining vehicle acceleration performance. If a powertrain is
downsized, a portion of the mass reduction may be attributed to the reduced torque requirement
which results from the lower vehicle mass. The lower torque requirement enables a reduction in
engine displacement, changes to transmission torque converter and gear ratios, and changes to

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
final drive gear ratio. The reduced powertrain torque enables the downsizing and/or mass
reduction of powertrain components and accompanying reduced rotating mass (e.g., for
transmission, driveshafts/halfshafts, wheels, and tires) without sacrificing powertrain
durability. Likewise, the combined mass reductions of the engine, drivetrain, and body in turn
reduce stresses on the suspension components, steering components, wheels, tires, and brakes,
which can allow further reductions in the mass of these subsystems. Reducing the unsprung
masses such as the brakes, control arms, wheels, and tires further reduce stresses in the
suspension mounting points, which will allow for further optimization and potential mass
reduction. However, pickup trucks have towing and hauling requirements which must be taken
into account when determining the amount of secondary mass reduction that is possible and so it
is less than that of passenger cars.
In September 2015, Ford announced that its MY 2017 F-Series Super duty pickup (F-
250) would be manufactured with an aluminum body and overall the truck will be 350 lbs lighter
(5 to 6 percent) than the current gen truck with steel.114'115 This is less overall mass reduction
than the resultant lightweighting effort on the MY 2015 F-150 which achieved up to a 750 lb
decrease in curb weight (12 to 13 percent) per vehicle.116 Strategies were employed in the F-250
to "improve the productivity of the Super Duty" in addition there were several safety systems
added including cameras, lane departure warning, brake assist, etc. If some of the mass
reduction efforts were not offset by other vehicle upgrades (size, towing, hauling, etc.), then
more mass reduction and greater fuel economy could have been realized. More details on the F-
250 will be known once it is released; however, a review of the F-150 vehicle aluminum
intensive design shows that it has an aluminum cab structure, body panels, and suspension
components, as well as a high strength steel frame and a smaller, lighter and more efficient
engine. The Executive Summary to Ducker Worldwide's 2014 report117 states that state that the
MY 2015 F-150 contains 1080 lbs of aluminum with at least half of this being aluminum sheet
and extrusions for body and closures. Ford's engine options for its light duty truck fleet includes
a 2.7L EcoBoost V-6. The integrated loop between Ford, the aluminum sheet suppliers, and the
aluminum scrap suppliers is integral to making aluminum a feasible lightweighting technology
option for Ford. It is also possible that the strategy of using aluminum body panels will be
applied to the heavy duty F-350 version when it is redesigned.118
We present cost estimates for this technology in Chapter 2.11 of this RIA.
2.6 Technology Application- SI Engines
This section summarizes the technologies the agencies project as a feasible path to
meeting the engine standards for spark-ignition engines used in vocational vehicles - that is
engines that are engine-certified and intended for vocational vehicles that will be GEM-certified.
These standards apply with respect to emissions measured over the FTP test cycle. This cycle is
described in Chapter 3.1. See Chapter 2.5 for spark-ignited engine technologies projected for the
Phase 2 HD pickup and van vehicle standards.
For the reasons discussed below, rather than setting a more stringent engine standard, the
agencies will maintain the MY 2016 fuel consumption and CO2 emission standards for SI
engines for use in vocational vehicles: 7.06 gallon/100 bhp-hr and 627 g CCh/bhp-hr, as
measured over the Heavy-duty FTP engine test cycle.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Heavy-duty spark-ignited (SI) engines are used in almost 30 percent of vocational
vehicles. Operators that choose gasoline engines do so for reasons similar to those for HD
complete pickups and vans. Gasoline engines have the advantage of being less expensive and
lower weight than diesels, but tend to also be less durable and have higher fuel consumption.
Thus, gasoline engines are most likely to be purchased for applications with lower annual VMT,
where fuel costs are less important than upfront costs.
When an Si-powered vocational vehicle is built by a non-integrated chassis manufacturer,
the engine is generally purchased from a company that also produces complete and/or
incomplete HD pickup trucks and vans. The primary certification path intended in this scenario
is for the engine to be engine-certified over the FTP and the vehicle to be GEM certified under
the GHG rules. This is common practice for CI engines, and in Phase 2 the agencies are
continuing this as the primary certification path for SI engines intended for vocational vehicles.
In Phase 1 we adopted a special provision aimed at simplifying compliance for
manufacturers of complete HD pickups and vans that also sell a relatively small number of
engines to non-integrated chassis manufacturers. This flexibility provision enables these
manufacturers to avoid meeting the separate SI engine standard, instead averaging them into the
applicable HD pickup and van fleet-wide average.0 These "loose" engine sales represent a very
small fraction of the Si-powered vocational vehicle market. The final Phase 2 program allows SI
engine manufacturers to sell a limited number of these "loose" SI engines to other chassis
manufacturers for use in vocational vehicles, through MY 2023.
The SI engines certified and sold as loose engines into the heavy-duty vocational vehicle
market are typically large V8 and V10 engines produced by General Motors and Ford. The
number of engine families certified in the past for this segment of vehicles is very limited and
has ranged between three and five engine models.119 Unlike the heavy-duty diesel engines
typical of this segment that are built for vocational vehicles, these SI engines are primarily
developed for chassis-certified heavy-duty pickup trucks and vans, but are also installed in
incomplete vocational vehicles.
Under the special Phase 1 provision, these loose engines need not be certified to engine-
based GHG and fuel consumption standards, but instead may be treated under the regulations as
though they are additional sales of the manufacturer's complete pickup and van products, on a
one-for-one basis. The pickup/van vehicle so chosen must be the vehicle with the highest
emission test weight that uses the engine (as this vehicle is likely to have the highest GHG
emissions and fuel consumption).0 However, if this vehicle is a credit-generator under the HD
pickup and van fleet averaging program, no credits would be generated by these engine-as-
vehicle contributors to the fleet average; they would be treated as just achieving the target
standard. If, on the other hand, the vehicle is a credit-user, the appropriate number of additional
credits would be needed to offset the engine-as-vehicle contributors. The purchaser of the
c See 40 CFR 1037.150(m) and 49 CFR 535.5(a)(7).
D Equivalent test weight is defined at 40 CFR 1037.104(d)(l 1) and is determined based on a vehicle's adjusted
loaded vehicle weight as specified in 40 CFR 86.129, except that for vehicles over 14,000 pounds, this may be
rounded to the nearest 500 pound increment.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
engine would treat it as any other certified engine, and would still need to meet applicable
vocational vehicle standards for the vehicles in which the engine is installed.
In deriving the stringency of the Phase 2 SI engine standard, the agencies first reviewed
the technology that was presumed in the MY 2010 Phase 1 baseline and the technology that was
projected to be adopted to meet the MY 2016 SI engine standard, finalized as part of the Phase 1
program. Engines certified to this standard would represent a logical level at which to set a
Phase 2 baseline performance level.
The agencies finalized MY 2016 standards that require manufacturers to achieve a 5
percent reduction in CO2 compared to the Phase 1 MY 2010 baseline. That MY 2010 baseline
engine was described in the Phase 1 Preamble at Section III.B.2.a.iii, as a naturally aspirated,
overhead valve V8 engine.120
In deriving the stringency of the MY 2016 gasoline engine standards, the agencies
projected 100 percent adoption of engine friction reduction, coupled cam phasing, and
stoichiometric gasoline direct injection (SGDI) to produce an overall 5 percent reduction from
the reference engine, over the engine FTP test cycle. Table 2-6 presents the technologies
projected to be present on an engine following this technology path.
Table 2-6 MY 2016 Technology Projection for SI Engines
TECHNOLOGY
ADOPTION
RATE
Coupled Cam Phasing
100%
Engine friction reduction
100%
SGDI
100%
In deciding whether to consider the above package as representing the Phase 2 baseline
performance of SI engines, the agencies reviewed available certification information and
consulted with stakeholders to determine the degree to which these projections match with
engines being produced today and engine product plans during the Phase 1 time frame. The
agencies have learned that no SI engine manufacturer has applied SGDI to this type of engine to
date, though cam phasing and engine friction reduction are widely being employed.
Section II.D.2(b) and Section V.C.l(b) of the Preamble discuss the agencies' response to
comments received on the application of SI engine technologies in the Phase 2 SI engine
standard and the vocational vehicle program, respectively. None of the comments received by
the agencies provided technical data on engine technology performance over the HD gasoline
engine FTP test procedure. Further, many engine technologies suggested to the agencies are
already presumed to be applied to SI engines, at application rates of 100 percent (see Table 2-6
above), to meet the MY 2016 engine standard. Because the agencies cannot count the
performance of those Phase 1 technologies in a Phase 2 standard, the difference between what
the commenters seek and what the agencies are adopting is considerably less than initially
appears (and that the commenters appear to believe).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.7 Technology Application and Estimated Costs - CI Engines
2.7.1 Phase 1 Engine Standards
The agencies' initial premise is that the baseline CI engine for purposes of the Phase 2
engine standard must be the engine needed to meet the Phase 1 CI engine standard. Table 2-7
shows CO2 performance at the end of Phase 1. However, as explained in the next few sections,
there are some issues associated with these baselines for both tractor and vocational engines.
Consequently, the agencies adjusted these baseline values from those proposed.
Table 2-7 Baseline Phase 1 CO2 Standards (g/bhp-hr)
LHDD - FTP
MHDD - FTP
HHDD - FTP
MHDD - SET
HHDD - SET
576
576
555
487
460
2.7.2	Individual Technology Feasibility and Cost
The cost for combustion system optimization includes costs associated with several
individual technologies, specifically, improved cylinder head, turbo efficiency improvements,
EGR cooler improvements, higher pressure fuel rail, improved fuel injectors and improved
pistons. The cost estimates for each of these technologies are presented in Chapter 2.7 of this
RIA for heavy HD, medium HD and light HD engines, respectively.
The agencies have included the costs of model-based control development in the research
and development costs applied separately to each engine manufacturer.
2.7.3	Test Cycle Weighting
The current SET modes used for tractor engine certification in Phase 1 have a relatively
large weighting in C speed as shown in the middle column of the following table:

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-8 SET Modes Weighting Factors
SPEED/% LOAD
WEIGHTING FACTOR IN
WEIGHTING FACTOR IN

PHASE 1 (%)
PHASE 2 (%)
Idle
15
12
A, 100
8
9
B, 50
10
10
B, 75
10
10
A, 50
5
12
A, 75
5
12
A, 25
5
12
Cd
o
o
9
9
B, 25
10
9
C, 100
8
2
C, 25
5
1
C, 75
5
1
C, 50
5
1
Total
100
100
A:
23
45
B:
39
38
C:
23
5
It can be seen from the above table that 23 percent weighting is in C speed, which is
typically in the range of 1800 rpm for HHD engines. However, many of today's HHD engines
do not commonly operate at such a high speed in real world driving conditions, specifically
during cruise vehicle speed between 55 and 65 mph. The agencies received confidential
business information from a few vehicle manufacturers that support this observation.
Furthermore, one of the key technology trends is to down speed, moving the predominant engine
speed from the range of 1300-1400 rpm to the range of 1150-1200 rpm at a vehicle speed of 65
mph. This trend would make the predominant engine speed even further away from C speed.
Therefore, it can be argued that, if the current SET weighting factors were retained in Phase 2,
the test would not properly reflect real-world driving operations. A more detailed explanation
with supportive data on this matter can be found in the article.121 Accordingly, the agencies are
adjusting the weighting of the various modes in the SET cycle as presented in the third column
of Table 2-8.
As shown, the new SET mode weighting basically moves most of the C speed weighting
to A speed. It also slightly reduces the weighting factor on the idle speed. These values are
based on the confidential business information obtained from vehicle manufacturers.
2.7.4 Phase 2 Baseline for Tractor and Vocational Engines
As mentioned above, the Phase 2 baseline engine numerical values are changed from
those used at proposal. However, the reasons for these changes differ for tractor and vocational
engines. For the tractor engine, the reason for the change in the SET cycle baseline values is due
to the new SET weighting factors, shown in Table 2-8, even though the engine fueling map as a
function of the engine torque and speed is the same whether Phase 1 or Phase 2 SET weighting
factors are used. Since the tractor engine standards are set up based on a composite value over
the 13 modes of the SET, using the weighting factors shown in Table 2-8, the new adjusted

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
standards with the new weighting factors result in a new set of numerical values shown in Table
2-9. Compared to the values in Table 2-7, the values are about 1.1 to 1.2 percent lower because
of the new SET weighting structure.122
Table 2-9 Tractor Engine Baseline CO2 Performance (g/bhp-hr)
MHDD - SET
HHDD - SET
481
455
For the vocational engine standard, the new baselines are required because GHG
performance of vocational vehicle engines has improved significantly since the inception of the
Phase 1 standards, and therefore, the baselines reflecting the level of the Phase 1 standard are
unrepresentative. The latest 2016 federal certification data, as well as data posted on California
Air Resource Board (CARB) websites, show that many of the Phase 1 engines are not only easily
achieving the Phase 1 2017 standard, but in some instances, the proposed 2027 engine standards
as well! See Figure 2-15 and Figure 2-16.
595
575
555
535
BO
K 515
O
U
495
475
455

2017 Phase I standard (555g/hp-hr)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Engines
Figure 2-15 2016 certified HHD engines over FTP cycle

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review
M
fSI
O
u
660
640
620
600
580
560
540
520
500
480
2017
Phase I standard (576 g/hp-hr)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Engines
Figure 2-16 2016 certified MHD/LHD engines over FTP cycle
The major contributor to this achievement in the vocational engine sector is transient
control related technologies, such as thermal management. This is one of the most challenging
areas for which to project improvement due to the nature of transient behaviors and the limited
data available. These improvements were not yet reflected in the 2010 certification data
available at proposal. Specifically, an integrated SCR and DPF system, including their
hardware, composition of catalytic material, urea dosing strategy, was in an early, non-optimized
stage in 2010. The early production SCR+DPF system had not been fully optimized for thermal
management and urea dosing strategy. As a result, some of the thermal management measures,
such as tailpipe back pressure control, post-fuel injection, and intake throttle control, tended to
be less efficient during transient operation. The agencies have also learned from the recent
certification data, illustrated in the figures above, that LHD engines perform differently than
MHD engines, and therefore that it makes more sense to separate MHD and LHD engines rather
than combine them in a single standard as in Phase 1. In view of this situation, after the agencies
analyzed all available certification data, we average the best possible engines from each
manufacturer, and consequently, the baselines of 2018 vocational engines for Phase 2 are
adjusted as follows.
Table 2-10 Vocational Engine Baseline CO2 Performance (g/bhp-hr)
LHD - FTP
MHD- FTP
HHD - FTP
576
558
525
2.7.5 Technology Packages
The agencies assessed the impact of technologies over each of the SET modes to project
an overall improvement for a tractor engine. It should be pointed out that the technology
packages discussed in this section are relevant for both tractor and vocational engines, with the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
exception of WHR related technologies. The agencies considered improvements in parasitic and
friction losses through bearing and piston ring designs to reduce friction, improved lubrication
and oils, and improved water pump and oil pump designs to reduce parasitic losses. The after-
treatment improvements are available through additional improvements that lower backpressure
of the systems, further optimization of the engine-out NOx levels, and further reduction in
ammonia slip from the SCR. Improvements to the EGR system and air flow through the intake
and exhaust systems, including through turbochargers, can also produce engine efficiency
improvements. Improvements in combustion chamber design and materials and fuel injection
control can reduce the fuel consumption of the engine. Engine downsizing is part of this
consideration with respect to improving efficiency, specifically when this technology is used
together with engine down-speeding. Although one of the most effective single technologies to
improve engine efficiency is the application of waste heat recovery (WHR) via the Rankine heat
engine cycle, the agencies do not project that this technology will have significant market
adoption until MY 2024. The reason for this is that this type of WHR system is currently only at
a pre-prototype stage of development. Furthermore, the system itself includes many components
that still require extensive field testing to assure reliability. The high technology cost, longer
payback period (if the cost and benefit of using WHR is considered in isolation), concern about
commercial acceptance (given the technology complexity, cost, concern about reliability leading
to demurrage costs and warranty claims in early model years) again point to a longer necessary
lead time for introducing this technology. See Chapter 2.3.9 above for more detailed discussions
on WHR. The agencies received detailed information from various stakeholders, who provided
information that was claimed as confidential business information (CBI). Examples include
technology improvement effectiveness information at each or some of 13 SET modes,
information on the list of components in the system, the working fluid of the system, and the
overall design.
While many effective technologies are considered for this rulemaking, it is important to
point out that the benefits of these technologies are not additive. For example, when multiple
technologies are applied to an engine, it is incorrect to simply sum the individual technologies'
effectiveness to arrive at an overall combined effectiveness of the technologies. We have
received a number of public comments regarding this non-additive effect. Most of them focus on
the agencies' projections of our so-called "dis-synergy" effect and our use of a dis-synergy factor
to account for this effect. This effect could also be called a negative synergy because it is a
decrease in technology effectiveness as a result of multiple technologies being applied to an
engine. Some commenters recommended that we adopt lower numeric values of our dis-synergy
factors, but a few commenters recommended higher dis-synergy factors than what we proposed.
A number of NGOs maintained that it was inappropriate to have a single dis-synergy factor. The
following paragraphs provide some background on this effect and our rationale for how we
developed numeric dis-synergy factors and applied them within our final stringency analysis.
As background, it is helpful to first review how engine fuel efficiency technologies
interact with one another. One example is the interaction between WHR and other technologies,
such as combustion, friction reduction, and fuel injection system improvements. WHR
effectiveness is directly proportional to the amount of thermodynamic available energy (i.e.,
energy available for conversion into mechanical work) provided from an engine's sources of
waste heat. In a modern internal combustion engine, these sources include exhaust gas energy
available from the EGR cooler and tailpipe, and from the coolant and lubricating oil systems.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Therefore, decreasing the amount of available energy from these sources reduces the
effectiveness of WHR. Some of the fuel efficiency technologies we identify in our stringency
analysis decrease the amount of available energy from these sources. For example, advancing
fuel injection timing will improve efficiency to a certain point, but it will also decrease available
exhaust energy by lowering exhaust temperature, and thus exhaust WHR effectiveness would
decrease. To a lesser extent, reducing bearing friction or piston ring-wall friction improves fuel
efficiency, but this also leads to less heat transfer to the coolant; and hence lower available
energy for WHR. As another example, increasing compression ratio can improve combustion
thermal efficiency (until the peak cylinder pressure rises past a given mechanical limit), but this
in turn increases friction losses at piston rings and bearings. As another example increasing fuel
injection pressure provides more opportunity for fuel injection optimization (e.g., enabling more
multiple injection events), which can improve fuel efficiency, but this will in turn increase fuel
pump parasitic energy losses. In another example, increasing turbocharger efficiency can
improve fuel efficiency, but this will also reduce EGR flow due to lower back pressure, thus
potentially increasing NOx, and also reducing the exhaust gas energy that can be utilized by
waste heat recovery devices, such as turbo-compound and Rankine cycle systems. Increasing
NOx would also put more demand on the after-treatment system or force less fuel efficient fuel
injection timing. There are more examples, but in conclusion, there are numerous complex
interactions between fuel efficiency technologies. In the next few paragraphs we describe how
we accounted for those interactions that lead to a dis-synergy effect.
If the agencies possessed the resources to conduct a multi-million dollar multi-year effort
to very accurately quantify all of the potential engine technology fuel efficiency dis-synergies,
we would have embarked on the development and calibration of a comprehensive engine cycle
computer simulation model several years ago. Such an effort would lead to the development of
an engine cycle simulation model, which would consist of all engine components, including sub-
models for fuel injection systems and combustion chambers; piston ring and bearing friction and
heat transfer; intake and exhaust systems, including EGR system, turbochargers, after-treatment
devices; and Rankine cycle or other WHR systems. Calibrating and validating such a model
would require tremendous laboratory testing resources to conduct the requisite component-level
and engine-level testing to gain confidence in the prediction capability of such a model. The
most challenging, and perhaps somewhat impossible, part of this comprehensive approach would
be to complete some sort of experimental validation step to demonstrate that the model
accurately predicts the combined performance of engine technologies that do not yet exist.
This level of effort is beyond the scope of the agencies' resources. However, fortunately,
other research and development programs have sufficiently reported on the magnitude of these
dis-synergies to the point that reliable estimates may be projected. The agencies were able to
rely upon information made available through research programs like DOE's SuperTruck
Program, where a number of major engine manufacturers partnered with DOE to co-fund
advanced high-efficiency engine development.23'25'26'30 In each of the manufacturer's
SuperTruck programs, more than five years and greater than ten million dollar budgets were
spent to model and develop pre-prototype engines. The agencies initially asked manufacturers if
they would share their proprietary SuperTruck engine cycle simulation models with the agencies.
This request was understandably declined because such models contain manufacturers' most
advanced and valuable competitive information. Therefore, based on the best information

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
available, the agencies developed a single set of empirical constants to account for these known
dis-synergies, and we applied these constants within our stringency analyses.
In this empirical approach, all technologies under consideration are combined according
to the National Academies recommended formula for combining the fuel efficiency benefits of
multiple engine (and vehicle) technologies:1
Equation 2-1: Formula for Combining Fuel Efficiency Benefits
%FEM=\-Yl(\-fr%FE:I)
i
In this equation,/; represents the market penetration of technology and %FE\ is the
percent fuel efficiency improvement (i.e., effectiveness) associated with technology i. The
resulting %FEtota\ is the combined fuel efficiency improvement due to all technologies, but with
no accounting of technology dis-synergies, like those described above. To account for dis-
synergies, %FEto\a\ is multiplied by a single numerical constant, which we call a dis-synergy
factor. This dis-synergy factor has two extreme bounds: a lower bound of 0.0 and an upper
bound of 1.0. And practically speaking, it is highly unlikely that adding a technology to an
engine that leads toward a dis-synergy factor on the order of 0.5 would even be considered a fuel
efficiency improving technology. Therefore, the agencies focused on determining where within
the range of 0.5-1.0 we should project this dis-synergy factor to be.
There are two key steps in determining an overall dis-synergy factor. The first step is to
determine the effectiveness of each key technology. For this step we relied upon our collection
of technology information from DOE's SuperTruck Program, from individual manufacturers and
technology suppliers, and from peer reviewed journal articles and presentations at technical
conferences. This information includes performance data on individual components and data on
engines with different combinations of technology. The second step is to iteratively solve for the
most probable single dis-synergy factor that matches the diverse set of data that we collected.
This step started by first running a simplified engine cycle simulation model (GT Power) to
simulate individual technology benefits, and then we ran the model with different technology
combinations. Finally, the results of the simplified model were compared to the data we had
collected. Note that while we were not able to validate this model to be accurate in an absolute
sense, the relative trends output by the model were consistent with the data we have in-hand.
With this model we determined a range of dis-synergy factors and the value of the factor
depended in part on the selection of technology packages. We found that this constant varies in
the range of 0.75 - 0.90. This range is further supported by separate, independent studies
performed by SwRI that were sponsored by SwRI report.7 Based upon our conclusion of this
range, the agencies are not going to adopt a dis-synergy factor of 0.95, which was requested of us
in comment. Based on our modeling and corroborative data, 0.95 would be inappropriately high
and likely not achievable.
Table 2-11 lists the potential emission reduction technologies together with the agencies'
estimated market penetration for tractor engines, along with the dis-synergy factors developed by
the agencies. A dis-synergy factor of 0.85 is adopted for 2021, and 0.90 is used for 2024 and
2027. This increase in the value of the dis-synergy factor represents the results of manufacturers
increasing their research and development efforts to optimize engine technologies together as a

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
package, in order to comply with the HD Phase 2 engine standards. The agencies have
accounted for our projected increased investment in research and design by including respective
incremental vehicle cost increases in our cost analysis. By increasing the dis-synergy factor
from 0.85 to 0.90 in MY 2024, our MY 2024 and MY 2027 engine standards are based on our
projections of increased technology package optimization. For example, we project that the
friction increase associated with the use of higher compression ratios leading to higher peak
cylinder pressures will be compensated for by friction reduction via improvements in piston ring
and crankshaft bearing design, as well as by improved oil lubricants. It should be noted that
Table 2-11 does not include individual modes of technology improvement over the 13 individual
modes of the SET. This is a result of the fact that we aggregated CBI data obtained from
manufacturers in order to avoid releasing proprietary intellectual property within this
presentation of our analysis.
Table 2-11 Projected Tractor Engine Technologies and Reduction, Percent Improvements Beyond Phase 1,
2017 Engine as Baseline
SET MODE
SET
WEIGHTED
REDUCTION
(%) 2020-2027
MARKET
PENETRATION
(2021)
MARKET
PENETRATION
(2024)
MARKET
PENETRATION
(2027)
Turbo compound with clutch
1.8%
5%
10%
10%
WHR (Rankine cycle)
3.6%
1%
5%
25%
Parasitic/Friction (Cyl Kits,
pumps, FIE), lubrication
1.4%
45%
95%
100%
Aftertreatment (lower dP)
0.6%
30%
95%
100%
EGR/Intake & exhaust
manifolds/Turbo /VVT/Ports
1.1%
45%
95%
100%
Combustion/FI/Control
1.1%
45%
95%
100%
Downsizing
0.3%
10%
20%
30%
Weighted reduction (%)

1.8%
4.0%
4.8%
Down speed impact on 13
modes

0.1%
0.2%
0.3%
Total reduction

1.8%
4.2%
5.1%
The agencies used the current market information and literature values to project what
technologies would be available in the time frame beyond 2021 and what their market
penetration would be. Chapter 2.3.9 details the reasons of why many of the technology market
penetration rates would follow an S-shape curve, which is most applicable to WHR with the
Rankine cycle technology. In spite of the fact that all trucks with WHR Ranking cycle
technology were still in the R/D stage or in the pre-prototype stage, the successful
demonstrations in real world driving conditions such as the DOE-sponsored SuperTruck
program, shows the technology that could be brought into market earlier because of the
technology's effectiveness. The agencies project that WHR with Rankine cycle will gain
momentum with time because of the potential for large emission reductions. It is unlikely that
we will see large scale production of WHR in the 2021 MY because of the many challenges that
industry faces, as described in Chapter 2.3.9. The agencies expect a market penetration of 1
percent in 2021. It will take time for WHR to have a sizeable market penetration due to system
complexity and it is estimated to be 5 percent in 2024; 25 percent in 2027, which follows an S-

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
shape curve, beginning with slow initial adoption, then more rapid adoption, and then a leveling
off as the market saturates. More discussion on WHR market penetration can be seen in Chapter
2.3.9. As there discussed, this projected trend is consistent with the finding reported by
NACEF36 in terms of the S-shape curve.
As for WHR with turbo-compound technology, only Daimler uses turbo-compound in
their DDI5 and DDI6 engines. They are phasing out turbo-compounding in the future and
replacing it with asymmetric turbo technology for most applications. Volvo just announced that
it would bring its newly-developed turbo-compound technology to market in mainly tractor
applications. Combining both manufacturers' market shares, the agencies estimate a 5 percent
market share for turbo-compound technologies in 2021. Additional production from these
manufacturers or from some additional manufacturers that could adopt this technology in some
of their trucks could push the market penetration up to 10 percent after 2024.
All other technologies, with the exception of downsizing, such as parasitic/friction loss,
aftertreatment, air breathing system, and combustion, which have been on the market already for
substantial periods and are relatively mature when compared to WHR, would follow the same
path for market penetration, 45 percent in 2021, 95 percent in 2024, and 100 percent in 2027.
The agencies don't expect high market penetration of engine downsizing, because downsizing
has a trade-off with reliability and resale values. We do see the potential for this type of
technology as it can be effective when combined with down speeding, specifically when power
demand drops due to more efficient engine and vehicle platforms. However, unlike other
technologies, such as parasitic/friction, aftertreatment, and combustion, the technology of down-
speeding together with downsizing would face the issue associated with resale value. As such,
the fleet may be reluctant to accept this technology as others until the reliability is proven.
Therefore, we don't expect that the market penetration would be as high as other technologies. It
comes down to a matter of choice. We project 10 percent, 20 percent, and 30 percent market
penetration rates in 2021, 2024, and 2027 respectively.
The tractor engine technology compliance pathway shown in Table 2-11 is only one of
many paths that manufacturers might adopt in order to achieve the 1.8 percent, 4.2, and 5.1
percent reduction goals in 2021, 2024 and 2027 respectively. This particular compliance
pathway relies on some use of WHR - small initial market penetration in 2021 and 2024,
increasing to 25 percent in MY 2027.E This projected rate of penetration in MY 2027 is greater
than projected at proposal (where the agencies' compliance pathway had WHR used in tractor
engines). One of the key reasons to increase the market penetration on WHR with Rankine cycle
technology was based on the valuable and credible CBI information obtained from a meeting
with Cummins.123 It can be mentioned that, during the meeting, Cummins provided detailed
technical information on both technology effectiveness and reliability on an entire engine system
level as well as a component level, indicating that the agency's early projection with 15 percent
on WHR was conservative, and should be increased even with their current engine platform.
Considering that sleeper cab and day cab are about 50-50 percent share on the market, and also
considering that Cummins' engine Class 8 market share is in the range of 35-45 percent in the
past few years and is expected to stay in the same range, this can be translated to 17.5-22.5
percent market share in the sleeper cab segment just from one manufacturer. Although the WHR
E As will be seen in Chapter 2.8, much higher market penetration of WHR is used in the sleeper cab engine.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
technology is most likely and most effectively applied to sleeper cabs, it would not be surprising
that a very small portion of day cabs could utilize this type of technology depending on their
driving routes. If other manufacturers can put their WHR Rankine cycle system in a pilot trial
manner with just a few percent market share, it could reach 25 percent share on the market.
In addition to the technologies mentioned above, down speeding effects are also part of a
projected technology package for tractor engines, and for vocational engines that share the same
hardware as tractor engines. Down speeding is performed by systematically shifting the engine
peak torque curve to a lower speed region of the engine map and also increasing the overall peak
torque at this lower speed. This allows you to take advantage of the use of a lower vehicle axle
ratio to enable the engine to spend much of its operating time in its most efficient spot on the
map. We expect that down speeding will take place in three sequential steps in 2021, 2024, and
2027, with engine peak torque shifting to the highest torque at the lowest speed in 2027.
The changes to engine peak torque and associated power for down speed engines has a
different effect on the 13 modes of the SET when compared to a 2018 baseline engine. The
effect is varied based on the engine map characteristics, such as the location of the sweet spot
and the shape of the peak torque curve. We utilized a large number of engine fuel maps to
investigate the impact of down speeding on composite fuel consumption over the SET
certification cycle for different engine fuel map shapes. We found that the benefit varied from
no improvement to 0.6 percent while the average benefit is around 0.3 percent for the 2027
torque curve used in our analysis. Engine fuel maps that are less aggressive in peak torque
behavior, such as 2021 engine map, show less of an effect on fuel consumption reduction.
Therefore, we conclude that fuel consumption reductions due solely to the changes in the 13
mode SET speed and load are 0.1 percent, 0.2 percent, and 0.3 percent for 2021, 2024, and 2027,
respectively.
Figure 2-17, Figure 2-18, and Figure 2-19 contain the 2018 baseline engine fuel maps for
350 Hp, 455 Hp, and 600 Hp rating engines. The 350 Hp engine will be used for class 7 tractors
and some HHD vocational vehicles. The 455 Hp engine will be used for all HHD tractors with
sleeper cabs and day cabs as well as some HHD vocational vehicles. The 600 Hp engine is only
used for Heavy Haul tractors.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
2018 Baseline Engine 350hp / 11L BSFC (g / kW * hr)
1600
1400
1200
1000
 1000
o-
o
H
800
600
400
235
200
600
800
1000 1200 1400 1600 1800 2000 2200
Speed ( RPM )
Figure 2-18 2018 Baseline Engine Fuel Map used in GEM for a 455 Hp Rating

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Engine 600hp / 15L BSFC ( g / kW * hr)
Figure 2-19 2018 Baseline Engine Fuel Map used in GEM for a 600 Hp Rating
The agencies considered the same technology package developed for the HHD diesel
engines for vocational LHD diesel and MHD diesel engines. The technology package includes
parasitic and friction reduction, improved lubrication, aftertreatment improvements, EGR system
and air flow improvements, and combustion improvements. WHR technology is not part of the
package as WHR is not as efficient over transient operation, which is the principal operating
mode for vocational vehicles, even regional vehicles, since transient operation still comprises a
large portion of overall regional vehicle operation. One difference between tractor and
vocational engines is the model based control used over transient operation, which is applied to
operation over the FTP cycle. Chapter 2.3.3 details the model based control. Table 2-12 below
lists technologies and projected penetration rates which are the predicate for the standard for the
various vocational vehicle engines. The same dis-synergy factors that were generated for
tractors are also used. As is true of all the projected compliance pathways/ there are other
(usually myriad) ways to achieve the standard.
The market penetration rate and technology effectiveness estimates shown in Table 2-12
were developed using CBI data provided by engine manufacturers in conjunction with the
agencies' engineering judgment using the same principles outlined previously for tractor engines.
In terms of effectiveness, the model based control used over transient operation, which is
described in Chapter 2.3.3, would be one of the most effective technologies, but it would take
significant effort to develop and put it into production. An example of this technology is the
neural network approach developed by Daimler.19-20 One concern surrounding the use of this
technology is that it is still not clear how it will interact with on-board diagnostics (OBD). For
example, one of the purposes of the model based control is to use physical models to predict the
engine performance. As a result of that, the number of sensors in theory could be reduced, such
as one of the NOx sensors, or a few temperature sensors. On the other hand, OBD would largely
F The exception being those standards where a design is mandated, as for certain non-aero trailers.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
rely on the sensors to collect data. If one of the engine components malfunctions, and the
sensors that were in place to identify the issue were removed because of model based control,
OBD would not be able to diagnose the issue correctly. It is not clear how this issue can be
effectively resolved if some of sensors would be removed. We expect a 25 percent market
penetration in 2021, 30 percent in 2024, and 40 percent in 2027. All other technologies in Table
2-10 are relatively more mature than model based control, and therefore, higher market
penetration is projected. It should be pointed out that in developing standard stringency, the
technologies' effectiveness is applied to all the engines including Regional, Multipurpose, and
Urban vehicles, since the same engine hardware will be used for all of these applications.
Table 2-12 Projected Vocational Engine Technologies and Reduction, Percent Improvements Beyond
Baseline Engine
TECHNOLOGY
GHG
EMISSIONS
REDUCTION
2020-2027
MARKET
PENETRATION
2021
MARKET
PENETRATION
2024
MARKET
PENETRATION
2027
Model based control
2.0%
25%
30%
40%
Parasitic /Friction
1.5%
60%
90%
100%
EGR/Air/WT /Turbo
1.0%
60%
90%
100%
Improved AT
0.5%
30%
60%
100%
Combustion Optimization
1.0%
60%
90%
100%
Weighted reduction (%)-
L/M/HHD

2.3%
3.6%
4.2%
Figure 2-20 and Figure 2-21 are the 2018 baseline engine fuel maps used in GEM for the
270 Hp and 200 Hp rated engines. The 2018 baseline engines with 350 Hp and 455 Hp that are
used for vocational vehicles share the same engines as tractors, and therefore, there is no need to
display their maps here.
2018 Baseline Engine 270hp / 7L BSFC ( g / kW * hr)
900
800
700
600
500
0)
F 400
o
I—
300
235
245
200
265
265
100
800 1000 1200 1400 1600 1800 2000 2200 2400 2600
Speed ( RPM )
Figure 2-20 2018 Baseline Engine Fuel Map used in GEM for a 270 Hp Rating

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
2018 Baseline Engine 200hp / 7L BSFC ( g / kW * hr)
900
800
700
600
6 500
h 400
300
235
250
270
200
100
800 1000 1200 1400 1600 1800 2000 2200 2400
Speed ( RPM)
Figure 2-21 2018 Baseline Engine Fuel Map used in GEM for a 200 Hp Rating
2.7.6 2021 Model Year HHP Diesel Engine Package for Tractors
As can be seen in Table 2-11, the composite CO2 reduction (the product of the
technology efficiency and projected technology penetration rates shown in that table) for a MY
2021 tractor engine over the SET cycle is 1.8 percent. With this reduction, the numerical
stringency values for 2021 can be derived from the baseline engine with new Phase 2 weighting
factors. Table 2-13 below shows the 2021 model year tractor engine standards.
Table 2-13 2021 Model Year Standards - Tractors

MHDD- SET
HHDD - SET
CO2 Emissions (g CO;/bhp-hr)
473
4474
Fuel Consumption (gal/100 bhp-lir)
4.6464
4.3910
The cost estimates for the MY 2021 HHD diesel engine packages can be developed from
the same information (i.e. technologies on which standard stringency is premised and projected
penetration rates) as shown in Table 2-14. We present technology cost estimates along with
adoption rates in Chapter 2.11 of this RIA. We present package cost estimates in greater detail
in Chapter 2.12 of this RIA.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-14 Technology Costs as Applied in Expected Packages for MY2021 Tractor Diesel Engines relative
to the Flat Baseline (2013$)a

MEDIUM
HEAVY

HD
HD
Aftertreatment system (improved effectiveness SCR, dosing, DPF)
$7
$7
Valve Actuation
$84
$84
Cylinder Head (flow optimized, increased firing pressure, improved thermal
management)
$3
$3
Turbocharger (improved efficiency)
$9
$9
Turbo Compounding
$51
$51
EGR Cooler (improved efficiency)
$2
$2
Water Pump (optimized, variable vane, variable speed)
$44
$44
Oil Pump (optimized)
$2
$2
Fuel Pump (higher working pressure, increased efficiency, improved pressure
regulation)
$2
$2
Fuel Rail (higher working pressure)
$5
$5
Fuel Injector (optimized, improved multiple event control, higher working
pressure)
$5
$5
Piston (reduced friction skirt, ring and pin)
$1
$1
Valve Train (reduced friction, roller tappet)
$39
$39
Waste Heat Recovery
$71
$71
"Right sized" engine
-$41
-$41
Total
$284
$284
Note:
a Costs presented here include projected technology penetration rates presented in Table 2-11. These costs include indirect costs
via markups along with learning impacts. For a description of the markups and learning impacts considered in this analysis and
how it impacts technology costs for other years, refer to Chapter 2 of the RIA (see RIA 2.11).
2.7.7 2021 Model Year LHD/MHD/HHD Diesel Engine Package for
Vocational Vehicles
From Table 2-12, the reduction of CO2 for 2021 model years of all LHD/MHD/HHD
vocational diesel engines is 2.3 percent. Table 2-15 below shows the 2021 model year
vocational engine standards.
Table 2-15 2021 Model Year Standards — Vocational

LHDD -
FTP
MHDD -
FTP
HHDD -
FTP
CO2 Emissions (g CCh/bhp-hr)
563
545
513
Fuel Consumption (gal/100 bhp-hr)
5.5305
5.3536
5.0393
The cost estimates for the MY 2021 vocational diesel engines are shown in Table 2-16.
We present technology cost estimates along with adoption rates in Chapter 2.11 of this RIA. We
present package cost estimates in greater detail in Chapter 2.12 of this RIA and adoption rates in
Chapter 2.9.1.2.2.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-16 Technology Costs as Applied in Expected Packages for MY2021 Vocational Diesel Engines
relative to the Flat Baseline (2013$)a

LIGHT
MEDIUM
HEAVY

HD
HD
HD
Aftertreatment system (improved effectiveness SCR, dosing, DPF)
$8
$8
$8
Valve Actuation
$93
$93
$93
Cylinder Head (flow optimized, increased firing pressure, improved
thermal management)
$6
$3
$3
Turbocharger (improved efficiency)
$10
$10
$10
EGR Cooler (improved efficiency)
$2
$2
$2
Water Pump (optimized, variable vane, variable speed)
$58
$58
$58
Oil Pump (optimized)
$3
$3
$3
Fuel Pump (higher working pressure, increased efficiency, improved
pressure regulation)
$3
$3
$3
Fuel Rail (higher working pressure)
$8
$6
$6
Fuel Injector (optimized, improved multiple event control, higher
working pressure)
$8
$6
$6
Piston (reduced friction skirt, ring and pin)
$1
$1
$1
Valve Train (reduced friction, roller tappet)
$70
$52
$52
Model Based Controls
$29
$29
$29
Total
$298
$275
$275
Note:
a Costs presented here includes projected technology penetration rates presented in Table 2-12. These costs include indirect costs
via markups along with learning impacts. For a description of the markups and learning impacts considered in this analysis and
how it impacts technology costs for other years, refer to Chapter 2 of the RIA (see RIA 2.11).
2.7.8 2024 Model Year HHDD Engine Package for Tractors
The agencies assessed the impact of technologies over each of the SET modes to project
an overall improvement in the 2024 model year. The agencies considered additional
improvements in the technologies included in the 2021 model year package. Compared to the
2021 technology package, the technology package in 2024 considers higher market adoption as
shown in Table 2-11, thus deriving a reduction of 4.2 percent. Table 2-17 below shows the 2024
model year tractor engine standards.
Table 2-17 2024 Model Year Standards - Tractors

MHDD- SET
HHDD - SET
CO2 Emissions (g CCh/bhp-hr)
461
436
Fuel Consumption (gal/100 bhp-hr)
4.5285
4.2829
The cost estimates for the MY 2024 tractor diesel engines are shown in Table 2-18. We
present technology cost estimates along with adoption rates in Chapter 2.11 of this RIA. We
present package cost estimates in greater detail in Chapter 2.12 of this RIA.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-18 Technology Costs as Applied in Expected Packages for MY2024 Tractor Diesel Engines relative
to the Flat Baseline (2013$)a

MEDIUM
HEAVY

HD
HD
Aftertreatment system (improved effectiveness SCR, dosing, DPF)
$14
$14
Valve Actuation
$169
$169
Cylinder Head (flow optimized, increased firing pressure, improved thermal
management)
$6
$6
Turbocharger (improved efficiency)
$17
$17
Turbo Compounding
$93
$93
EGR Cooler (improved efficiency)
$3
$3
Water Pump (optimized, variable vane, variable speed)
$85
$85
Oil Pump (optimized)
$4
$4
Fuel Pump (higher working pressure, increased efficiency, improved pressure
regulation)
$4
$4
Fuel Rail (higher working pressure)
$9
$9
Fuel Injector (optimized, improved multiple event control, higher working
pressure)
$10
$10
Piston (reduced friction skirt, ring and pin)
$3
$3
Valve Train (reduced friction, roller tappet)
$77
$77
Waste Heat Recovery
$298
$298
"Right sized" engine
-$82
-$82
Total
$712
$712
Note:
a Costs presented here reflect projected technology penetration rates presented in Table 2-11. These costs include indirect costs
via markups along with learning impacts. For a description of the markups and learning impacts considered in this analysis and
how it impacts technology costs for other years, refer to Chapter 2 of the RIA (see RIA 2.11).
2.7.9 2024 Model Year LHD/MHD/HHD Diesel Engine Package for Vocational
Vehicles
The agencies developed the 2024 model year LHD/MHD/HHD vocational diesel engine
package based on additional improvements in the technologies included in the 2021 model year
package as shown in Table 2-12. The projected impact of these technologies provides an overall
reduction of 3.6 percent over the 2018 model year baseline. Table 2-19 below shows the 2024
model year vocational engine standards.
Table 2-19 2024 Model Year Standards - Vocational

LHDD -
FTP
MHDD -
FTP
HHDD -
FTP
CO2 Emissions (g CCh/bhp-hr)
555
538
506
Fuel Consumption (gal/100 bhp-hr)
5.4519
5.2849
4.9705
Costs for the MY 2024 vocational diesel engines are shown in Table 2-20. We present
technology cost estimates along with adoption rates in Chapter 2.11 of this RIA. We present
package cost estimates in greater detail in Chapter 2.12 of this RIA.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-20 Technology Costs as Applied in Expected Packages for MY2024 Vocational Diesel Engines
relative to the Flat Baseline (2013$)a

LIGHT
MEDIUM
HEAVY

HD
HD
HD
Aftertreatment system (improved effectiveness SCR, dosing, DPF)
$14
$14
$14
Valve Actuation
$160
$160
$160
Cylinder Head (flow optimized, increased firing pressure, improved
thermal management)
$10
$6
$6
Turbocharger (improved efficiency)
$16
$16
$16
EGR Cooler (improved efficiency)
$3
$3
$3
Water Pump (optimized, variable vane, variable speed)
$81
$81
$81
Oil Pump (optimized)
$4
$4
$4
Fuel Pump (higher working pressure, increased efficiency, improved
pressure regulation)
$4
$4
$4
Fuel Rail (higher working pressure)
$11
$9
$9
Fuel Injector (optimized, improved multiple event control, higher
working pressure)
$13
$10
$10
Piston (reduced friction skirt, ring and pin)
$2
$2
$2
Valve Train (reduced friction, roller tappet)
$97
$73
$73
Model Based Controls
$32
$32
$32
Total
$446
$413
$413
Note:
a Costs presented here include project technology penetration rates presented in Table 2-12. These costs include indirect costs via
markups along with learning impacts. For a description of the markups and learning impacts considered in this analysis and how
it impacts technology costs for other years, refer to Chapter 2 of the RIA (see RIA 2.11).
2.7.10 2027 Model Year HHDD Engine Package for Tractor
The agencies assessed the impact of technologies over the SET composite test cycle to
project an overall improvement in the 2027 model year. The agencies considered additional
improvements in the technologies included in the 2024 model year package. Compared to 2021
technology package, the technology package in 2027 considers higher market adoption, thus
deriving emission reductions of 5.1 percent as shown in Table 2-11. Table 2-21 below shows the
2027 model year tractor engine standards.
Table 2-21 2027 Model Year Standards - Tractors

MHDD- SET
HHDD - SET
CO2 Emissions (g CCh/bhp-hr)
457
432
Fuel Consumption (gal/100 bhp-hr)
4.4892
4.2436
The costs for the MY 2027 tractor diesel engines are shown in Table 2-22. We present
technology cost estimates along with adoption rates in Chapter 2.12 of this RIA. We present
package cost estimates in greater detail in Chapter 2.13 of this RIA.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-22 Technology Costs as Applied in Expected Packages for MY2027 Tractor Diesel Engines relative
to the Flat Baseline (2013$)a

MEDIUM
HEAVY

HD
HD
Aftertreatment system (improved effectiveness SCR, dosing, DPF)
$15
$15
Valve Actuation
$172
$172
Cylinder Head (flow optimized, increased firing pressure, improved thermal
management)
$6
$6
Turbocharger (improved efficiency)
$17
$17
Turbo Compounding
$89
$89
EGR Cooler (improved efficiency)
$3
$3
Water Pump (optimized, variable vane, variable speed)
$85
$85
Oil Pump (optimized)
$4
$4
Fuel Pump (higher working pressure, increased efficiency, improved pressure
regulation)
$4
$4
Fuel Rail (higher working pressure)
$9
$9
Fuel Injector (optimized, improved multiple event control, higher working
pressure)
$10
$10
Piston (reduced friction skirt, ring and pin)
$3
$3
Valve Train (reduced friction, roller tappet)
$77
$77
Waste Heat Recovery
$1,208
$1,208
"Right sized" engine
-$123
-$123
Total
$1,579
$1,579
Note:
a Costs presented here include projected technology penetration rates presented in Table 2-11. These costs include indirect costs
via markups along with learning impacts. For a description of the markups and learning impacts considered in this analysis and
how it impacts technology costs for other years, refer to Chapter 2 of the RIA (see RIA 2.11).
2.7.11 2027 Model Year LHD/MHD/HHD Diesel Engine Package for Vocational
Vehicles
The agencies developed the 2027 model year LHD/MHD/HHD vocational diesel engine
package based on additional improvements in the technologies included in the 2021 model year
package as shown in Table 2-12. The projected impact of these technologies provides an overall
emission reduction of 4.2 percent over the 2017 model year baseline. Table 2-23 below shows
the 2027 model year standards.
Table 2-23 2027 Model Year Standards - Vocational

LHDD - FTP
MHDD- FTP
HHDD - FTP
CO2 Emissions (g CCh/bhp-hr)
552
535
503
Fuel Consumption (gal/100 bhp-hr)
5.4224
5.2554
4.9411
Costs for MY 2027 vocational diesel engines are shown in Table 2-24. We present
individual technology cost estimates in Chapter 2.11 of this RIA and adoption rates for
vocational vehicle engines in Chapter 2.9.1 of this RIA. We present package cost estimates in
greater detail in Chapter 2.12 of this RIA.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-24 Technology Costs as Applied in Expected Packages for MY2027 Vocational Diesel Engines
relative to the Flat Baseline (2013$)a

LIGHT
MEDIUM
HEAVY

HD
HD
HD
Aftertreatment system (improved effectiveness SCR, dosing, DPF)
$15
$15
$15
Valve Actuation
$172
$172
$172
Cylinder Head (flow optimized, increased firing pressure, improved
thermal management)
$10
$6
$6
Turbocharger (improved efficiency)
$17
$17
$17
EGR Cooler (improved efficiency)
$3
$3
$3
Water Pump (optimized, variable vane, variable speed)
$85
$85
$85
Oil Pump (optimized)
$4
$4
$4
Fuel Pump (higher working pressure, increased efficiency, improved
pressure regulation)
$4
$4
$4
Fuel Rail (higher working pressure)
$11
$9
$9
Fuel Injector (optimized, improved multiple event control, higher
working pressure)
$14
$10
$10
Piston (reduced friction skirt, ring and pin)
$3
$3
$3
Valve Train (reduced friction, roller tappet)
$102
$77
$77
Model Based Controls
$41
$41
$41
Total
$481
$446
$446
Note:
a Costs presented here include projected technology penetration rates presented in Table 2-12. These costs include indirect costs
via markups along with learning impacts. For a description of the markups and learning impacts considered in this analysis and
how it impacts technology costs for other years, refer to Chapter 2 of the RIA (see RIA 2.11).
2.8 Technology Application and Estimated Costs - Tractors
2.8.1 Defining the Baseline Tractors
The fuel efficiency and CO2 emissions of combination tractors vary depending on the
configuration of the tractor. Many aspects of the tractor impact its performance, including the
engine fuel map (independent of improvements measured under the engine standard), the
transmission, drive axle, aerodynamics, and rolling resistance. For each tractor subcategory, the
agencies selected a theoretical tractor to represent the average 2017 model year tractor that meets
the Phase 1 standards (see 76 FR 57212, September 15, 2011). These tractors are used as
baselines from which to evaluate costs and effectiveness of additional technologies and
standards. The specific attributes of each tractor subcategory baseline are listed below in Table
2-25. Using these values, the agencies assessed the CO2 emissions and fuel consumption
performance of the baseline tractors using the final version of Phase 2 GEM. The results of these
simulations are shown below in Table 2-26.
The Phase 1 2017 model year tractor standards and the baseline 2017 model year tractor
results are not directly comparable. The same set of aerodynamic and tire rolling resistance
technologies were used in both setting the Phase 1 standards and determining the baseline of the
Phase 2 tractors. However, there are several aspects that differ. First, a new version of GEM
was developed and validated to provide additional capabilities, including more refined modeling
of transmissions and engines. Second, the determination of the HD Phase 2 CdA value takes into

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
account a revised test procedure, a new standard reference trailer, and wind averaged drag.
Additionally, the HD Phase 2 version of GEM includes road grade in the 55 mph and 65 mph
highway cycles, as discussed in Preamble Section III.E.
The agencies used the same adoption rates of tire rolling resistance for the Phase 2
baseline as we used to set the Phase 1 2017 MY standards. See 76 FR 57211. The tire rolling
resistance level assumed to meet the 2017 MY Phase 1 standard high roof sleeper cab is
considered to be a weighted average of 10 percent pre-Phase 1 baseline rolling resistance, 70
percent Level 1, and 20 percent Level 2. The tire rolling resistance to meet the 2017MY Phase 1
standards for the high roof day cab, low roof sleeper cab, and mid roof sleeper cab includes 30
percent pre-Phase 1 baseline level, 60 percent Level 1 and 10 percent Level 2. Finally, the low
and mid roof day cab 2017 MY standards were premised on a weighted average rolling
resistance consisting of 40 percent baseline, 50 percent Level 1, and 10 percent Level 2. The
agencies did not receive comments on the tire packages in the NPRM used to develop the Phase
2 baseline.
The Phase 2 baseline in the NPRM was determined based on the aerodynamic bin
adoption rates used to determine the Phase 1 MY 2017 tractor standards. The vehicles that were
tested prior to the NPRM were used to develop the proposed aerodynamic bin structure for Phase
2. In both the NPRM and this final rulemaking, we developed the Phase 2 bins such that there is
an alignment between the Phase 1 and Phase 2 aerodynamic bins after taking into consideration
the changes in aerodynamic test procedures and reference trailers required in Phase 2. The Phase
2 bins were developed so that tractors that performed as a Bin III in Phase 1 would also perform
as Bin III tractors in Phase 2. The baseline aerodynamic value for the Phase 2 final rulemaking
was determined in the same manner as the NPRM, using the adoption rates of the bins used to
determine the Phase 1 standards, but reflect the final Phase 2 bin CdA values.
The agencies determined the rear axle ratio and final drive ratio in the 2017 MY baseline
tractor based on axle market information shared by Meritor,124 one of the primary suppliers of
heavy-duty axles and confidential business information provided by Daimler. Our assessment of
this information found that a rear axle ratio of 3.70 and a top gear ratio of 0.73 (equivalent to a
final drive ratio of 2.70) is a commonly spec'd tractor. Meritor's white paper on downspeeding
stated that final drive ratios of less than 2.64 are considered to be "downsped."125 The agencies
recognize that there is a significant range in final drive ratios that will be utilized by tractors built
in 2017 MY, we do not believe that the average (i.e., baseline) tractor in 2017 MY will
downsped.
In the proposal, the agencies noted that the manufacturers were not using tamper-proof
automatic engine shutdown systems (AESS) to comply with the Phase 1 standards. As a result
the agencies reverted back to the baseline auxiliary power unit (APU) adoption rate of 30 percent
used in the Phase 1 baseline. In response to comments, the agencies reassessed the baseline idle
reduction adoption rates. The latest NACFE confidence report found that 9 percent of tractors
had auxiliary power units and 96 percent of vehicles are equipped with adjustable automatic
engine shutdown systems.126 Therefore, the agencies are projecting that 9 percent of sleeper
cabs will contain an adjustable AESS and APU, while the other 87 percent will only have an
adjustable AESS.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-25 GEM Inputs for the 2017 Baseline Class 7 and 8 Tractors
CLASS 7
CLASS 8
Day Cab
Day Cab
Sleeper Cab
Low
Mid Roof
High Roof
Low Roof
Mid Roof
High Roof
Low Roof
Mid Roof
High
Roof







Roof
Engine
2017 MY
2017 MY
2017 MY
2017 MY
2017 MY
2017 MY
2017 MY
2017 MY
2017
11L
11L
11L
15L
15L
15L
15L
15L Engine
MY
Engine
Engine
Engine
Engine
Engine
Engine
Engine
455 HP
15L
350 HP
350 HP
350 HP
455 HP
455 HP
455 HP
455 HP

Engine








455 HP
Aerodynamics (CdA in m2)
5.41
6.48
6.38


6.38


5.90



5.41
6.48

5.41
6.48

Steer Tires (CRR in kg/metric ton)
6.99
6.99
6.87
6.99
6.99
6.87
6.87
6.87
6.54
Drive Tires (CRR in kg/metric ton)
7.38
7.38
7.26
7.38
7.38
7.26
7.26
7.26
6.92
Extended Idle Reduction - Adjustable AESS with no Idle Red Tech Adoption Rate @ 1% Effectiveness
N/A
N/A
N/A
N/A
N/A
N/A
87%
87%
87%
Extended Idle Reduction - Adjustable AESS with Diesel APU Adoption Rate @
3% Effectiveness
N/A
N/A
N/A
N/A
N/A
N/A
9%
9%
9%


Transmission = 10 Speed Manual Transmission



Gear Ratios =
12.8, 9.25, 6.76, 4.90, 3.58, 2.61, 1.89, 1.38, 1.00, 0.73

Drive Axle Configuration = 4x2
Drive Axle Configuration = 6x4



Tire Revs/Mile =
= 512






Drive Axle Ratio
= 3.70



Table 2-26 Class 7 and 8 Tractor 2017 Baseline CO2 Emissions and Fuel Consumption

CLASS 7
CLASS 8

Day Cab
Day Cab
Sleeper Cab

Low
Roof
Mid
Roof
High
Roof
Low
Roof
Mid
Roof
High
Roof
Low
Roof
Mid
Roof
High
Roof
CO 2 (grams
CCh/ton-mile)
119.1
127.2
129.7
91.3
96.6
98.2
84.0
90.2
87.8
Fuel
Consumption
(gal/1,000 ton-
mile)
11.699
41
12.4950
9
12.7406
7
8.9685
7
9.4891
9
9.64637
8.25147
8.86051
8.62475
The 2017 model year baseline fuel maps in the HD Phase 2 version of GEM are different
than those used in 2017 year fuel maps in the HD Phase 1 version. The baseline map in the HD
Phase 2 version takes two major factors into consideration. The first is the likelihood of engine
down speeding beyond the 2020 model year and the second is making the gradient of brake

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
specific fuel consumption rate (BSFC) around the fuel consumption sweet spot less radical when
compared to the HD Phase 1 version's engine fuel map. All the baseline engine fuel maps for
use in 2017 can be seen in Chapter 2.7.5. All other maps from 2021 to 2027 can be seen in
Chapter 2.8.4.1.
The agencies received comments regarding the heavy-haul baseline vehicle with respect
to the transmission and axle ratio. Upon consideration of these comments, the agencies find that
the baseline heavy-haul tractor is better represented by an 18-speed transmission with a 3.73 rear
axle ratio. The heavy-haul tractor baseline configuration inputs to GEM for the Phase 2 final
rule are shown below in Table 2-27. The baseline 2017 MY heavy-haul tractor will emit 56.9
grams of CO2 per ton-mile and consume 5.59 gallons of fuel per 1,000 ton-mile.
Table 2-27 Heavy-Haul Tractor Baseline Configuration
BASELINE HEAVY-HAUL TRACTOR CONFIGURATION
Engine = 2017 MY 15L Engine with 600 HP
	Aerodynamics (CdA in m2) = 5.00	
Steer Tires (CRR in kg/metric ton) = 7.0
Drive Tires (CRR in kg/metric ton) = 7.4
Transmission = 18 speed Manual Transmission
Gear ratio= 14.4, 12.29, 8.51, 7.26, 6.05, 5.16, 4.38, 3.74, 3.2, 2.73, 2.28, 1.94,
	1.62, 1.38, 1.17, 1.00,0.86,0.73	
Drive axle Ratio = 3.73
All Technology Improvement Factors = 0%
2.8.2 Defining the Tractor Technology Packages
The agencies' assessment of the technology effectiveness was developed through the use
of GEM in coordination with modeling conducted by Southwest Research Institute. The
agencies developed the standards through a three-step process, similar to the approach used in
Phase 1. First, the agencies developed technology performance characteristics or effectiveness
for each technology, as described below. Each technology is associated with an input parameter
which in turn is used as an input to the Phase 2 GEM simulation tool (i.e. the final version of
GEM used both to develop standard stringency and to evaluate compliance at certification) and
its effectiveness thereby modeled. Second, the agencies combined the technology performance
levels with a projected technology adoption rate to determine the GEM inputs used (in step 3) to
set the stringency of the final standards. Third, the agencies input these parameters into Phase 2
GEM and used the output to determine the final CO2 emissions and fuel consumption levels. All
percentage improvements noted below are over the 2017 baseline tractor.
2.8.2.1 Engine
Please see RIA Chapter 2.7 for a discussion on engine technologies.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.8.2.2 Aerodynamics
The aerodynamic packages are categorized as Bin I, Bin II, Bin III, Bin IV, Bin V, Bin
VI, or Bin VII based on the wind averaged drag aerodynamic performance determined through
testing conducted by the manufacturer. In general, the CdA values for each package and tractor
subcategory were developed through EPA's coastdown testing of tractor-trailer combinations,
the 2010 NAS report, and SAE papers.127'128 The agencies also discuss aerodynamic
technologies for tractors in Chapter 2.4.1 of the RIA.
As noted in Section III.D of the Preamble, the agencies received comments from
manufacturers about the feasibility of developing tractors with aerodynamics that could achieve
the proposed Bins V and above. After the proposal, the agencies reviewed new information
regarding the aerodynamic improvements achieved in the SuperTruck program for high roof
sleeper cabs and box trailers. Also after the proposal, the truck manufacturers conducted CFD
analysis of a "typical Bin III" high roof sleeper cab tractor with a Phase 2 standard trailer (with a
trailer skirt), a SuperTruck tractor with a Phase 2 standard trailer, and a SuperTruck tractor with
a SuperTruck trailer. Even though the agencies did not conduct the CFD testing, we agree with
the methodology and the results. As shown in Figure 2-22, the difference between a Bin III high
roof sleeper cab tractor and a SuperTruck tractor, both with a Phase 2 standard trailer, is
approximately 1.0 m2 As shown in Table 2-28, the CdA difference between Bin III and Bin IV
is approximately 0.5 m2 and the difference between Bin III and Bin V is approximately 1.0 m2
Therefore, a SuperTruck tractor would be able to achieve a Bin V level in Phase 2 with the Phase
2 standard trailer.
Table 2-28 Phase 2 Aerodynamic Bin Values for High Roof Sleeper
PHASE 2 AERO BINS FOR HIGH
ROOF SLEEPER CABS
Phase 2 Bin
CdA Range (m2)
Bin III
5.7-6.2
Bin IV
5.2-5.6
Bin V
4.7-5.1

-------
*** E. O. 12866 Review — Revised - Do Not Cite, Quote, or Release During Review ***
Aero Stringency
ifcioecy Magfrtude [m-'wc]

BIN V (4.7
-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review **"*
The effectiveness of aerodynamic improvements depends on the drive cycle. As shown
below in Figure 2-23, aerodynamics on sleeper cabs that operate a higher fraction of their miles
at highway speeds have a greater impact on fuel consumption and CO2 emissions.
Aerodynamic Impact on Tractor Fuel Consumption
Sleeper Cab
Figure 2-23 Aerodynamic Impact on Tractor CO2 Emissions based on Phase 2 GEM Simulations
2.8.2.3 Tire Rolling Resistance
The rolling resistance coefficient target for the Phase 2 NPRM was developed from
SmartWay's tire testing to develop the SmartWay certification and testing a selection of tractor
tires as part of the Phase 1 and Phase 2 programs. Even though the coefficient of tire rolling
resistance comes in a range of values, to analyze this range, the tire performance was evaluated
at four levels for both steer and drive tires, as determined by the agencies. The four levels in the
Phase 2 proposal included the baseline (average) from 2010, Level I and Level 2 from Phase 1,
and Level 3 that achieves an additional 25 percent improvement over Level 2. The Level 1
rolling resistance performance represents the threshold used to develop SmartWay designated
tires for long haul tractors. The Level 2 threshold represents an incremental step for
improvements beyond today's SmartWay level and represents the best in class rolling resistance
of the tires we tested for Phase 1. The Level 3 values in the NPRM represented the long-term
rolling resistance value that the agencies predicts could be achieved in the 2025 timeframe.
Given the multiple year phase-in of the standards, the agencies expect that tire manufacturers
will continue to respond to demand for more efficient tires and will offer increasing numbers of
tire models with rolling resistance values significantly better than today's typical low rolling
resistance tires.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
ICCT found in their workshop that opportunities exist for improvements in rolling
resistance for tractor tires that could lead to a two to six percent improvement in fuel
consumption when compared to a 2010 baseline tractor.130 A fuel consumption improvement in
this range would require a six to 18 percent improvement in the tractor tire rolling resistance
levels. Michelin commented that the proposed values for the drive tires seem reasonable, though
the 4.5 kg/ton level would require significantly higher adoption rate of new generation wide base
single tires. Michelin also stated that the value of 4.3 kg/ton target for steer tires is highly
unlikely based on current evolution and that research shows that 5.0 kg/ton would be more
likely.
The agencies have evaluated this comment and find it persuasive. The agencies analyzed
the 2014MY certification data for tractors between the NPRM and final rulemaking. We found
that the lowest rolling resistance value submitted for 2014 MY GHG and fuel efficiency
certification for tractors was 4.9 and 5.1 kg/metric ton for the steer and drive tires respectively,
while the highest rolling resistance tire had a CRR of 9.8 kg/metric ton.131 We have accordingly
increased the coefficient of rolling resistance for Level 3 tires in the final rule based on the
comments and the certification data.
Figure 2-24 shows the impact of changing the rolling resistance on CO2 emissions and
fuel consumption of tractors.
Rolling Resistance Impact on Tractor Fuel Consumption
4.0%
Sleeper Cab
u
-3.0%
Tractor-Trailer Weighted CRR (kg/metric ton)
Figure 2-24 Impact of the Coefficient of Rolling Resistance (CRR) on Fuel Consumption based on Phase 2
GEM Simulations

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.8.2.4	Tire Pressure Monitoring and Automatic Tire Inflation Systems
As noted in RIA Chapter 2.4.3.3, automatic tire inflation systems (ATIS) provide fuel
consumption improvement opportunities because they keep the tires at the proper inflation
pressure. Tire pressure monitoring systems (TPMS) notify the operator of tire pressure, but
require the operator to manually inflate the tires to the optimum pressure. The agencies did not
propose to include TPMS as a GEM input because of this dependence on the operator. Instead,
we requested comment and sought data to support a reduction level. Many commenters
suggested that the agencies should recognize TPMS in GEM and provided some additional
studies.
After consideration of the comments, the agencies are adopting provisions in Phase 2 to
allow GEM inputs for either ATIS or TPMS. The agencies believe there is sufficient incentive
for truck operators to address low tire pressure conditions if they are notified that the condition
exists by TPMS.
The agencies considered the comments and the studies to determine the effectiveness of
TPMS and ATIS. ICCT found in their workshop that opportunities exist for ATIS that could
lead to a 0.5 to two percent improvement in fuel consumption.132 The agencies conducted a
further review of the FCMSA study cited by commenters and we interpret the results of the study
to indicate that overall a combination of TPMS and ATIS in the field achieved 1.4 percent
reduction. However, it did not separate the results from each technology, therefore it did not
indicate that TPMS and ATIS achieved the same levels of reduction. Therefore, we set the
effectiveness of TPMS slightly lower than ATIS to reflect that operators will be required to take
some action to insure that the proper inflation pressure is maintained. The input values to the
Phase 2 GEM are set to 1.2 percent reduction in CO2 emissions and fuel consumption for ATIS
and 1.0 percent reduction for TPMS.
2.8.2.5	Idle Reduction
The benefits for the extended idle reductions were developed from literature, SmartWay
work, and the 2010 NAS report. Additional details regarding the comments and calculations are
included in RIA Chapter 2.4.8.1.
2.8.2.6	Transmission
The benefits for automated manual (AMT) and automatic (AT) transmissions were
developed from literature, from simulation modeling conducted by Southwest Research Institute,
and powertrain testing at Oak Ridge National Laboratory. The agencies' assessment of the
comments is that Allison, ICCT, and Volvo support the proposed two percent effectiveness for
AT and AMT transmission types. In addition, the agencies reviewed the NACFE report on
electronically controlled transmissions (AT, AMT, and DCT).133 This report had similar
findings as those noted above in the NAS 2010 report. Electronically controlled transmissions
were found to be more fuel efficient than manual transmissions, though the amount varied
significantly. The report also stated that fleets found that electronically controlled transmissions
also reduced the fuel efficiency variability between drivers. Therefore after considering the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
comments related to effectiveness and additional reports, the agencies are adopting as proposed a
two percent effectiveness for AMT.
The agencies conducted powertrain testing at Oak Ridge National Laboratory using the
same HD diesel engine paired with an Eaton AMT and an Allison TC10 AT to evaluate the
impact of the transmission type on the CO2 emissions and fuel consumption.134 The Allison
TC10 transmission is their newest and most efficient heavy-duty automatic transmission and
contains the neutral-idle and first gear lock-up features. The agencies swept final drive ratio
during the testing to recognize that the proper spec'ing of the rear axle ratio will vary depending
on the type of transmission and the top gear ratio of the transmission. As shown in Figure 2-25
and Figure 2-26, the fuel consumption over the highway cycles simulating a Class 8 tractor-
trailer was similar between the two transmissions. Figure 2-27 shows that the TC10 automatic
transmission had lower fuel consumption over the transient cycle, but because the drive cycle
weighting of the ARB transient cycle is low in tractors, the agencies expect that automatic
transmissions designed for long haul operation and automated manual transmissions to perform
similarly and have similar effectiveness when compared to a manual transmission.
The benefit of the AMT's automatic shifting compared to a manual transmission is
recognized in GEM by simulating the MT as an AMT and increasing the emission results from
the simulation by two percent. For ATs, the agencies developed the default automatic
transmission inputs to GEM to represent a typical heavy-duty automatic transmission, which is
less efficient than the TC10. The agencies selected more conservative default transmission
losses in GEM so that we would not provide a false efficiency improvement for the less efficient
automatic transmissions that exist in the market today. Under the regulations in this rulemaking,
manufacturers that certify using the TC10 transmission would need to either conduct the optional
transmission gear efficiency testing or powertrain testing to recognize the benefits of this type of
automatic transmission in GEM. However, as noted in Section II.C.5 of the FRM Preamble, the
agencies could determine in a future action that it would be appropriate to modify GEM to be
equivalent to powertrain testing technology, rather than to require manufacturers to perform
powertrain testing to be credited for the full benefits of technologies such as advanced
transmissions. In such a case, the agencies would not consider the modification to GEM to
impact the effective stringency of the Phase 2 standards because the new version of GEM would
be equivalent to performing powertrain testing. Thus, we encourage manufacturers to work with
us in the coming years to investigate the potential to streamline the process for fully recognizing
advanced transmissions in GEM.

-------
E. O. 12866 Review - Revised - Do Not Cite, Quote, or Release During Review

590
JE
580
ap
570
_oj
560
o

u
550
s_
(D
540
>

O
530
~o

TC10
>AMT
2.5
3.5
Final Drive Ratio
Figure 2-25 Powertrain Test Results of AMT and AT over the 65 mph Cycle
Powertrain Testing Results - 55 mph Cycle
/ion

£ 475
— 470
aj
u 465
s-
^ 460
(D
> 455
"S 450
| 445
to
c 440
!=! 435
aj
£ 430



•





















•
•

• TC10



• AMT
•
• -
* •






• *



I
2.5 3 3.5
Final Drive Ratio
Figure 2-26 Powertrain Test Results of AMT and AT over the 55 mph Cycle

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Powertrain Testing - ARB Transient Cycle

765
JE

ap
760
(D

O



U

i_
755
(D

>

O

~o
TC10
>AMT
2.5	3	3.5
Final Drive Ratio
Figure 2-27 Powertrain Test Results of AMT and AT over the Transient Cycle
2.8.2.6.1	Transmission Efficiency
The agencies also proposed standards that considered the efficiency benefit of
transmissions that operate with top gear direct drive instead of overdrive. In the proposal, we
estimated that direct drive had 2 percent higher gear efficiency than an overdrive gear. 80 FR
40229. The benefit of direct drive was recognized through the transmission gear ratio inputs to
GEM. Direct drive leads to greater CO2 emissions and fuel consumption reductions in highway
operation, but virtually none in transient operation. ICCT cited a finding that highlighted
opportunities to improve transmission efficiency, including direct drive, which would provide
about two percent fuel consumption reduction.135 The agencies did not receive any negative
comments regarding the efficiency difference between direct drive and overdrive; therefore, we
continued to include the default transmission gear efficiency advantage of 2 percent for a gear
with a direct drive ratio in the version of GEM adopted for the final Phase 2 rules.
The agencies are also adopting in Phase 2 an optional transmission efficiency test (40
CFR 1037.565) for generating an input to GEM that overrides the default efficiency of each gear
based on the results of the test. Although optional, the transmission efficiency test will allow
manufacturers to reduce the CO2 emissions and fuel consumption by designing better
transmissions with lower friction due to better gear design and/or mandatory use of better
lubricants. The agencies project that transmission efficiency could improve 1 percent over the
2017 baseline transmission in Phase 2. Our assessment was based on comments received and
discussions with transmission manufacturers.136
2.8.2.6.2	Neutral Idl e
Automatic transmissions historically apply torque to an engine when in gear at zero speed
because of torque converter, such as when stopped at a traffic light. A neutral idle technology
can disengage transmission with torque converter, thus reducing power loss to a minimum. The

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
agencies simulated the impact of reducing the load on the engine at idle in GEM for tractors. As
expected, neutral idle had zero impact on the highway cycles because those cycles do not include
any idle time. During the ARB Transient cycle, neutral idle reduced CO2 emissions and fuel
consumption by 3.8 percent. The composite impact of neutral idle on CO2 emissions and fuel
consumption for day cabs is 1.2 percent and is 0.3 percent for sleeper cabs.
2.8.2.7 Drivetrain and Engine Downspeeding
Axle Configurations: Please see RIA Chapter 2.4.5.3 for the discussion on axle
configurations.
The agencies' assessments of these technologies show that the reductions are in the range
of 2 to 3 percent. For the final rule, the agencies are simulating 6x2, 4x2, and disengageable
axles within GEM instead of providing a fixed value for the reduction. This approach is more
technically sound because it will take into account future changes in axle efficiency. Tractor
simulations using Phase 2 GEM indicated that 6x4 and 4x2 axle configurations lead to a 2
percent improvement in day cab and sleeper cab tractor efficiency.
Downspeeding: Downspeeding would be as demonstrated through the Phase 2 GEM
inputs of transmission gear ratio, drive axle ratio, and tire diameter. Volvo offers an XE package
for fuel efficiency in 2017 MY that includes a downspeed package with a 2.64 rear axle ratio and
0.78 top transmission gear ratio, equivalent to a 2.06 final drive ratio (FDR). The agencies
evaluated the impact of downspeeding during a powertrain test of a heavy HD diesel engine and
automated manual transmission while simulating a Class 8 tractor-trailer.137 The results are
shown in Figure 2-28. Downspeeding from a 2.6 FDR to a 2.3 FDR reduced fuel consumption
by 2.5 percent.
540
535
oT 530
~! 525
"S 520
£
3 515
£
8 510
"ai
j? 505
500
495
1.
Final Drive Ratio Impact on Composite Fuel Consumption
Powertrain Test Results













































































.1
.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 3.
Final Drive Ratio
Figure 2-28 Downspeeding Impact on Fuel Consumption

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Axle Efficiency: Please see RIA Chapter 2.4.5.1 for additional discussion on
opportunities to improve axle efficiency. The 2010 NAS report assessed low friction lubricants
for the drivetrain as providing a 1 percent improvement in fuel consumption based on fleet
testing.138 The light-duty 2012-16 MY vehicle rule and the pickup truck portion of this program
estimate that low friction lubricants can have an effectiveness value between 0 and 1 percent
compared to traditional lubricants. In the Phase 2 proposal, the agencies proposed the reduction
in friction due to low viscosity axle lubricants of 0.5 percent. 80 FR 40217.
The agencies' assessment of axle improvements found that axles built in the Phase 2
timeline could be 2 percent more efficient than a 2017 baseline axle.139 In lieu of a fixed value
for low friction axle lubricants, the agencies are adopting an axle efficiency test procedure (40
CFR 1037.560), as discussed in the NPRM. 80 FR 40185. The axle efficiency test will be
optional, but will allow manufacturers to recognize in GEM reductions in CO2 emissions and
fuel consumption through improved axle gear designs and/or mandatory use of low friction
lubricants.
2.8.2.8 Accessories and Other Technologies
Reducing the mechanical and electrical loads of accessories reduce the power
requirement of the engine and in turn reduces the fuel consumption and CO2 emissions.
Modeling in GEM, as shown in Table 2-29, demonstrates the impact of reducing 1 kW of
accessory load for each tractor subcategory.
Table 2-29 Impact of 1 kW Accessory Load Reduction on CO2 Emissions

Tractor Subcategory
%C02 per kW
Class 8 High Roof Sleeper
0.5%
Class 8 Mid Roof Sleeper
0.5%
Class 8 Low Roof Sleeper
0.6%
Class 8 High Roof Day
0.6%
Class 8 Mid Roof Day
0.6%
Class 8 Low Roof Day
0.7%
Class 7 High Roof Day
0.8%
Class 7 Mid Roof Day
0.8%
Class 7 Low Roof Day
0.8%
Heavy Haul
0.5%
Compared to 2017 MY air conditioners, air conditioners with improved efficiency
compressors could reduce CO2 emissions by 0.5 percent. Improvements in accessories, such as
power steering, can lead to an efficiency improvement of 1 percent over the 2017MY baseline
(also see RIA Chapter 2.4.10). The agencies received several comments related to accessories.
Due to the complexity in determining a definition of that qualifies as an efficient accessory, we
are maintaining the proposed language for tractor accessories which provides defined
effectiveness values only for electric or high efficiency air conditioning compressors, electric

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
power steering pumps, and electric coolant pumps (if not already accounted for during the engine
fuel mapping procedure).
The agencies proposed to provide a two percent reduction for intelligent controls, such
as predictive cruise. Control. ICCT found in their workshop that opportunities exist for road
load optimization through predictive cruise, GPS, and driver feedback that could lead to a zero to
five percent reduction in fuel consumption and CO2 emissions.140 Daimler commented that
eCoast should also be recognized as an intelligent control within GEM. Eaton offers similar
technology, known as Neutral Coast Mode. The feature places an automated transmission in
neutral on downhill grades which allows the engine speed to go idle speed. A fuel savings is
recognized due to the difference in engine operating conditions. Based on literature information,
the agencies are adopting intelligent controls such as predictive cruise control with an
effectiveness of two percent (also see RIA Chapter 2.4.11) and neutral coasting with an
effectiveness of 1.5 percent.
2.8.2.9	Weight Reduction
The weight reductions were developed from tire manufacturer information, the
Aluminum Association, the Department of Energy, SABIC and TIAX. The fuel consumption
and CO2 emissions impact of a 1,000 pound weight reduction on tractors is approximately 1.2 to
1.5 percent based on simulations conducted in Phase 2 GEM. This reduction includes the impact
of both reducing the overall weight of the vehicle for the fraction of the fleet that is cubed-out
and the increase in payload capability for the fraction of the fleet that is weighed-out.
2.8.2.10	Vehicle Speed Limiter
The agencies did not include vehicle speed limiters in setting the Phase 1 stringency
levels. The agencies likewise are not including vehicle speed limiters in the technology package
for setting the standards for Class 7 and 8 tractors in Phase 2. The effectiveness of VSLs depend
on the type of tractor because it is dependent on the drive cycle. The greater the amount of time
spent at 65 mph, the greater the impact of a VSL set below 65 mph. Figure 2-29 shows the
effectiveness of VSL on sleeper and day cab tractors based on modeling conducted using Phase 2
GEM.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
12.0%
C
o
O
-g 10.0%
(D
C
£ 8.0%
Q.
E
3
c
o
u
(D
3
¦a
c
c
o
f\J
O
U
6.0%
4.0%
2.0%
0.0%
54
Vehicle Speed Limiter Effectiveness
Composite Result
-Sleeper Cab
¦ Day Cab
58	60	62
Vehicle Speed Limiter Setting (mph)
66
Figure 2-29 Vehicle Speed Limiter Impact on Tractor Fuel Consumption
2.8.2.11 Consideration of Phase 1 Credits in Phase 2 Stringency Setting
The agencies requested comment regarding the treatment of Phase 1 credits, as discussed
in Section I.C. 1 .b. See 80 FR 40251. As examples, the agencies discussed limiting the use of
Phase 1 credits in Phase 2 and factoring credit balances into the 2021 standards. Daimler
commented that allowing Phase 1 credits in Phase 2 is necessary to smooth the transition into a
new program that is very complex and that HD manufacturers cannot change over an entire
product portfolio at one time. The agencies evaluated the status of Phase 1 credit balances in
2015 by sector. For tractors, we found that manufacturers are generating significant credits, and
that it appears that many of the credits result from their use of an optional provision for
calculating aerodynamic drag. However, we also believe that manufacturers will generate fewer
credits in MY 2017 and later when the final Phase 1 standards begin. Still, the agencies believe
that manufacturers will have significant credits balances available to them for MYs 2021-2023,
and that much of these balances would be the result of the test procedure provisions rather than
pull ahead of any technology. Based on confidential product plans for MYs 2017 and later, we
expect this total windfall amount to be three percent of the MY 2021 standards or more.
Therefore, the agencies are factoring in a total credit amount equivalent to this three percent
credit (i.e. three years times 1 percent per year). Thus, we are increasing the stringency of the
CO2 and fuel consumption tractor standards for MYs 2021-2023 by 1 percent to reflect these
credits.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.8.2.12 Summary of Technology Performance
Table 2-30 describes the performance levels for the range of Class 7 and 8 tractor
technologies.
Table 2-30 Phase 2 Technology Inputs for Tractors

CLASS 7
CLASS 8

Day Cab
Day Cab
Sleeper Cab

Low
Roof
Mid
Roof
High
Roof
Low
Roof
Mid
Roof
High
Roof
Low
Roof
Mid
Roof
High
Roof
Engine

2021MY
11L
Engine
350 HP
2021MY
11L
Engine
350 HP
2021MY
11L
Engine
350 HP
2021MY
15L
Engine
455 HP
2021MY
15L
Engine
455 HP
2021MY
15L
Engine
455 HP
2021MY
15L
Engine
455 HP
2021MY
15L
Engine
455 HP
2021MY
15L
Engine
455 HP
Aerodynamics (Cd A in m2)
Bin I
6.00
7.00
7.45
6.00
7.00
7.45
6.00
7.00
7.15
Bin II
5.60
6.65
6.85
5.60
6.65
6.85
5.60
6.65
6.55
Bin III
5.15
6.25
6.25
5.15
6.25
6.25
5.15
6.25
5.95
Bin IV
4.75
5.85
5.70
4.75
5.85
5.70
4.75
5.85
5.40
Bin V
4.40
5.50
5.20
4.40
5.50
5.20
4.40
5.50
4.90
Bin VI
4.10
5.20
4.70
4.10
5.20
4.70
4.10
5.20
4.40
Bin VII
3.80
4.90
4.20
3.80
4.90
4.20
3.80
4.90
3.90
Steer Tires (CRR in kg/metric ton)
Base
7.8
7.8
7.8
7.8
7.8
7.8
7.8
7.8
7.8
Level 1
6.6
6.6
6.6
6.6
6.6
6.6
6.6
6.6
6.6
Level 2
5.7
5.7
5.7
5.7
5.7
5.7
5.7
5.7
5.7
Level 3
4.9
4.9
4.9
4.9
4.9
4.9
4.9
4.9
4.9
Drive Tires (CRR in kg/metric ton)
Base
8.1
8.1
8.1
8.1
8.1
8.1
8.1
8.1
8.1
Level 1
6.9
6.9
6.9
6.9
6.9
6.9
6.9
6.9
6.9
Level 2
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
Level 3
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
Idle Reduction (% reduction)
Tamper
Proof AESS
N/A
N/A
N/A
N/A
N/A
N/A
4%
4%
4%
Tamper
Proof AESS
with Diesel
APU
N/A
N/A
N/A
N/A
N/A
N/A
4%
4%
4%
Tamper
Proof AESS
with Battery
APU
N/A
N/A
N/A
N/A
N/A
N/A
6%
6%
6%

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Tamper
Proof AESS
with
Automatic
Stop-Start
N/A
N/A
N/A
N/A
N/A
N/A
3%
3%
3%
Tamper
Proof AESS
with FOH
N/A
N/A
N/A
N/A
N/A
N/A
3%
3%
3%
Adjustable
AESS
N/A
N/A
N/A
N/A
N/A
N/A
1%
1%
1%
Adjustable
AESS with
Diesel APU
N/A
N/A
N/A
N/A
N/A
N/A
3%
3%
3%
Adjustable
AESS with
Battery
APU
N/A
N/A
N/A
N/A
N/A
N/A
5%
5%
5%
Adjustable
AESS with
Automatic
Stop-Start
N/A
N/A
N/A
N/A
N/A
N/A
5%
5%
5%
Adjustable
AESS with
FOH
N/A
N/A
N/A
N/A
N/A
N/A
2%
2%
2%
Transmission (% reduction)
Manual
0%
0%
0%
0%
0%
0%
0%
0%
0%
AMT
2%
2%
2%
2%
2%
2%
2%
2%
2%
Auto
2%
2%
2%
2%
2%
2%
2%
2%
2%
Dual Clutch
2%
2%
2%
2%
2%
2%
2%
2%
2%
Top Gear
Direct Drive
2%
2%
2%
2%
2%
2%
2%
2%
2%
Transmissio
n Efficiency
Improvemen
ts
1%
1%
1%
1%
1%
1%
1%
1%
1%
Neutral Idle
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Driveline (% reduction)
Axle
Efficiency
Improvemen
ts
2%
2%
2%
2%
2%
2%
2%
2%
2%
6x2, 6x4
Axle
Disconnect
or 4x2 Axle
N/A
N/A
N/A
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Downspeed
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Modeled
in GEM
Accessory Improvements (% reduction)
A/C
Efficiency
0.5%
0.5%
0.5%
0.5%
0.5%
0.5%
0.5%
0.5%
0.5%
Electric
Access.
1%
1%
1%
1%
1%
1%
1%
1%
1%

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Other Technologies (% reduction)
Predictive
Cruise
Control
2%
2%
2%
2%
2%
2%
2%
2%
2%
Automated
Tire
Inflation
System
1.2%
1.2%
1.2%
1.2%
1.2%
1.2%
1.2%
1.2%
1.2%
Tire
Pressure
Monitoring
System
1%
1%
1%
1%
1%
1%
1%
1%
1%
Neutral
Coast
1.5%
1.5%
1.5%
1.5%
1.5%
1.5%
1.5%
1.5%
1.5%
2.8.3 Tractor Technology Adoption Rates
Often tractor manufacturers introduce major product changes together, as a package.
This allows manufacturers to optimize their available resources, including engineering,
development, manufacturing and marketing activities to create a product with multiple new
features. In some limited cases, manufacturers may implement an individual technology outside
of a vehicle's redesign cycle. It is recognized by the manufacturers that a vehicle design will
need to remain competitive over the intended life of the design and meet future regulatory
requirements.
With respect to the levels of technology adoption used to develop the HD Phase 2
standards, NHTSA and EPA established two types of technology adoption constraints. The first
type of constraint was established based on the application of fuel consumption and CO2
emission reduction technologies into the different types of tractors. For example, extended idle
reduction technologies are limited to Class 8 sleeper cabs based on the (reasonable) assumption
that day cabs are not used for overnight hoteling.
A second type of constraint was applied to most other technologies and limited their
adoption based on factors reflecting the real world operating conditions that some combination
tractors encounter. This second type of constraint was applied to the aerodynamic, tire,
powertrain, vehicle speed limiter, and other technologies. Table 2-34,

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-35 and Table 2-36 specify the adoption rates that EPA and NHTSA used to
develop the final Phase 2 standards
NHTSA and EPA believe that within each of these individual vehicle categories there are
particular applications where the use of the identified technologies would be either ineffective or
not technically feasible. The addition of ineffective technologies provides no environmental or
fuel efficiency benefit, increases costs and is not a basis upon which to set a maximum feasible
improvement under 49 USC Section 32902 (k), or appropriate under 42 U.S.C. Section 7521
(a)(2). For example, the agencies are not predicating the standards on the use of full
aerodynamic vehicle treatments on 100 percent of tractors, because we know that in many
applications (for example gravel truck engaged in local aggregate delivery) the added weight of
the aerodynamic technologies would increase fuel consumption and hence CO2 emissions to a
greater degree than the reductions from the aerodynamic technology. .
Discussions related to our responses to comments received on technology adoption rates
for each of the technologies are included in Preamble Section III.D.l.c and in Section 4.3 of the
response to comments document. The sections below contain the final decisions based on the
consideration of these comments and any new data or information.
2.8.3.1 Aerodynamics Adoption Rate
The impact of aerodynamics on a tractor-trailer's efficiency increases with vehicle speed.
Therefore, the usage pattern of the vehicle will determine the benefit of various aerodynamic
technologies. Sleeper cabs are often used in line haul applications and drive the majority of their
miles on the highway travelling at speeds greater than 55 mph. The industry has focused
aerodynamic technology development, including SmartWay tractors, on these types of trucks.
Therefore the most aggressive aerodynamic technologies are applied to this regulatory
subcategory. All of the major manufacturers today offer at least one SmartWay sleeper cab
tractor model, which is represented as Bin III aerodynamic performance. For the NPRM, the
agencies developed a technology package for 2027 MY that included the aerodynamic adoption
rates shown in Table 2-31. 80 FR 40227.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-31 Proposed Aerodynamic Bin Adoption Rates for 2027 MY Tractors

CLASS 7
CLASS 8

Day Cab
Day Cab
Sleeper Cab

Low
Roof
Mid
Roof
High
Roof
Low
Roof
Mid
Roof
High
Roof
Low
Roof
Mid
Roof
High
Roof
Aerodynamics
Bin I
0%
0%
0%
0%
0%
0%
0%
0%
0%
Bin II
50%
50%
0%
50%
50%
0%
50%
50%
0%
Bin III
40%
40%
20%
40%
40%
20%
40%
40%
20%
Bin IV
10%
10%
20%
10%
10%
20%
10%
10%
20%
Bin V
N/A
N/A
35%
N/A
N/A
35%
N/A
N/A
35%
Bin VI
N/A
N/A
20%
N/A
N/A
20%
N/A
N/A
20%
Bin VII
N/A
N/A
5%
N/A
N/A
5%
N/A
N/A
5%
In Phase 1, the agencies determined the stringency of the tractor standards through the
use of a mix of aerodynamic bins in the technology packages. For example, we included 10
percent Bin II, 70 percent Bin III, and 20 percent Bin IV in the high roof sleeper cab tractor
standard. The weighted average aerodynamic performance of this technology package is
equivalent to Bin III. 76 FR 57211. In consideration of the comments, the agencies have
adjusted the aerodynamic adoption rate for Class 8 high roof sleeper cabs used to set the final
standards in 2021, 2024, and 2027 MYs {i.e., the degree of technology adoption on which the
stringency of the standard is premised). Upon further analysis of simulation modeling of a
SuperTruck tractor with a Phase 2 reference trailer with skirts, we agree with the manufacturers
that a SuperTruck tractor technology package would only achieve the Bin V level of CdA, as
discussed above in RIA Chapter 2.8.2.2. Consequently, the final standards are not premised on
any adoption of Bin VI and VII technologies. Accordingly, we determined the adoption rates in
the technology packages developed for the final rule using a similar approach as Phase 1 -
spanning three aerodynamic bins and not setting adoption rates in the most aerodynamic bin(s) -
to reflect that there are some vehicles whose operation limits the applicability of some
aerodynamic technologies. We set the MY 2027 high roof sleeper cab tractor standards using a
technology package that included 20 percent of Bin III, 30 percent Bin IV, and 50 percent Bin V
reflecting our assessment of the fraction of high roof sleeper cab tractors that we project could
successfully apply these aerodynamic packages with this amount of lead time. The weighted
average of this set of adoption rates is equivalent to a tractor aerodynamic performance near the
border between Bin IV and Bin V. We believe that there is sufficient lead time to develop
aerodynamic tractors that can move the entire high roof sleeper cab aerodynamic performance to
be as good as or better than today's SmartWay designated tractors.
The agencies phased-in the aerodynamic technology adoption rates within the technology
packages used to determine the MY 2021 and 2024 standards so that manufacturers can
gradually introduce these technologies. The changes required for Bin V performance reflect the
kinds of improvements projected in the Department of Energy's SuperTruck program. That
program has demonstrated tractor-trailers in 2015 with significant aerodynamic technologies.
For the final rule, the agencies are projecting that truck manufacturers will be able to begin
implementing some of these aerodynamic technologies on high roof tractors as early as 2021 MY
on a limited scale. For example, in the 2021 MY technology package, the agencies have

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
assumed that 10 percent of high roof sleeper cabs will have aerodynamics better than today's
best tractors. This phase-in structure is consistent with the normal manner in which
manufacturers introduce new technology to manage limited research and development budgets as
well as to allow them to work with fleets to fully evaluate in-use reliability before a technology
is applied fleet-wide. The agencies believe the phase-in schedule will allow manufacturers to
complete these normal processes. Overall, while the agencies are now projecting slightly less
benefit from aerodynamic improvements than we did in the NPRM, the actual aerodynamic
technologies being projected are very similar to what was projected at the time of NPRM
(however, these vehicles fall into Bin V in the final rule, instead of Bin VI and VII in the
NPRM). Importantly, our averaging, banking and trading provisions provide manufacturers with
the flexibility (and incentive) to implement these technologies over time even though the
standard changes in a single step.
The agencies also received comment regarding our aerodynamic assessment of the other
tractor subcategories. Aerodynamic improvements through new tractor designs and the
development of new aerodynamic components is an inherently slow and iterative process. The
agencies recognize that there are tractor applications that require on/off-road capability and other
truck functions which restrict the type of aerodynamic equipment applicable. We also recognize
that these types of trucks spend less time at highway speeds where aerodynamic technologies
have the greatest benefit. The 2002 VIUS data ranks trucks by major use.141 The heavy trucks
usage indicates that up to 35 percent of the trucks may be used in on/off-road applications or
heavier applications. The uses include construction (16 percent), agriculture (12 percent), waste
management (5 percent), and mining (2 percent). Therefore, the agencies analyzed the
technologies to evaluate the potential restrictions that will prevent 100 percent adoption of more
advanced aerodynamic technologies for all of the tractor regulatory subcategories and developed
standards with new penetration rates reflecting that these vehicles spend less time at highway
speeds. For the final rule, the agencies evaluated the certification data to assess how the
aerodynamic performance of high roof day cabs compare to high roof sleeper cabs. In 2014, the
high roof day cabs on average are certified to one bin lower than the high roof sleeper cabs.142
Consistent with the public comments, and the certification data, the aerodynamic adoption rates
used to develop the final Phase 2 standards for the high roof day cab regulatory subcategories are
less aggressive than for the Class 8 sleeper cab high roof tractors. In addition, the agencies are
also accordingly reducing the adoption rates in the highest bins for low and mid roof tractors to
follow the changes made to the high roof subcategories because we neither proposed nor expect
the aerodynamics of a low or mid roof tractor to be better than a high roof tractor.
2.8.3.2 Low Rolling Resistance Tire Adoption Rate
For the tire manufacturers to further reduce tire rolling resistance, the manufacturers must
consider several performance criteria that affect tire selection. The characteristics of a tire also
influence durability, traction control, vehicle handling, comfort, and retreadability. Tire design
requires balancing performance, since changes in design may change different performance
characteristics in opposing directions. A single performance parameter can easily be enhanced,
but an optimal balance of all the criteria would require improvements in materials and tread
design at a higher cost, as estimated by the agencies.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
For the final rulemaking, the agencies evaluated the tire rolling resistance levels in the
Phase 1 certification data.143 We found that high roof sleeper cabs are certified today with steer
tire rolling resistance levels that ranged between 4.9 and 7.6 kg/ton and with drive tires ranging
between 5.1 and 9.8 kg/ton. In the same analysis, we found that high roof day cabs are certified
with rolling resistance levels ranging between 4.9 and 9.0 kg/ton for steer tires and between 5.1
and 9.8 kg/ton for drive tires. This range spans the baseline through Level 3 rolling resistance
performance levels. Therefore, for the final rule we took an approach similar to the one taken in
Phase 1 and proposed in Phase 2 that considers adoption rates across a wide range of tire rolling
resistance levels to recognize that operators may have different needs. 76 FR 57211 and 80 FR
40227.
In our analysis of the Phase 1 certification data, we found that the drive tires on low and
mid roof sleeper cab tractors on average had 10 to 17 percent higher rolling resistance than the
high roof sleeper cabs. But we found only a minor difference in rolling resistance of the steer
tires between the tractor subcategories. Based on comments received and further consideration
of our own analysis of the difference in tire rolling resistance levels that exist today in the
certification data, the agencies are adopting Phase 2 standards using a technology pathway that
utilizes higher rolling resistance levels for low and mid roof tractors than the levels used to set
the high roof tractor standards. This is also consistent with the approach that we took in setting
the Phase 1 tractor standards. 76 FR 57211. In addition, the final rule reflects a reduction in
Level 3 adoption rates for low and mid roof tractors from 25 percent in MY 2027 used at
proposal (80 FR 40227) to zero percent adoption rate. The technology packages developed for
the low and mid roof tractors used to determine the stringency of the MY 2027 standards in the
final rule do not include any adoption rate of Level 3 drive tires to recognize the special needs of
these applications, consistent with the comments noted above raising concerns about applications
that limit the use of low rolling resistance tires.
The agencies phased-in the low rolling resistance tire adoption rates within the
technology packages used to determine the MY 2021 and 2024 standards so that manufacturers
can gradually introduce these technologies. In addition, the levels of rolling resistance used in
all of the technology packages are achievable with either dual or wide based single tires, so the
agencies are not forcing one technology over another. See Table 2-34 through Table 2-36 for the
adoption rates of each tractor subcategory.
2.8.3.3 Tire Pressure Monitoring System and Automatic Tire Inflation System
Adoption Rates
The agencies used a 20 percent adoption rate of ATIS in MY 2021 and a 40 percent
adoption rate in setting the proposed Phase 2 MY 2024 and 2027 tractor standards. 80 FR
40227.
The agencies received a number of comments on ATIS and TPMS. The agencies find the
comments related to a greater acceptance of TPMS in the tractor market to be persuasive.
However, available information indicates that it is feasible to utilize either TPMS or ATIS to
reduce the prevalence on underinflated tires in-use on all tractors. As a result, we are finalizing
tractor standards that are predicated on the performance of a mix of TPMS and ATIS adoption
rates in all tractor subcategories. The agencies are using adoption rates of 30 percent of ATIS

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
and 70 percent of TPMS in the technology packages used in setting the final Phase 2 MY 2027
tractor standards. This represents a lower adoption rate of ATIS than used in the NPRM, but the
agencies have added additional adoption rate of TPMS because none of the comments or
available information disputed the ability to use it on all tractors. The agencies have developed
technology packages for setting the 2021 and 2024 MY standards which reflect a phase in of
adoption rates of each of these technologies. In 2021 MY, the adoption rates consist of 20
percent TPMS and 20 percent ATIS. In 2024 MY, the adoption rates are 50 percent TPMS and
25 percent ATIS.
2.8.3.4	Weight Reduction Technology Adoption Rate
The agencies set the 2021 through 2027 model year tractor standards without using
weight reduction as a technology on whose performance the standard is predicated. The
agencies view weight reduction as a technology with a high cost that offers a small benefit in the
tractor sector. For example, our estimate of a 400 pound weight reduction would cost $2,050
(2012$) in MY2021, but offer a 0.3 percent reduction in fuel consumption and CO2 emissions.
Nonetheless, the agencies are adopting an expanded list of weight reduction options which could
be input into the GEM by the manufacturers to reduce their certified CO2 emission and fuel
consumption levels.
2.8.3.5	Idle Reduction Technology Adoption Rate
Idle reduction technologies provide significant reductions in fuel consumption and CO2
emissions for Class 8 sleeper cabs and are available on the market today. There are several
different technologies available to reduce idling. These include APUs, diesel fired heaters, and
battery powered units. Our discussions with manufacturers prior to the Phase 2 NPRM indicated
that idle technologies are sometimes installed in the factory, but that it is also a common practice
to have the units installed after the sale of the truck. We want to continue to incentivize this
practice and to do so in a manner that the emission reductions associated with idle reduction
technology occur in use. We proposed to continue the Phase 1 approach into Phase 2 where we
recognize only idle emission reduction technologies that include a tamper-proof automatic
engine shutoff system (AESS) with some override provisions.0
We used an overall 90 percent adoption rate of tamper-proof AESS for Class 8 sleeper
cabs in setting the proposed MY 2024 and 2027 standards. Id. The agencies stated in the Phase
2 NPRM that we were unaware of reasons why AESS with extended idle reduction technologies
could not be applied to this high fraction of tractors with a sleeper cab, except those deemed a
vocational tractor, in the available lead time.
The agencies received numerous comments on idle reduction adoption rates and the need
to consider adjustable AESS (see Section III.D.l.c.v of the Preamble). The agencies find the
comments regarding the concerns for using 90 percent adoption rates of tamper-proof AESS to
be persuasive. For the final rule, the agencies developed a menu of idle reduction technologies
that include both tamper-proof and adjustable AESS (as discussed in Section III.D.l.b) that are
G The agencies are retaining the HD Phase 1 AESS override provisions included in 40 CFR 1037.660(b) for driver
safety.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
recognized at different levels of effectiveness in GEM. As discussed in the discussion of tractor
baselines (Section III.D. 1 .a), the latest NACFE confidence report found that 96 percent of HD
vehicles are equipped with adjustable automatic engine shutdown systems.144 Therefore, the
agencies built this level of idle reduction into the baseline for sleeper cab tractors. Due to the
high percentage acceptance of adjustable AESS today, the agencies project that by 2027 MY it is
feasible for 100 percent of sleeper cabs to contain some type of AESS and idle reduction
technology to meet the hoteling needs of the driver. However, we recognize that there are a
variety of idle reduction technologies that meet the various needs of specific customers and not
all customers will select diesel powered APUs due to the cost or weight concerns highlighted in
the comments. Therefore, we developed an idle reduction technology package for each MY that
reflects this variety. The idle reduction packages developed for the final rule contain lower
AESS adoption rates than used at proposal. The AESS used during the NPRM assumed that it
also included a diesel powered APU in terms of determining the effectiveness and costs. In the
final rule, the idle reduction technology mix actually has an overall lower cost (even after
increasing the diesel APU technology cost for the final rule) than would have been developed for
the final rule. In addition, the stringency of the tractor standards are not affected because the
higher penetration rate of other idle reduction technologies, which are not quite as effective, but
will be deployed more. We developed the technology package to set the 2027 MY sleeper cab
tractor standards that includes 15 percent adoption rate of adjustable AESS only, 40 percent of
adjustable AESS with a diesel powered APU, 15 percent adjustable AESS with a battery APU,
15 percent adjustable AESS with automatic stop/start, and 15 percent adjustable AESS with a
fuel operated heater. We continued the same approach of phasing in different technology
packages for the 2021 and 2024 MY standards, though we included some type of idle reduction
on 100 percent of the sleeper cab tractors. The 2021 MY technology package had a higher
adoption rate of adjustable AESS with no other idle reduction technology and lower adoption
rates of adjustable AESS with other idle reduction technologies.
2.8.3.6 Transmission Adoption Rates
The agencies' proposed standards included a 55, 80, and 90 percent adoption rate of
automatic, automated manual, and dual clutch transmissions in MYs 2021, 2024, and 2027
respectively. 80 FR 40225-7. The agencies did not receive any comments regarding these
proposed transmission adoption rates, and have not found any other information suggesting a
change in approach. Therefore, we are including the same level of adoption rates in setting the
final rule standards. The MY 2021 and 2024 standards are likewise premised on the same
adoption rates of these transmission technologies as at proposal.
The agencies have added neutral idle as a technology input to GEM for Phase 2 in the
final rulemaking. The TC10 that was tested by the agencies for the final rule included this
technology. Therefore, we projected that neutral idle would be included in all of the automatic
transmissions and therefore the adoption rates of neutral idle match the adoption rates of the
automatic transmission in each of the MYs.
Transmissions with direct drive as the top gear and numerically lower axles are better
suited for applications with primarily highway driving with flat or low rolling hills. Therefore,
this technology is not appropriate for use in 100 percent of tractors. The agencies proposed
standards reflected the projection that 50 percent of the tractors would have direct drive in top

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
gear in MYs 2024 and 2027. 80 FR 40226-7. The agencies did not receive any comments
regarding the adoption rates of transmissions with direct drive in those MYs. We therefore are
including the same level of adoption rates in setting the final rule standards for MYs 2024 and
2027. Transmissions with direct drive top gears exist in the market today, therefore, the agencies
determined it is feasible to also include this technology in the package for setting the 2021 MY
standards. For the final rule, the agencies included a 20 percent adoption rate of direct drive in
the 2021 MY technology package.
The agencies received comments supporting establishing a transmission efficiency test
that measures the efficiency of each transmission gear and could be input into GEM. In the final
rule, the agencies are adopting Phase 2 standards that project that 20, 40, and 70 percent of the
AMT and DCT transmissions will be tested and achieve a fuel consumption and CO2 emissions
reduction of one percent in MYs 2021, 2024, and 2027, respectively.
2.8.3.7 Engine Downspeeding Adoption Rates
The agencies proposed to include lower final drive ratios in setting the Phase 2 standards
to account for engine downspeeding. In the NPRM, we used a transmission top gear ratio of
0.73 and baseline drive axle ratio of 3.70 in 2017 going down to a rear axle ratio of 3.55 in 2021
MY, 3.36 in 2024 MY, and 3.20 in 2027 MY. 80 FR 40228-30.
UCS commented that downspeeding was only partially captured as proposed. The
agencies also received additional information from vehicle manufacturers and axle
manufacturers that we believe supports using lower numerical drive axle ratios in setting the
final Phase 2 standards for sleeper cabs that spend more time on the highway than day cabs,
directionally consistent with the UCS comment. For the final rules, the agencies have used 3.70
in the baseline and 3.16 for sleeper cabs and 3.21 for day cabs in MY 2027 to account for
continued downspeeding opportunities. The final drive ratios used for setting the other model
years are shown in Table 2-32. These values represent the "average" tractor in each of the MYs,
but there will be a range of final drive ratios that contain more aggressive engine downspeeding
on some tractors and less aggressive on others.
Table 2-32 Final Drive Ratio for Tractor Technology Packages
MODEL
YEAR
REAR AXLE
RATIO
TRANSMISSION
TOP GEAR
RATIO
FINAL DRIVE
RATIO
Sleeper Cabs
2018
3.70
0.73
2.70
2021
3.31
0.73
2.42
2024
3.26
0.73
2.38
2027
3.16
0.73
2.31

Day Cabs
2018
3.70
0.73
2.70
2021
3.36
0.73
2.45
2024
3.31
0.73
2.42
2027
3.21
0.73
2.34

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.8.3.8 Drivetrain Adoption Rates
The agencies' proposed standards included 6x2 axle adoption rates in high roof tractors
of 20 percent in 2021 MY and 60 percent in MYs 2024 and 2027. Because 6x2 axle
configurations could raise concerns of traction, the agencies proposed standards that reflected
lower adoption rates of 6x2 axles in low and mid roof tractors recognizing that these tractors may
require some unique capabilities. The agencies proposed standards for low and mid roof tractors
that included 6x2 axle adoption rates of 10 percent in MY 2021 and 20 percent in MYs 2024 and
2027. 80 FR 40225-7.
ATA and others commented that limitations to a high penetration rate of 6x2 axles
include curb cuts, other uneven terrain features that could expose the truck to traction issues,
lower residual values, traction issues, driver dissatisfaction, tire wear, and the legality of their
use. Upon further consideration, the agencies have reduced the adoption rate of 6x2 axles and
projected a 30 percent adoption rate in the technology package used to determine the Phase 2
2027 MY standards. The 2021 MY standards include an adoption rate of 15 percent and the
2024 MY standards include an adoption rate of 25 percent 6x2 axles. This adoption rate
represents a combination of liftable 6x2 axles (which as noted in ATA's comments are allowed
in all states but Utah, and Utah is expected to revise their law) and 4x2 axles. In addition, it is
worth recognizing that state regulations related to 6x2 axles could change significantly over the
next ten years.
In the NPRM, the agencies projected that 20 percent of 2021 MY and 40 percent of the
2024 and 2027 MY axles would use low friction axle lubricants. 80 FR 40225-7. In the final
rule, we are requiring that manufacturers conduct an axle efficiency test if they want to include
the benefit of low friction lubricant or other axle design improvements when certifying in GEM.
The axle efficiency test will be optional, but will allow manufacturers to reduce CO2 emissions
and fuel consumption if the manufacturers have improved axle gear designs and/or mandatory
use of low friction lubricants. The agencies' assessment of axle improvements found that 80
percent of the axles built in MY 2027 could be two percent more efficient than a 2017 baseline
axle. Because it will take time for axle manufacturers to make improvements across the majority
of their product offerings, the agencies phased in the amount of axle efficiency improvements in
the technology packages in setting the 2021 and 2024 MY standards to include 30 and 65 percent
adoption rates, respectively.
2.8.3.9 Accessories and Other Technology Adoption Rates
In the NPRM, the agencies projected adoption rates as show in Table 2-33. 80 FR 40227.
The agencies are adopting the same level of adoption rates for setting the final Phase 2 standards
because we did not receive any comments or new data to support a change in the adoption rates
used in the proposal.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-33 Adoption Rates used in the Tractor Technology Packages in the NPRM
MODEL YEAR
PREDICTIVE
CRUISE
CONTROL
ELECTRIFIED
ACCESSORIES
HIGHER
EFFICIENCY AIR
CONDITIONING

2021
20%
10%
10%
2024
40%
20%
20%
2027
40%
30%
30%
2.8.3.10 Vehicle Speed Limiter Adoption Rate
As adopted in Phase 1, we are continuing the approach where vehicle speed limiters may
be used as a technology to aid in meeting the standard. In setting the standard, however, we
assumed a zero percent adoption rate of vehicle speed limiters. Although we believe vehicle
speed limiters are a simple, easy to implement, and inexpensive technology, we want to leave the
use of vehicle speed limiters to the truck purchaser. Since truck fleets purchase tractors today
with owner-set vehicle speed limiters, we considered not allowing GEM to recognize
performance of VSLs due to potential issues regarding whether any reductions would accrue
from installing VSLs, since they can be turned off. We ultimately concluded, as we did in Phase
1, that we should allow the use of VSLs that cannot be overridden by the operator as a means of
compliance for vehicle manufacturers that wish to offer it and truck purchasers that wish to
purchase the technology. In doing so, we are providing another means of meeting that standard
that can lower compliance cost and provide a more optimal vehicle solution for some truck
fleets. For example, a local beverage distributor may operate trucks in a distribution network of
primarily local roads. Under those conditions, aerodynamic fairings used to reduce aerodynamic
drag provide little benefit due to the low vehicle speed while adding additional mass to the
vehicle. A vehicle manufacturer could choose to install a VSL set at 55 mph for this customer.
The resulting tractor would be optimized for its intended application and would be fully
compliant with our program all at a lower cost to the ultimate tractor purchaser.H
However, as in Phase 1, we have chosen not to base the standards on performance of
VSLs because of concerns about how to set a realistic adoption rate that avoids unintended
adverse impacts. Although we expect there will be some use of VSL, currently it is used when
the fleet involved decides it is feasible and practicable and increases the overall efficiency of the
freight system for that fleet operator. To date, the compliance data provided by manufacturers
indicate that none of the tractor configurations include a tamper-proof VSL setting less than 65
mph. At this point the agencies are not in a position to determine in how many additional
situations use of a VSL would result in similar benefits to overall efficiency or how many
customers would be willing to accept a tamper-proof VSL setting. We are not able at this time to
quantify the potential loss in utility due to the use of VSLs. Absent this information, we cannot
make a determination regarding the reasonableness of setting a standard based on a particular
H The agencies note that because a VSL value can be input into GEM, its benefits can be directly assessed with the
model and off cycle credit applications therefore are not necessary even though the standard is not based on
performance of VSLs (i.e. VSL is an on-cycle technology).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
VSL level. Therefore, the agencies are not premising the standards on use of VSL, and instead
would continue to rely on the industry to select VSL when circumstances are appropriate for its
use. The agencies have not included either the cost or benefit due to VSLs in analysis of the
program's costs and benefits.
2.8.3.11 Adoption Rates Used to Set the Heavy-Haul Tractor Standards
The agencies recognize that certain technologies used to determine the stringency of the
Phase 2 tractor standards are less applicable to heavy-haul tractors. Heavy-haul tractors are not
typically used in the same manner as long-haul tractors with extended highway driving, and
therefore will experience less benefit from aerodynamics. Aerodynamic technologies are very
effective at reducing the fuel consumption and GHG emissions of tractors, but only when
traveling at highway speeds. At lower speeds, the aerodynamic technologies may have a
detrimental impact due to the potential of added weight. The agencies therefore proposed not
considering the use of aerodynamic technologies in the development of the Phase 2 heavy-haul
tractor standards. Moreover, because aerodynamics will not play a role in the heavy-haul
standards, the agencies proposed to combine all of the heavy-haul tractor cab configurations (day
and sleeper) and roof heights (low, mid, and high) into a single heavy-haul tractor subcategory.
The agencies received comments regarding the applicability of aerodynamic technologies
on heavy-haul vehicles. After considering these comments, the agencies are using a technology
package that does not use aerodynamic improvements in setting the Phase 2 heavy-haul tractor
standards, as we proposed. 1
Certain powertrain and drivetrain components are also impacted during the design of a
heavy-haul tractor, including the transmission, axles, and the engine. Heavy-haul tractors
typically require transmissions with 13 or 18 speeds to provide the ratio spread to ensure that the
tractor is able to start pulling the load from a stop. Downspeed powertrains are typically not an
option for heavy-haul operations because these vehicles require more torque to move the vehicle
because of the heavier load. Finally, due to the loading requirements of the vehicle, it is not
likely that a 6x2 axle configuration can be used in heavy-haul applications.
We received comments from stakeholders about the application of technologies other
than aerodynamics for heavy-haul tractors. After considering these comments and the
information regarding the tire rolling resistance improvement opportunities, discussed in Section
III.D.l.b.iii, the agencies have adjusted the adoption rate of low rolling resistance tires.
Consistent with the changes made in the final rule for the adoption of low rolling resistance tires
in low and mid roof tractors, the agencies did not project any adoption of Level 3 tires for heavy-
haul tractors in the final rule.
2.8.3.12 Summary of the Adoption Rates used to determine the Standards
Table 2-34,
1 Since aerodynamic improvements are not part of the technology package, the agencies likewise are not adopting
any aero bin structure for the heavy-haul tractor subcategory.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-35, and Table 2-36 provide the adoption rates of each technology broken down
by weight class, cab configuration, and roof height.
Table 2-34 Technology Adoption Rates for Class 7 and 8 Tractors for Determining the 2021 MY Standards

CLASS 7
CLASS 8

Day Cab
Day Cab
Sleeper Cab

Low
Roof
Mid Roof
High
Roof
Low Roof
Mid Roof
High
Roof
Low Roof
Mid Roof
High Roof
Engine

2021MY
11L
Engine
350 HP
2021MY
11L
Engine
350 HP
2021MY
11L
Engine
350 HP
2021MY
15L
Engine
455 HP
2021MY
15L
Engine
455 HP
2021MY
15L
Engine
455 HP
2021MY
15L
Engine
455 HP
2021MY
15L
Engine
455 HP
2021MY
15L
Engine 455
HP
Aerodynamics
Bin I
10%
10%
0%
10%
10%
0%
0%
10%
0%
Bin II
10%
10%
0%
10%
10%
0%
20%
10%
0%
Bin III
70%
70%
60%
70%
70%
60%
60%
70%
60%
Bin IV
10%
10%
35%
10%
10%
35%
20%
10%
30%
Bin V
0%
0%
5%
0%
0%
5%
0%
0%
10%
Bin VI
0%
0%
0%
0%
0%
0%
0%
0%
0%
Bin VII
0%
0%
0%
0%
0%
0%
0%
0%
0%
Steer Tires
Base
5%
5%
5%
5%
5%
5%
5%
5%
5%
Level 1
35%
35%
35%
35%
35%
35%
35%
35%
35%
Level 2
50%
50%
50%
50%
50%
50%
50%
50%
50%
Level 3
10%
10%
10%
10%
10%
10%
10%
10%
10%
Drive Tires
Base
15%
15%
5%
15%
15%
5%
15%
15%
5%
Level 1
35%
35%
35%
35%
35%
35%
35%
35%
35%
Level 2
50%
50%
50%
50%
50%
50%
50%
50%
50%
Level 3
0%
0%
10%
0%
0%
10%
0%
0%
10%
Idle Reduction
Tamper Proof
AESS
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Tamper Proof
AESS with
Diesel APU
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Tamper Proof
AESS with
Battery APU
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Tamper Proof
AESS with
Automatic
Stop-Start
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Tamper Proof
AESS with
FOH
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
40%
40%
40%
AESS









Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
30%
30%
30%
AESS with









Diesel APU









Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
10%
10%
10%
AESS with









Battery APU









Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
10%
10%
10%
AESS with









Automatic









Stop-Start









Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
10%
10%
10%
AESS with









FOH









Transmission
Manual
0%
0%
0%
0%
0%
0%
0%
0%
0%
AMT
40%
40%
40%
40%
40%
40%
40%
40%
40%
Auto
10%
10%
10%
10%
10%
10%
10%
10%
10%
Dual Clutch
5%
5%
5%
5%
5%
5%
5%
5%
5%
Top Gear
20%
20%
20%
20%
20%
20%
20%
20%
20%
Direct Drive









Transmission
20%
20%
20%
20%
20%
20%
20%
20%
20%
Efficiency
Improvement









Neutral Idle
10%
10%
10%
10%
10%
10%
10%
10%
10%
Driveline
Axle
30%
30%
30%
30%
30%
30%
30%
30%
30%
Efficiency
Improvement









6x2, 6x4 Axle
N/A
N/A
N/A
15%
15%
15%
15%
15%
15%
Disconnect or









4x2 Axle









Downspeed
3.36
3.36
3.36
3.36
3.36
3.36
3.31
3.31
3.31
(Rear Axle









Ratio)









Accessory Improvements
A/C
10%
10%
10%
10%
10%
10%
10%
10%
10%
Efficiency









Electric
10%
10%
10%
10%
10%
10%
10%
10%
10%
Access.









Other Technologies
Predictive
20%
20%
20%
20%
20%
20%
20%
20%
20%
Cruise









Control









Automated
20%
20%
20%
20%
20%
20%
20%
20%
20%
Tire Inflation









System









Tire Pressure
20%
20%
20%
20%
20%
20%
20%
20%
20%
Monitoring
System









Neutral Coast
0%
0%
0%
0%
0%
0%
0%
0%
0%

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-35 Technology Adoption Rates for Class 7 and 8 Tractors for Determining the 2024 MY Standards

CLASS 7
CLASS 8

Day Cab
Day Cab
Sleeper Cab

Low Roof
Mid Roof
High
Roof
Low Roof
Mid Roof
High
Roof
Low Roof
Mid Roof
High Roof
Engine

2024MY
11L
Engine
350 HP
2024MY
11L
Engine
350 HP
2024MY
11L
Engine
350 HP
2024MY
15L
Engine
455 HP
2024MY
15L
Engine
455 HP
2024MY
15L
Engine
455 HP
2024MY
15L
Engine
455 HP
2024MY
15L
Engine
455 HP
2024MY
15L
Engine
455 HP
Aerodynamics
Bin I
0%
0%
0%
0%
0%
0%
0%
0%
0%
Bin II
20%
20%
0%
20%
20%
0%
20%
20%
0%
Bin III
60%
60%
40%
60%
60%
40%
60%
60%
40%
Bin IV
20%
20%
40%
20%
20%
40%
20%
20%
40%
Bin V
0%
0%
20%
0%
0%
20%
0%
0%
20%
Bin VI
0%
0%
0%
0%
0%
0%
0%
0%
0%
Bin VII
0%
0%
0%
0%
0%
0%
0%
0%
0%
Steer Tires
Base
5%
5%
5%
5%
5%
5%
5%
5%
5%
Level 1
25%
25%
15%
25%
25%
15%
25%
25%
15%
Level 2
55%
55%
60%
55%
55%
60%
55%
55%
60%
Level 3
15%
15%
20%
15%
15%
20%
15%
15%
20%
Drive Tires
Base
10%
10%
5%
10%
10%
5%
10%
10%
5%
Level 1
25%
25%
15%
25%
25%
15%
25%
25%
15%
Level 2
65%
65%
60%
65%
65%
60%
65%
65%
60%
Level 3
0%
0%
20%
0%
0%
20%
0%
0%
20%
Idle Reduction
Tamper
Proof AESS
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Tamper
Proof AESS
with Diesel
APU
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Tamper
Proof AESS
with Battery
APU
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Tamper
Proof AESS
with
Automatic
Stop-Start
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Tamper
Proof AESS
withFOH
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Adjustable
AESS
N/A
N/A
N/A
N/A
N/A
N/A
30%
30%
30%

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
40%
40%
40%
AESS with









Diesel APU









Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
10%
10%
10%
AESS with









Battery APU









Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
10%
10%
10%
AESS with









Automatic









Stop-Start









Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
10%
10%
10%
AESS with









FOH









Transmission
Manual
0%
0%
0%
0%
0%
0%
0%
0%
0%
AMT
50%
50%
50%
50%
50%
50%
50%
50%
50%
Auto
20%
20%
20%
20%
20%
20%
20%
20%
20%
Dual Clutch
10%
10%
10%
10%
10%
10%
10%
10%
10%
Top Gear
50%
50%
50%
50%
50%
50%
50%
50%
50%
Direct Drive









Transmission
40%
40%
40%
40%
40%
40%
40%
40%
40%
Efficiency









Improvement









Neutral Idle
20%
20%
20%
20%
20%
20%
20%
20%
20%
Driveline
Axle
65%
65%
65%
65%
65%
65%
65%
65%
65%
Efficiency









Improvement









6x2, 6x4
N/A
N/A
N/A
25%
25%
25%
25%
25%
25%
Axle









Disconnect









or 4x2 Axle









Downspeed
3.31
3.31
3.31
3.31
3.31
3.31
3.26
3.26
3.26
(Rear Axle









Ratio)









Accessory Improvements
A/C
20%
20%
20%
20%
20%
20%
20%
20%
20%
Efficiency









Electric
20%
20%
20%
20%
20%
20%
20%
20%
20%
Access.









Other Technologies
Predictive
40%
40%
40%
40%
40%
40%
40%
40%
40%
Cruise









Control









Automated
25%
25%
25%
25%
25%
25%
25%
25%
25%
Tire Inflation









System









Tire Pressure
50%
50%
50%
50%
50%
50%
50%
50%
50%
Monitoring
System









Neutral
0%
0%
0%
0%
0%
0%
0%
0%
0%
Coast










-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-36 Technology Adoption Rates for Class 7 and 8 Tractors for Determining the 2027 MY Standards

CLASS 7
CLASS 8

Day Cab
Day Cab
Sleeper Cab

Low Roof
Mid Roof
High
Roof
Low Roof
Mid Roof
High
Roof
Low Roof
Mid Roof
High Roof
Engine

2027MY
11L
Engine
350 HP
2027MY
11L
Engine
350 HP
2027MY
11L
Engine
350 HP
2027MY
15L
Engine
455 HP
2027MY
15L
Engine
455 HP
2027MY
15L
Engine
455 HP
2027MY
15L
Engine
455 HP
2027MY
15L
Engine
455 HP
2027MY
15L
Engine
455 HP
Aerodynamics
Bin I
0%
0%
0%
0%
0%
0%
0%
0%
0%
Bin II
20%
20%
0%
20%
20%
0%
20%
20%
0%
Bin III
50%
50%
30%
50%
60%
30%
40%
50%
20%
Bin IV
30%
30%
30%
30%
20%
30%
40%
30%
30%
Bin V
0%
0%
40%
0%
0%
40%
0%
0%
50%
Bin VI
0%
0%
0%
0%
0%
0%
0%
0%
0%
Bin VII
0%
0%
0%
0%
0%
0%
0%
0%
0%
Steer Tires
Base
5%
5%
5%
5%
5%
5%
5%
5%
5%
Level 1
20%
20%
10%
20%
20%
10%
20%
20%
10%
Level 2
50%
50%
50%
50%
50%
50%
50%
50%
50%
Level 3
25%
25%
35%
25%
25%
35%
25%
25%
35%
Drive Tires
Base
5%
5%
5%
5%
5%
5%
5%
5%
5%
Level 1
10%
10%
10%
10%
10%
10%
10%
10%
10%
Level 2
85%
85%
50%
85%
85%
50%
85%
85%
50%
Level 3
0%
0%
35%
0%
0%
35%
0%
0%
35%
Idle Reduction
Tamper
Proof AESS
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Tamper
Proof AESS
with Diesel
APU
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Tamper
Proof AESS
with Battery
APU
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Tamper
Proof AESS
with
Automatic
Stop-Start
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Tamper
Proof AESS
withFOH
N/A
N/A
N/A
N/A
N/A
N/A
0%
0%
0%
Adjustable
AESS
N/A
N/A
N/A
N/A
N/A
N/A
15%
15%
15%

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
40%
40%
40%
AESS with









Diesel APU









Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
15%
15%
15%
AESS with









Battery APU









Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
15%
15%
15%
AESS with









Automatic









Stop-Start









Adjustable
N/A
N/A
N/A
N/A
N/A
N/A
15%
15%
15%
AESS with









FOH









Transmission
Manual
0%
0%
0%
0%
0%
0%
0%
0%
0%
AMT
50%
50%
50%
50%
50%
50%
50%
50%
50%
Auto
30%
30%
30%
30%
30%
30%
30%
30%
30%
Dual Clutch
10%
10%
10%
10%
10%
10%
10%
10%
10%
Top Gear
50%
50%
50%
50%
50%
50%
50%
50%
50%
Direct Drive









Transmission
70%
70%
70%
70%
70%
70%
70%
70%
70%
Efficiency









Improvement









Neutral Idle
30%
30%
30%
30%
30%
30%
30%
30%
30%
Driveline
Axle
80%
80%
80%
80%
80%
80%
80%
80%
80%
Efficiency









Improvement









6x2, 6x4
N/A
N/A
N/A
30%
30%
30%
30%
30%
30%
Axle









Disconnect









or 4x2 Axle









Downspeed
3.21
3.21
3.21
3.21
3.21
3.21
3.16
3.16
3.16
(Rear Axle









Ratio)









Accessory Improvements
A/C
30%
30%
30%
30%
30%
30%
30%
30%
30%
Efficiency









Electric
30%
30%
30%
30%
30%
30%
30%
30%
30%
Access.









Other Technologies
Predictive
40%
40%
40%
40%
40%
40%
40%
40%
40%
Cruise









Control









Automated
30%
30%
30%
30%
30%
30%
30%
30%
30%
Tire Inflation









System









Tire Pressure
70%
70%
70%
70%
70%
70%
70%
70%
70%
Monitoring
System









Neutral
0%
0%
0%
0%
0%
0%
0%
0%
0%
Coast










-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-37 includes the adoption rates of each technology used in setting the heavy-haul
tractor standards for 2021, 2024, and 2027 MY.
Table 2-37 Technology Adoption Rates for Heavy-Haul Tractors for Determining the 2021,2024, and 2027
MY Standards
HEAVY-HAUL TRACTOR APPLICATION RATES

2021MY
2024MY
2027MY
Engine
2021 MY 15L Engine with
600 HP with 2% reduction
over 2018 MY
2024 MY 15L Engine with
600 HP with 4.2%
reduction over 2018 MY
2027 MY 15L Engine with
600 HP with 5.4% reduction
over 2018 MY
Aerodynamics - 0%
Steer Tires
Phase 1 Baseline
15%
10%
5%
Level I
35%
30%
10%
Level 2
50%
60%
85%
Level 3
0%
0%
0%
Drive Tires
Phase 1 Baseline
15%
10%
5%
Level I
35%
30%
10%
Level 2
50%
60%
85%
Level 3
0%
0%
0%
Transmission
AMT
40%
50%
50%
Automatic with Neutral
Idle
10%
20%
20%
DCT
5%
10%
10%
Other Technologies
6x2 Axle
0%
0%
0%
Transmission Efficiency
20%
40%
70%
Axle Efficiency
30%
65%
80%
Predictive Cruise
Control
20%
40%
40%
Accessory
Improvements
10%
20%
20%
Air Conditioner
Efficiency
Improvements
10%
20%
20%
Automatic Tire Inflation
Systems
20%
25%
30%
Tire Pressure
Monitoring System
20%
50%
70%
The agencies are also adopting in Phase 2 provisions that allow the manufacturers to meet
an optional heavy Class 8 tractor standard that reflects both aerodynamic improvements, along
with the powertrain requirements that go along with higher GCWR. Table 2-38 reflects the
adoption rates for each of the technologies for each of the subcategories in MY 2021. The
technology packages closely reflect those in the primary Class 8 tractor program. The
exceptions include less aggressive targets for low rolling resistance tires, no 6x2 axle adoption
rates, and no downspeeding due to the heavier loads of these vehicles.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-38 GEM Inputs for 2021 MY Heavy Class 8 Tractor Standards
OPTIONAL HEAVY CLASS 8 TRACTOR APPLICATION RATES - 2021 MY

Low/Mid Roof Day
Cab
High Roof Day Cab
Low/Mid Roof Sleeper
Cab
High Roof Sleeper Cab
Engine
2021 MY 15L Engine
with 600 HP with 2%
reduction over 2018
MY
2021 MY 15L Engine
with 600 HP with 2%
reduction over 2018
MY
2021 MY 15L Engine
with 600 HP with 2%
reduction over 2018
MY
2021 MY 15L Engine
with 600 HP with 2%
reduction over 2018 MY
Aerodynamics
Bin I
10%
0%
10%
0%
Bin II
10%
0%
10%
0%
Bin III
70%
60%
70%
60%
Bin IV
10%
35%
10%
30%
Bin V
0%
5%
0%
10%
Bin VI
0%
0%
0%
0%
Bin VII
0%
0%
0%
0%
Steer Tires
Phase 1 Baseline
10%
5%
10%
5%
Level I
25%
35%
25%
35%
Level 2
65%
60%
65%
60%
Level 3
0%
0%
0%
0%
Drive Tires
Phase 1 Baseline
20%
10%
20%
10%
Level I
40%
30%
40%
30%
Level 2
40%
60%
40%
60%
Level 3
0%
0%
0%
0%
Transmission
AMT
40%
40%
40%
40%
Automatic with Neutral
Idle
10%
10%
10%
10%
DCT
5%
5%
5%
5%
Other Technologies
Adjustable AESS w/
Diesel APU
N/A
N/A
30%
30%
Adjustable AESS w/
Battery APU
N/A
N/A
10%
10%
Adjustable AESS w/
Automatic Stop-Start
N/A
N/A
10%
10%
Adjustable AESS w/
FOH Cold, Main Engine
Warm
N/A
N/A
10%
10%
Adjustable AESS
programmed to 5
minutes
N/A
N/A
40%
40%
Transmission Efficiency
20%
20%
20%
20%
Axle Efficiency
30%
30%
30%
30%
Predictive Cruise
Control
20%
20%
20%
20%
Accessory
Improvements
10%
10%
10%
10%

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Air Conditioner
Efficiency
Improvements
10%
10%
10%
10%
Automatic Tire Inflation
Systems
20%
20%
20%
20%
Tire Pressure
Monitoring System
20%
20%
20%
20%
2.8.4 Derivation of the Tractor Standards
The agencies used the technology effectiveness inputs and technology adoption rates to
develop GEM inputs to derive the HD Phase 2 fuel consumption and CO2 emissions standards
for each subcategory of Class 7 and 8 combination tractors. Note that we have analyzed one
technology pathway for each level of stringency, but tractor manufacturers are free to use any
combination of technology to meet the standards on average.
2.8.4.1 2021 through 2027 MY Engine Fuel Maps
One of the most significant changes in the HD Phase 2 version of GEM is the allowance
for manufacturers to enter their own engine fuel maps by following the test procedure described
in the Chapter 3 Test Procedure section of this RIA. The GEM engine fuel map input file
consists of information in csv format. It contains a steady-state engine fueling map that includes
three columns: engine speed in rpm, engine torque in Nm, and engine fueling rate in g/s. New
for the final Phase 2 rule, the input file also includes a cycle average fuel map represented by
engine cycle work, the cycle-average engine speed to vehicle speed ratio, and the fuel mass in
grams. The input file also contains the engine full torque or lug curve in two columns: engine
speed in rpm and torque in NM. The input file also contains the motoring torque and uses the
same format and units as the full load torque curve. The idle fuel map is also included.
The agencies developed default engine fuel maps for all tractor subcategories, utilizing
the same format that the manufacturers would be required to provide. Fuel maps were developed
for the 2021, 2024, and 2027 model years by applying the technologies assumed in deriving the
engine standards and the additional technology effectiveness of new engine platforms (for 2027)
to the 2018 baseline engine fuel maps. Those default maps are derived from multiple sources of
confidential business information from different stakeholders together with engineering
judgment. A list of all of the engine fuel maps used in setting the standards for each subcategory
is given in Table 2-39. The model years covered by the maps are 2021, 2024, and 2027 are
shown from Figure 2-30 to Figure 2-38. In lieu of using 2021, 2024, and 2027 MY fuel maps for
the 15L 600 HP engine used in heavy-haul tractor standards and optional 2021 MY Heavy Class
8 tractor standards, we used the 2018 MY fuel map shown in Figure 2-19. We then applied a 2
percent reduction in 2021 MY, a 4.2 percent reduction in 2024 MY, and a 5.4 percent reduction
in 2027 MY in the GEM runs to determine the stringency of the standards.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-39 GEM Default CI Engine Fuel Maps for Tractors
REGULATORY SUBCATEGORY
ENGINE FUEL MAP
Class 8 Combination
Sleeper Cab - High Roof
15L-455 HP
Class 8 Combination
Sleeper Cab - Mid Roof
15L-455 HP
Class 8 Combination
Sleeper Cab - Low Roof
15L-455 HP
Class 8 Combination
Day Cab - High Roof
15L-455 HP
Class 8 Combination
Day Cab - Mid Roof
15L-455 HP
Class 8 Combination
Day Cab - Low Roof
15L-455 HP
Class 7 Combination
Day Cab - High Roof
11L - 350 HP
Class 7 Combination
Day Cab - Mid Roof
11L - 350 HP
Class 7 Combination
Day Cab - Low Roof
11L - 350 HP
Heavy Haul
Heavy-Haul and Heavy
Class 8 Tractors
15L - 600 HP
In vehicle applications, considering that market penetration of WHR would be different
between sleeper cab (SC) and day cab (DC) engines due to the nature of their different driving
cycles, the emission reductions should be different, and therefore, the engine fuel maps used in
GEM can be different as well. In addition, at least one new engine platform would be taken into
consideration, which means that more aggressive technology effectiveness is included in the
tractor vehicles in addition to higher market penetration of WHR. See Chapter 2.7.5 above.
As discussed in Section III.D(l)(b)(i) of the FRM Preamble, the agencies project that at
least one engine manufacturer (and possibly more) will have completed a redesign for tractor
engines by 2027. Accordingly, we project that 50 percent of tractor engines in 2027 will be
redesigned engines and be 1.6 percent more efficient than required by the engine standards, so
the average engine would be 0.8 percent better.145 However, we could have projected the same
overall improvement by projecting 25 percent of engine get 3.2 percent better. Based on the CBI
information available to us, we believe projecting a 0.8 percent improvement is somewhat
conservative.
Adding this 0.8 percent improvement to the 5.1 reduction required by the standards
means we project the average 2027 tractor engine would be 5.9 percent better than Phase 1.
Because engine improvements for tractors are applied separately for day cabs and sleeper cabs in
the vehicle program, we estimated separate improvements for them here. Specifically, we
project a 5.4 percent reduction for day cabs and a 6.4 percent reduction in fuel consumption in
sleeper cabs beyond Phase 1. It is important to also note that manufacturers that do not achieve
this level would be able to make up for the difference by applying one of the many other tractor
technologies to a greater extent than we project, or to achieve greater reductions by optimizing
technology efficiency further. We are not including the cost of developing these new engines in
our cost analysis because we believe these engines are going to be developed due to market
forces (i.e., the new platform, already contemplated) rather than due to this rulemaking.
The default fuel maps are created for use in GEM. As just explained, use of different
WHR market penetration rates between sleeper cabs and day cabs results in unique fuel maps for
each.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Figure 2-30 to Figure 2-38 show all the engine fuel maps used in GEM for years 2021 to
2027, for sleeper cab and day cab vehicles with 455hp rating engines and 350hp rating engines.
2021 Engine 455hp / 15L BSFC ( g / kW * hr)
2000
1800
1600
1400
1200
1000
800
600
400
200
0™
600
800
1000 1200 1400 1600 1800 2000 2200
Speed ( RPM )
Figure 2-30 2021 Engine Fuel Map with 455hp Rating Used For Sleeper Cab
2021 Engine 455hp / 15L BSFC ( g / kW * hr)
2000
00
1800
1600
1400
z 1200
S 1000
800
600
400
200
600
800 1 000 1200 1400 1600 1 800 2000 2200
Speed ( RPM )
Figure 2-31 2021 Engine Fuel Map with 455hp Rating Used For Day Cab

-------
E. O. 12866 Review - Revised - Do Not Cite, Quote, or Release During Review
2021 Engine 350hp / 11L BSFC ( g / kW * hr)
1600
CO
1400
1200
1000
800

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
2024 Engine 455hp / 15L BSFC ( g / kW * hr)
2000
1800
1600
1400
E
z
1200
0)
= 1000
b
i—
800
600
200
210
225
400
225
200
600
800
1000 1200 1400 1600 1800 2000 2200
Speed ( RPM )
Figure 2-34 2024 Engine Fuel Map with 455hp Rating Used For Day Cab
2024 Engine 350hp/11L BSFC ( g / kW * hr)
1600
1400
1200
1000
E
z
800
0)
3
0-
b
1—
600
400
--25!
200
600
800 1000 1200 1400 1600 1800 2000 2200
Speed ( RPM )
Figure 2-35 2024 Engine Fuel Map with 350hp Rating Used For Class 7 Tractor

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
2027 Engine 455hp / 15L BSFC ( g / kW * hr)
2000
1800
1600
1400
Z 1200
.0=
1000
800
600
-\95
205
220
400
220
200
600
800
1000 1200 1400 1600 1800 2000 2200
Speed ( RPM)
Figure 2-36 2027 Engine Fuel Map with 455hp Rating Used For Sleeper Cab
2027 Engine 455hp / 15L BSFC ( g / k\/V * hr)
2000
600
400
E
z
200
(D
g- 1000
b
H 800
600
195
205
-225
400
225
200
600
800
1000 1200 1400 1600
Speed ( RPM )
2000 2200
Figure 2-37 2027 Engine Fuel Map with 455hp Rating Used For Day Cab

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
2027 Engine 350hp/ 11L BSFC ( g / kW * hr)
1600
1400
1200
1000
E
z
§ 800
w
o
i—
600
400
230
200
600 800 1000 1200 1400 1600 1 800 2000 2200
Speed ( RPM )
Figure 2-38 2027 Engine Fuel Map with 350hp Rating Used For Class 7 Tractor
2.8.4.2 GEM Inputs Used in Setting the Tractor Standards
As such, the agencies derived a standard for each subcategory by weighting the
individual GEM input parameters included in Table 2-30 with the adoption rates in Table 2-34,

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-35, and Table 2-36. For example, the CdA value for a 2021MY Class 8 Sleeper
Cab High Roof scenario case was derived as 60 percent times 5.95 plus 30 percent times 5.40
plus 10 percent times 4.90, which is equal to a CdA of 5.68 m2 Similar calculations were made
for tire rolling resistance, transmission types, idle reduction, and other technologies. To account
for the engine standards and engine technologies, the agencies developed engine fuel maps for
GEM, as described in the section above.J The agencies then ran GEM with a single set of
vehicle inputs, as shown in Table 2-40, to derive the standards for each subcategory.
Table 2-40 GEM Inputs for the 2021MY Class 7 and 8 Tractor Standard Setting
CLASS 7
CLASS 8
Day Cab
Day Cab
Sleeper Cab
Low
Mid
High Roof
Low Roof
Mid
High Roof
Low Roof
Mid
High
Roof
Roof


Roof


Roof
Roof
Engine
2021MY
2021MY
2021MY
2021MY
2021MY
2021MY
2021MY
2021MY
2021MY
11L
11L
11L
15L
15L
15L
15L
15L
15L
Engine
Engine
Engine
Engine
Engine
Engine
Engine
Engine
Engine
350 HP
350 HP
350 HP
455 HP
455 HP
455 HP
455 HP
455 HP
455 HP
Aerodynamics (CdA in m2)
5.24
6.33
6.01
5.24
6.33
6.01
5.24
6.33
5.68
Steer Tires (CRR in kg/metric ton)
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
Drive Tires (CRR in kg/metric ton)
6.6
6.6
6.3
6.6
6.6
6.3
6.6
6.6
6.3
Extended Idle Reduction Weighted Effectiveness
N/A
N/A
N/A
N/A
N/A
N/A
2.3%
2.3%
2.3%


Transmission = 10 speed Manual Transmission



Gear Ratios = 12.8, 9.25, 6.76, 4.90, 3.58, 2.61, 1.89, 1.38, 1.00, 0.73

Drive Axle Ratio = 3.36 for day cabs, 3.31 for sleeper cabs
6x2 Axle Weighted Effectiveness
N/A
N/A
N/A
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%


Transmission Type Weighted Effectiveness =
1.1%


Neutral Idle Weighted Effectiveness
0.1%
0.1%
0.1%
0.1%
0.1%
0.1%
0.02%
0.02%
0.02%
Direct Drive Weighted Effectiveness = 0.4%
Transmission Efficiency Weighted Effectiveness = 0.2%
Axle Efficiency Improvement = 0.6%


Air Conditioner Efficiency Improvements =
0.1%


Accessory Improvements = 0.1%
Predictive Cruise Control =0.4%
Automatic Tire Inflation Systems = 0.3%
Tire Pressure Monitoring System = 0.2%
Phase 1 Credit Carry-over = 1%
1 See RIA Chapter 2.7 explaining the derivation of the engine standards.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-41 GEM Inputs for the 2024MY Class 7 and 8 Tractor Standard Setting
CLASS 7
CLASS 8
Day Cab
Day Cab
Sleeper Cab
Low
Mid
High Roof
Low Roof
Mid
High Roof
Low Roof
Mid
High
Roof
Roof


Roof


Roof
Roof
Engine
2024MY
2024MY
2024MY
2024MY
2024MY
2024MY
2024MY
2024MY
2024MY
11L
11L
11L
15L
15L
15L
15L
15L
15L
Engine
Engine
Engine
Engine
Engine
Engine
Engine
Engine
Engine
350 HP
350 HP
350 HP
455 HP
455 HP
455 HP
455 HP
455 HP
455 HP
Aerodynamics (CdA in m2)
5.16
6.25
5.82
5.16
6.25
5.82
5.16
6.25
5.52
Steer Tires (CRR in kg/metric ton)
5.9
5.9
5.8
5.9
5.9
5.8
5.9
5.9
5.8
Drive Tires (CRR in kg/metric ton)
6.4
6.4
6.0
6.4
6.4
6.0
6.4
6.4
6.0
Extended Idle Reduction Weighted Effectiveness
N/A
N/A
N/A
N/A
N/A
N/A
2.5%
2.5%
2.5%


Transmission = 10 speed Manual Transmission



Gear Ratios = 12.8, 9.25, 6.76, 4.90, 3.58, 2.61, 1.89, 1.38, 1.00, 0.73

Drive Axle Ratio = 3.31 for day cabs, 3.26 for sleeper cabs
6x2 Axle Weighted Effectiveness
N/A
N/A
N/A
0.5%
0.5%
0.5%
0.5%
0.5%
0.5%


Transmission Type Weighted Effectiveness =
1.6%


Neutral Idle Weighted Effectiveness
0.2%
0.2%
0.2%
0.2%
0.2%
0.2%
0.03%
0.03%
0.03%
Direct Drive Weighted Effectiveness = 1.0%
Transmission Efficiency Weighted Effectiveness = 0.4%
Axle Efficiency Improvement =1.3%


Air Conditioner Efficiency Improvements =
0.1%


Accessory Improvements = 0.2%
Predictive Cruise Control =0.8%
Automatic Tire Inflation Systems = 0.3%
Tire Pressure Monitoring System = 0.5%

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-42 GEM Inputs for the 2027MY Class 7 and 8 Tractor Standard Setting
CLASS 7
CLASS 8
Day Cab
Day Cab
Sleeper Cab
Low
Mid
High Roof
Low Roof
Mid
High Roof
Low Roof
Mid
High
Roof
Roof


Roof


Roof
Roof
Engine
2027MY
2027MY
2027MY
2027MY
2027MY
2027MY
2027MY
2027MY
2027MY
11L
11L
11L
15L
15L
15L
15L
15L
15L
Engine
Engine
Engine
Engine
Engine
Engine
Engine
Engine
Engine
350 HP
350 HP
350 HP
455 HP
455 HP
455 HP
455 HP
455 HP
455 HP
Aerodynamics (CdA in m2)
5.12
6.21
5.67
5.12
6.21
5.67
5.08
6.21
5.26
Steer Tires (CRR in kg/metric ton)
5.8
5.8
5.6
5.8
5.8
5.6
5.8
5.8
5.6
Drive Tires (CRR in kg/metric ton)
6.2
6.2
5.8
6.2
6.2
5.8
6.2
6.2
5.8
Extended Idle Reduction Weighted Effectiveness
N/A
N/A
N/A
N/A
N/A
N/A
3%
3%
3%


Transmission = 10 speed Manual Transmission



Gear Ratios = 12.8, 9.25, 6.76, 4.90, 3.58, 2.61, 1.89, 1.38, 1.00, 0.73

Drive Axle Ratio = 3.21 for day cabs, 3.16 for sleeper cabs
6x2 Axle Weighted Effectiveness
N/A
N/A
N/A
0.6%
0.6%
0.6%
0.6%
0.6%
0.6%


Transmission Type Weighted Effectiveness =
1.6%


Neutral Idle Weighted Effectiveness
0.2%
0.2%
0.2%
0.2%
0.2%
0.2%
0.03%
0.03%
0.03%
Direct Drive Weighted Effectiveness = 1.0%
Transmission Efficiency Weighted Effectiveness = 0.7%
Axle Efficiency Improvement = 1.6%


Air Conditioner Efficiency Improvements =
0.3%


Accessory Improvements = 0.2%
Predictive Cruise Control =0.8%
Automatic Tire Inflation Systems = 0.4%
Tire Pressure Monitoring System = 0.7%

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-43 GEM Inputs for the 2021,2024, and 2027MY Heavy-Haul Tractor Standard Setting
2021MY
2024MY
2027MY
Engine = 2021 MY 15L
Engine with 600 HP
Engine = 2024 MY 15L
Engine with 600 HP
Engine = 2027 MY 15L
Engine with 600 HP
Aerodynamics (CdA in m2) = 5.00
Steer Tires (CRR in
kg/metric ton) = 6.2
Steer Tires (CRR in
kg/metric ton) = 6.0
Steer Tires (CRR in
kg/metric ton) = 5.8
Drive Tires (CRR in
kg/metric ton) = 6.6
Drive Tires (CRR in
kg/metric ton) = 6.4
Drive Tires (CRR in
kg/metric ton) = 6.2
Transmission = 18 speed
Manual Transmission
Transmission = 18 speed
Manual Transmission
Transmission = 18 speed
Manual Transmission
Drive axle Ratio = 3.70
Drive axle Ratio = 3.70
Drive axle Ratio = 3.70
6x2 Axle Weighted
Effectiveness = 0%
6x2 Axle Weighted
Effectiveness = 0%
6x2 Axle Weighted
Effectiveness = 0%
Transmission benefit = 1.1%
Transmission benefit =
1.8%
Transmission benefit = 1.8%
Transmission
Efficiency=0.2%
Transmission
Efficiency=0.4%
Transmission
Efficiency=0.7%
Axle Efficiency=0.3%
Axle Efficiency=0.7%
Axle Efficiency=1.6%
Predictive Cruise
Control=0.4%
Predictive Cruise Control
=0.8%
Predictive Cruise Control
=0.8%
Accessory Improvements =
0.1%
Accessory Improvements =
0.2%
Accessory Improvements =
0.3%
Air Conditioner Efficiency
Improvements= 0.1%
Air Conditioner Efficiency
Improvements = 0.1%
Air Conditioner Efficiency
Improvements = 0.2%
Automatic Tire Inflation
Systems = 0.3%
Automatic Tire Inflation
Systems = 0.3%
Automatic Tire Inflation
Systems = 0.4%
Tire Pressure Monitoring
System= 0.2%
Tire Pressure Monitoring
System= 0.5%
Tire Pressure Monitoring
System= 0.7%
The agencies ran GEM with a single set of vehicle inputs, as shown in Table 2-44, to
derive the optional standards for each subcategory of the Heavy Class 8 tractors.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-44 GEM Inputs for 2021 MY Heavy Class 8 Tractor Standards
HEAVY CLASS 8 GEM INPUTS FOR 2021 MY
Day Cab
Sleeper Cab
Low Roof
Mid Roof
High Roof
Low Roof
Mid Roof
High Roof
2021MY 15L Engine 600 HP
Aerodynamics (CdA in m2)
5.2
6.3
6.0
5.2
6.3
5.7
Steer Tires (CRR in kg/metric ton)
6.1
6.1
6.1
6.1
6.1
6.1
Drive Tires (CRR in kg/metric ton)
6.S
6.S
6.5
6.S
6.S
6.5
Extended Idle Reduction Weighted Effectiveness
N/A
N/A
N/A
2.3%
2.3%
2.3%
Transmission = 18 speed Manual Transmission
Drive Axle Ratio = 3.73
Transmission Type Weighted Effectiveness = 1.1%
Neutral Idle Weighted Effectiveness
0.1%
0.1%
0.1%
0.1%
0.1%
0.1%
Direct Drive Weighted Effectiveness = 0.4%
Transmission Efficiency Weighted Effectiveness = 0.2%
Axle Efficiency Improvement = 0.6%
Air Conditioner Efficiency Improvements = 0.1%
Accessory Improvements = 0.1%
Predictive Cruise Control =0.4%
Automatic Tire Inflation Systems = 0.3%
Tire Pressure Monitoring System = 0.2%

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The levels of the 2021, 2024, and 2027 model year standards for each subcategory are
included in Table 2-45.
Table 2-45 2021,2024, and 2027 Model Year Tractor Standards
2021 MODEL YEAR C02 GRAMS PER TON-MILE

Day Cab
Sleeper Cab
Heavy-Haul

Class 7
Class 8
Class 8
Class 8
Low Roof
105.5
80.5
72.3
52.4
Mid Roof
113.2
85.4
78.0
High Roof
113.5
85.6
75.7
2021 Model Year Gallons of Fuel per 1,000 Ton-Mile

Day Cab
Sleeper Cab
Heavy-Haul

Class 7
Class 8
Class 8
Class 8
Low Roof
10.36346
7.90766
7.10216
5.14735
Mid Roof
11.11984
8.38900
7.66208
High Roof
11.14931
8.40864
7.43615
2024 Model Year CO2 Grams per r
"on-Mile

Day Cab
Sleeper Cab
Heavy-Haul

Class 7
Class 8
Class 8
Class 8
Low Roof
99.8
76.2
68.0
50.2
Mid Roof
107.1
80.9
73.5
High Roof
106.6
80.4
70.7
2024 Model Year and Later Gallons of Fuel per 1,000 Ton-Mile

Day Cab
Sleeper Cab
Heavy-Haul

Class 7
Class 8
Class 8
Class 8
Low Roof
9.80354
7.48527
6.67976
4.93124
Mid Roof
10.52063
7.94695
7.22004
High Roof
10.47151
7.89784
6.94499
2027 Model Year CO2 Grams per r
"on-Milea

Day Cab
Sleeper Cab
Heavy-Haul

Class 7
Class 8
Class 8
Class 8
Low Roof
96.2
73.4
64.1
48.3
Mid Roof
103.4
78.0
69.6
High Roof
100.0
75.7
64.3
2027 Model Year and Later Gallons of Fuel per 1,000 Ton-Mile

Day Cab
Sleeper Cab
Heavy-Haul

Class 7
Class 8
Class 8
Class 8
Low Roof
9.44990
7.21022
6.29666
4.74460
Mid Roof
10.15717
7.66208
6.83694
High Roof
9.82318
7.43615
6.31631
The 2027 MY standards for the high roof day cabs and high roof sleeper cab include the
0.3 m2 reduction in CdA built into GEM to reflect a change in the standard trailer (see Preamble
Section III.E.2.a.viii). This change lowers the numerical value of the standard, but does not
impact the stringency (i.e., the effectiveness of the technology packages that need to be installed
on a tractor to be compliant with the standards). Therefore, the percent reductions reported

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
throughout the Preamble to the final rule reflect only the effectiveness of the technology package
needed to meet the standard and does not include the change in CdA built into GEM. See Table
2-46 for the percent reduction calculations for high roof tractors in MY 2027.
Table 2-46 Percent Reductions for 2027MY High Roof Tractors

CLASS 7
HIGH ROOF
TRACTOR
CLASS 8 HIGH
ROOF DAY
CAB
CLASS 8 HIGH
ROOF SLEEPER
CAB
Baseline GEM Output
2018 MY (g/ton-mile)
129.7
98.2
87.8
2027 MY GEM Output
with 0.3 m2 CdA (g/ton-
mile)
100.0
75.7
64.3
2027 MY GEM Output
without 0.3 m2 CdA
(g/ton-mile)
102.0
77.0
65.7
% Reduction in Stringency
due to Technology
Package Only
21%
22%
25%
The level of the Phase 2 2021 model year optional Heavy Class 8 standards for each
subcategory is included in Table 2-47.
Table 2-47 Phase 2 Optional Heavy Class 8 Standards
OPTIO]
\AL HEAVY CLASS 8 TRACTOR STANDARDS
Low Roof Day
Cab
Mid Roof
Day Cab
High Roof
Day Cab
Low Roof
Sleeper Cab
Mid Roof
Sleeper Cab
High Roof
Sleeper Cab
2021 Model Year CO2 Standards (Grams per Ton-Mile)
51.8
54.1
54.1
45.3
47.9
46.9
2021 MY and Later Fuel Consum
ption (gallons of Fuel per 1,000 Ton-Mile)
5.08841 5.31434 5.31434
4.44990 4.70530 4.60707
2.8.5 Tractor Package Costs of the Standards
A summary of the technology package costs under the final standard and relative to the
flat baseline is included in Table 2-48 through Table 2-51 for MYs 2021, 2024, and 2027,
respectively. RIA Chapter 2.11 includes the technology costs for the individual technologies.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-48 Class 7 and 8 Tractor Technology Incremental Costs in the 2021 Model Yeara b
Final Standards vs. the Flat Baseline (2013$ per vehicle)

CLASS 7
CLASS 8

Day Cab
Day Cab
Sleeper Cab

Low/Mid
High
Low/ Mid
High
Low
Mid
High

Roof
Roof
Roof
Roof
Roof
Roof
Roof
Engine0
$284
$284
$284
$284
$284
$284
$284
Aerodynamics
$164
$299
$164
$299
$119
$119
$349
Tires
$39
$9
$61
$16
$61
$56
$16
Tire inflation







system
$259
$259
$300
$300
$300
$300
$300
Transmission
$4,096
$4,096
$4,096
$4,096
$4,096
$4,096
$4,096
Axle & axle







lubes
$71
$71
$101
$101
$101
$101
$101
Idle reduction







with APU
$0
$0
$0
$0
$1,998
$1,998
$1,909
Air conditioning
$17
$17
$17
$17
$17
$17
$17
Other vehicle







technologies
$204
$204
$204
$204
$204
$204
$204
Total
$5,134
$5,240
$5,228
$5,317
$7,181
$7,175
$7,276
Notes:
a Costs shown are for the 2021 model year and are incremental to the costs of a baseline Phase 2 tractor. These costs include
indirect costs via markups along with learning impacts. For a description of the markups and learning impacts considered in this
analysis and how it impacts technology costs for other years, refer to Chapter 2 of the RIA (see RIA 2.11).
b Note that values in this table include projected technology penetration rates. Therefore, the technology costs shown reflect the
average cost expected for each of the indicated tractor classes. To see the actual estimated technology costs exclusive of
adoption rates, refer to Chapter 2.11 of this RIA.
c Engine costs are for a heavy HD diesel engine meant for a combination tractor (see Table 2-14).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-49 Class 7 and 8 Tractor Technology Incremental Costs in the 2024 Model Yeara b
Final Standards vs. the Flat Baseline (2013$ per vehicle)

CLASS 7
CLASS 8

Day Cab
Day Cab
Sleeper Cab

Low/Mid
High
Low/ Mid
High
Low
Mid
High

Roof
Roof
Roof
Roof
Roof
Roof
Roof
Engine0
$712
$712
$712
$712
$712
$712
$712
Aerodynamics
$264
$465
$264
$465
$217
$217
$467
Tires
$40
$12
$65
$20
$65
$65
$20
Tire inflation







system
$383
$383
$477
$477
$477
$477
$477
Transmission
$6,092
$6,092
$6,092
$6,092
$6,092
$6,092
$6,092
Axle & axle







lubes
$139
$139
$185
$185
$185
$185
$185
Idle reduction







with APU
$0
$0
$0
$0
$2,946
$2,946
$2,946
Air conditioning
$32
$32
$32
$32
$32
$32
$32
Other vehicle







technologies
$374
$374
$374
$374
$374
$374
$374
Total
$8,037
$8,210
$8,201
$8,358
$11,100
$11,100
$11,306
Notes:
a Costs shown are for the 2021 model year and are incremental to the costs of a baseline Phase 2 tractor. These costs include
indirect costs via markups along with learning impacts. For a description of the markups and learning impacts considered in this
analysis and how it impacts technology costs for other years, refer to Chapter 2.11 of this RIA.
b Note that values in this table include projected technology penetration rates. Therefore, the technology costs shown reflect the
average cost expected for each of the indicated tractor classes. To see the actual estimated technology costs exclusive of
adoption rates, refer to Chapter 2.11 of this RIA.
c Engine costs are for a heavy HD diesel engine meant for a combination tractor (see Table 2-18).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-50 Class 7 and 8 Tractor Technology Incremental Costs in the 2027 Model Yeara b
Final Standards vs. the Flat Baseline (2013$ per vehicle)

CLASS 7
CLASS 8

Day Cab
Day Cab
Sleeper Cab

Low/Mid
High
Low/ Mid
High
Low
Mid
High

Roof
Roof
Roof
Roof
Roof
Roof
Roof
Engine0
$1,579
$1,579
$1,579
$1,579
$1,579
$1,579
$1,579
Aerodynamics
$453
$547
$453
$547
$415
$415
$639
Tires
$43
$12
$70
$20
$70
$70
$20
Tire inflation







system
$469
$469
$594
$594
$594
$594
$594
Transmission
$7,098
$7,098
$7,098
$7,098
$7,098
$7,098
$7,098
Axle & axle







lubes
$168
$168
$220
$220
$220
$220
$220
Idle reduction







with APU
$0
$0
$0
$0
$3,134
$3,173
$3,173
Air conditioning
$45
$45
$45
$45
$45
$45
$45
Other vehicle







technologies
$380
$380
$380
$380
$380
$380
$380
Total
$10,235
$10,298
$10,439
$10,483
$13,535
$13,574
$13,749
Notes:
a Costs shown are for the 2021 model year and are incremental to the costs of a baseline Phase 2 tractor. These costs include
indirect costs via markups along with learning impacts. For a description of the markups and learning impacts considered in this
analysis and how it impacts technology costs for other years, refer to Chapter 2.11 of this RIA.
b Note that values in this table include projected technology penetration rates. Therefore, the technology costs shown reflect the
average cost expected for each of the indicated tractor classes. To see the actual estimated technology costs exclusive of
adoption rates, refer to Chapter 2.11 of this RIA.
c Engine costs are for a heavy HD diesel engine meant for a combination tractor (see Table 2-22).
Table 2-51 Heavy-Haul Tractor Technology Incremental Costs in the 2021,2024, and 2027 Model Yeara b
Final Standards vs. the Less Dynamic Baseline (2013$ per vehicle)

2021 MY
2024 MY
2027 MY
Engine0
$284
$712
$1,579
Tires
$61
$65
$70
Tire inflation system
$300
$477
$594
Transmission
$4,096
$6,092
$7,098
Axle Efficiency
$101
$185
$220
Air conditioning
$17
$32
$45
Other vehicle technologies
$204
$374
$380
Total
$5,063
$7,937
$9,986
Notes:
a Costs shown are for the specified model year and are incremental to the costs of a baseline
Phase 2 tractor. These costs include indirect costs via markups along with learning impacts.
For a description of the markups and learning impacts considered in this analysis and how it
impacts technology costs for other years, refer to Chapter 2 of the RIA (see RIA 2.11).
b Note that values in this table include projected technology penetration rates. Therefore,
the technology costs shown reflect the average cost expected for each of the indicated
tractor classes. To see the actual estimated technology costs exclusive of adoption rates,
refer to Chapter 2 of the RIA (see RIA 2.11 in particular).
c Engine costs are for a heavy HD diesel engine meant for a combination tractor.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.9 Technology Application and Estimated Costs - Vocational Vehicles
This section describes the technical analysis supporting the derivation of the vocational
vehicle standards, including technology effectiveness and adoption rates. For purposes of setting
standards, the agencies have established a unique baseline vocational vehicle configuration for
each of the vocational vehicle regulatory subcategories, including nine diesel subcategories, nine
gasoline subcategories, and seven custom chassis subcategories. For purposes of demonstrating
compliance, some of the attributes and parameters are fixed by the agencies and are not available
as manufacturer inputs to GEM, while some are available to manufacturers when identifying
configurations to certify in the model years of the HD Phase 2 program.
2.9.1 Vocational Engines
This section describes the engines the agencies selected to incorporate into the baseline
vehicle configurations for the vocational vehicle subcategories, and how we used the GEM tool
to establish performance levels of these baseline vehicles. The agencies have developed models
for engines that represent performance of the technologies we expect would be installed in
vocational vehicles in the baseline year of 2017. A description of the technologies applied to our
2017 diesel engine models can be found above in Chapter 2.7 of this RIA, and gasoline engine
technologies are described in RIA Chapter 2.6. A description of the GEM engine simulation can
be found in RIA Chapter 4.
One of the most significant changes in the HD Phase 2 version of GEM is the provision
for manufacturers to enter their own engine fuel maps by following the test procedure described
in the RIA Chapter 3. The GEM engine fuel map input file consists of information in csv format.
It contains a steady-state engine fueling map that includes three columns: engine speed in rpm,
engine torque in Nm, and engine fueling rate in g/s. New for the final Phase 2 rules, the input
file also includes a cycle average fuel map represented by engine cycle work, the cycle-average
engine speed to vehicle speed ratio, and the fuel mass in grams. The input file also contains the
engine full torque or lug curve in two columns: engine speed in rpm and torque in NM. The
input file also contains the motoring torque and uses the same format and units as the full load
torque curve. The idle fuel map is also included.
2.9.1.1 Baseline Vocational Engines
The agencies have developed the vehicle standards using engine fuel maps described in
this section for all vocational vehicle sub-categories, utilizing the same format that the OEMs
will be required to provide when demonstrating compliance. Four sets of diesel engine maps
cover the nine primary diesel vocational vehicle regulatory subcategories and the seven custom
chassis subcategories, and one gasoline engine map covers the six gasoline vocational vehicle
regulatory subcategories, as summarized in Table 2-52. This means that some of the
subcategories share the same engine fuel map (and appropriately so; the agencies anticipate
common use of these engine platforms in real world application; see Chapter 2.7.5 above). For
example, all MHD diesel subcategories are powered by the same 7L engine with 270 hp rating,
as this is a very popular rating for engines in class 6-7 vocational vehicles in the U.S.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The agencies selected the 15L as the primary engine for the Regional HHD subcategory
because these vocational vehicles often require a similar level of power as a day cab tractor.
Also, the same engine hardware is often used for both tractor and vocational vehicles. It would
not be cost effective to develop two complete engines from one manufacturer in order to meet
two different market needs. The same principle is applied to 11L engines. We have made
changes to this 11L engine since proposal, from a 345hp to a 350hp rating for the HHD
subcategories. As proposed, the engine displacements and power ratings for the diesel MHD and
LHD vocational subcategories are the same as those simulated in GEM for Phase 1. More
details about the comments received on vocational engines and our responses with respect to
selection of baseline engines can be found in the Preamble at Section V.C and in the RTC
Section 6.
Table 2-52 GEM Engines for Vocational Vehicles
REGULATORY SUBCATEGORY AND DUTY CYCLE
ENGINE FUEL MAP
CI Heavy Heavy-Duty (Class 8)
Regional and Multipurpose Duty Cycles
15L - 455 HP
CI Heavy Heavy-Duty (Class 8)
Regional, Multi-Purpose, and Urban
Duty Cycles
11L- 350HP
CI Medium Heavy-Duty (Class 6-
7)
Regional, Multi-Purpose, and Urban
Duty Cycles
7L - 270 HP
CI Light Heavy-Duty (Class 2b-5)
Regional, Multi-Purpose, and Urban
Duty Cycles
7L - 200 HP
SI Heavy-Duty (Class 2b-8)
Regional, Multi-Purpose, and Urban
Duty Cycles
6.8L - 300 HP
Working with SwRI, the agencies have developed a baseline fuel map for an SI engine
intended for vocational vehicles. Based on testing at SwRI from a 2015 Ford 6.8L gasoline
engine, two key technologies are introduced to develop this baseline engine: cam phasing and
cooled EGR through a comprehensive engine modeling using GT-Power. It is recognized that it
would be very challenging to develop a map that can exactly match the proposed standards of
627 g/hp-hr numerically with the engine modeling approach taken. Consequently, the small
adjustment would have to be taken in order to match 627 g/hp-hr exactly. This can be done by
taking the ratio of whatever value obtained from modeling to 627g/hp-hr, and multiplying it to
the entire map if the final numerical values derived from GT-Power engine modeling is different
from the standards. More detailed process of this map development can be seen in Chapter 5.4
of the SwRI report146. It should be pointed out that this technology path is just one of many other
potential road maps that can achieve the standards. We believe this reasonably represents a
gasoline engine that complies with the applicable MY 2016 engine standard as shown in Figure
2-39.146

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Figure 2-39 Gasoline Engine Fuel Map for 300hp Rating
Vocational diesel baseline engine maps for MY 2018 are presented in Chapter 2.7 above.
Specifically, see Figure 2-17 to see the map of the 350 hp engine, Figure 2-18 for a map of the
455 hp engine, Figure 2-20 for a map of the 270 hp engine, and Figure 2-21 for a map of the 200
hp engine.
2.9.1.2 Improved Vocational Engines for Phase 2 Standard-Setting
The agencies developed four model year versions of these engine maps for each of these
four diesel engines: one set for MY 2017 as the baseline, a set of maps for MY2021, a set for
MY2024, and a set for MY 2027, as improved over the 2017 baseline engine maps.
2.9.1.2.1 Vocational Gasoline Engine Technology for Standard-Setting
Although the agencies will retain the Phase 1 SI separate engine standard for all
implementation years of Phase 2, we developed the Phase 2 standards for vocational vehicles
powered by SI engines, in part, to reflect performance of additional engine technology.K When
developing improvement levels for the stringency of the MY 2021, MY 2024, and MY 2027
vehicle standards, the agencies analyzed adoption rates, effectiveness, and cost of cylinder
deactivation and SI engine technologies that reduce friction. Consistent with our projection of
adoption rates of advanced engine friction reduction on HD gasoline pickup trucks, the agencies
projected that 44 percent of SI engines intended for vocational vehicles would already have
technologies applied that achieve performance equivalent to Level 2 engine friction reduction,
enabling a projected adoption rate of 56 percent of SI vocational engines that could upgrade to
Level 2. In terms of effectiveness, the agencies relied on the data presented in the Joint
Technical Support Document (TSD) published in support of the LD GHG final rulemaking.147
K The agencies did so in part in response to comments indicating that improvements in SI engine performance over
the baseline were feasible.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
In Chapter 3 of that document, the agencies present effectiveness values for upgrading from
baseline levels of engine friction reduction to Level 2 (EFR2) as ranging from 3.4 percent to 4.8
percent, for a range of LD vehicle types, and with large trucks falling in the middle of this range.
The TSD describes example technologies as including low-tension piston rings, piston skirt
design, roller cam followers, improved crankshaft design and bearings, material coatings,
material substitution, more optimal thermal management, and piston and cylinder surface
treatments. For this Phase 2 HD rulemaking, the agencies derived incremental EFR2
effectiveness values from the combined EFR1+EFR2 values that were relative to baseline-level
friction reduction. We were able to do this because the TSD also presented incremental
improvements for upgrading from EFR1 to EFR2 as ranging from 0.83 to 1.37. Using the same
reasoning as explained at proposal, the effectiveness and adoption rate of Level 2 engine friction
reduction is estimated to yield a fuel efficiency improvement of 0.6 percent.
Cylinder deactivation is considered as a technology in the HD pickup and van program,
and it can be an effective technology for vocational vehicles with high power to vehicle weight
ratios in driving conditions that don't demand full load operation. Table VI-6 in Preamble
Section VI shows that expected improvements in fuel consumption due to application of cylinder
deactivation on HD pickups and vans are on the order of two to three percent over the applicable
chassis dynamometer test cycle. The discussion in Section VI.E. 8 of the Preamble explains the
reasoning behind the agencies' decision to predicate the HD SI pickup standards on 56 percent
adoption of cylinder deactivation. Because of differences in offerings between engines sold in
complete pickup trucks and those sold in vocational vehicles, we are applying only 30 percent
adoption of cylinder deactivation for SI vocational vehicle-level improvements. Because of
differences in driving patterns and test procedures between HD pickup trucks and vocational
vehicles, we are not applying the same effectiveness as for the pickups, instead we are applying a
cycle average effectiveness of one percent. Further, because friction reduction and cylinder
deactivation act in some overlapping ways to improve efficiency of engines, we are applying a
dis-synergy factor of 0.9. Thus the combination of these technologies results in a calculated
package effectiveness value of 0.8 percent, which we apply in each model year of Phase 2
standards. In terms of costs, the agencies have presented the costs of upgrading from EFR1 to
EFR2, as shown in Chapter 2.11.2.17 below. The costs of cylinder deactivation are shown in
Chapter 2.11.2.18. By applying our market adoption rates and incremental costs of these two
technologies, we estimate a vocational vehicle package cost due to improved SI engines of $138
in MY 2021 for this technology.
2.9.1.2.2 Improved Vocational Diesel Engine Technology for Standard-
Setting
As pointed out above, we consider that vocational and tractor vehicles share the same
engine hardware with 455hp and 350hp rating, since the same engines would likely be applied to
both tractor and vocational sectors, consistent with the current market structure. However,
moving to 2021, and 2024 and 2027 years, those maps between tractor and vocational vehicles
could start to deviate, even though the engine hardware remains the same, because of different
technology paths. Since the benefits obtained from WHR would be minimal for vocational
applications, we do not expect that WHR would be used in this sector (and the vocational vehicle
standards consequently do not reflect any use of engines with WHR). On the other hand,
transient control technology is one of the major enabling technologies in the vocational sector.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
In addition, the weighting of the composite certification cycles is much higher in the transient
cycle than in the 55 mph and 65 mph cruise speed cycles. In the vehicle standard-setting
process, we use the steady state map for the 55 and 65 mph cruise speed cycles, while the cycle
average maps are used for transient ARB cycle. The technology effectiveness map without
WHR and transient control technology is used to develop an engine fuel map for 55 and 65mph
cycles, where the same principle of engine fuel map from the tractor vehicle described in Chapter
2.8.4 is used. The second map is for the transient ARB cycle, where the total reduction of
technology effectiveness map without WHR but with transient control technology is used for the
cycle average map. After two maps are created, a weighting factor derived from three weighting
factors of 55mph, 65mph cruise speed cycles and transient ARB cycle is used to determine the
final reduction of emissions. For the sake of simplicity, it is noted that engines with 455hp and
350hp are the same ones as the tractor engines largely with the same technology path, and
therefore they can be grouped together by using one unique mapping methodology. On the other
hand, the engine with 200 hp and 270 hp for Class 2b-7 vehicles can be grouped into a second
group by using another set of mapping procedures, since the agencies used a different technology
path for these than for tractor engines.
Compared to the tractor engine technology table (Table 2-11) or with potential new
engine platform, the SET weighted reductions are identical except WHR setting to zero, and a
technology called model based control for transient operations is added. It is also noted that
market penetrations are different from Table 2-12. This is because new engine calibrations must
be developed without the WHR device, and portions of new engine platform may be less likely
applied to vocational sectors as opposed to the tractor market. Again, this is just one of the
technology paths proposed, and there could be many other ways to achieve the same goal. It is
also noted that the total reduction from each table is different, with more reductions predicted
from transient control than for control under steady state conditions. This reflects a different
technology path for each, and, specifically, that model based control for the transient operation
can play a significant role in reducing vehicle CO2 emissions.
The maps reflect that certain additional benefits from engine improvements can
appropriately be included in the vehicle standard, specifically, improvements based on will total
and more optimal integration between engine and transmission during transient operation. (As
explained in 2.8 above, the same approach is reflected with respect to engine improvements in
the tractor standard).
We next used these steady state and transient technology maps to translate the reductions
into the engine fuel maps used for GEM during the stringency standard runs. Figure 2-40
highlights the principle of the final mapping procedure. In this figure, SS stands for steady state.
Starting with the 2018 baseline engine fuel map (the top of this figure), the baseline cycle
average map is created with a 1.05 transient correction factor, which is used to multiply the fuel
rate obtained from a normal GEM simulation with a steady state engine fuel map. How the cycle
average map is created can be seen in Chapter 3 of the RIA. The transient factor of 1.05 is
derived from a large experimental data set to account for transient behavior. Next, 2018 baseline
technology maps, such as Figure 2-20 and Figure 2-21, are used to generate steady state engine
fuel maps for 2021, 2024, and 2027, following the exactly same procedure for HHD engines as
the tractor engine fuel maps, and the same procedure for Class 2b-7 as vocational engines (i.e.,
engines used in vocational vehicles). The cycle average maps for 2021, 2024, and 2027 will be

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review **'
generated based on the new derived cycle average multiplier as shown in Table 2-53 and Table
2-54.
1.05 transient correction
2018 map
Cycle Average map
SS Map
Reduction
Cycle average multiplier
SS 21/24/27
maps
Cycle Average
21/24/27 maps
Figure 2-40 Vocational engine fuel map for GEM run
The cycle average multipliers are shown in the table below, which are calculated by
subtracting the difference between the transient technology map reduction and the SS technology
map reduction from 1.05.
Table 2-53 Cycle Average Multiplier for HHD Engines
YEARS
SS TECHNOLOGY MAP
REDUCTION USED IN
GEM
TRANSIENT TECHNOLOGY
MAP REDUCTION USED IN
CYCLE AVERAGE MAP
CYCLE
AVERAGE
MULTIPLIER
2021
2.0%
2.8%
1.042
2024
3.4%
4.8%
1.036
2027
3.9%
5.5%
1.034
Table 2-54 Cycle Average Multiplier for LHD and MHD Engines
YEARS
SS TECHNOLOGY MAP
REDUCTION USED IN
GEM
TRANSIENT TECHNOLOGY
MAP REDUCTION USED IN
CYCLE AVERAGE MAP
CYCLE
AVERAGE
MULTIPLIER
2021
1.8%
2.6%
1.043
2024
3.4%
4.4%
1.036
2027
3.5%
5.2%
1.033
The overall reduction over the composite cycles differ as a result of combining steady
state mapping with transient mapping for the final vehicle stringency standard runs using GEM.
It should be between the total reduction shown in the steady state technology map and transient
technology maps. Since more aggressive model based control for transient operation is used in

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
the vehicle standards than for the engine standards, it can be expected that overall reduction
would be more than engine standards, which vehicle standard is in the range of 4.8 percent on
average over all vocational vehicles.
With the engine fuel map procedure developed, all vocational engine fuel maps can be
created. Figures shown below are the engine fuel maps used for vocational vehicles from 2021
to 2027, including 455hp, 350hp, 270hp, and 200hp engines.
2021 Engine 455hp / 15L BSFC ( g / kW * hr)
2000
1800
1600
1400
1 1200

-------
E. O. 12866 Review - Revised - Do Not Cite, Quote, or Release During Review
2021 Engine 270hp / 7L BSFC ( g / kW * hr)
900
800
700
600
500
Q)
F 400
o
H
300
230
245
200
265
100
800 1000 1200 1400 1600 1800 2000 2200 2400 2600
Speed ( RPM )
Figure 2-43 2021 Vocational Engine Fuel Map with 270hp Rating
2021 Engine 200hp / 7L BSFC ( g / kW * hr)
900
800
700
600
500
o
F 400
o
I—
300
230
245
265
2.65
230
245
265
200
100
800 1000 1200 1400 1600 1800 2000 2200 2400
Speed ( RPM )
Figure 2-44 2021 Vocational Engine Fuel Map with 200hp Rating

-------
E. O. 12866 Review - Revised - Do Not Cite, Quote, or Release During Review
2024 Engine 455hp / 15L BSFC ( g / kW * hr)
2000
1800
1600
1400
Z 1200
5. 1000
800
600
200
210
230
400
230
200
600
800
1000 1200 1400 1600 1800 2000 2200
Speed ( RPM )
Figure 2-45 2024 Vocational Engine Fuel Map with 455hp Rating
2024 Engine 350hp/11L BSFC ( g / kW * hr)
1600
1400
1200
1000
E
z
§ 800
w
o
i—
600
1\ 5
400
200
600 800 1000 1200 1400 1600 1 800 2000 2200
Speed ( RPM)
Figure 2-46 2024 Vocational Engine Fuel Map with 350hp Rating

-------
E. O. 12866 Review - Revised - Do Not Cite, Quote, or Release During Review
2024 Engine 270hp / 7L BSFC ( g / kW * hr)
900
800
700
600
E
z
500
0)
F 400
o
i—
300
240
200
260
260
100
800 1000 1200 1400 1600 1800 2000 2200 2400 2600
Speed ( RPM)
Figure 2-47 2024 Vocational Engine Fuel Map with 270hp Rating
2024 Engine 200hp / 7L BSFC ( g / kW * hr)
900
800
700
600
E
Z
500
<1>
F 400
O
I-
300
^15"
225
240
260
-285"
225
240
260
200
100
800 1000 1200 1400 1600 1800 2000 2200 2400
Speed ( RPM)
Figure 2-48 2024 Vocational Engine Fuel Map with 200hp Rating

-------
E. O. 12866 Review - Revised - Do Not Cite, Quote, or Release During Review
2027 Engine 455hp / 15L BSFC ( g / kW * hr)
2000
1800
1600
1400
Z 1200
1000
800
600
200
210
225
400
225
200
600
800
1000 1200 1400 1600 1800 2000 2200
Speed ( RPM )
Figure 2-49 2027 Vocational Engine Fuel Map with 455hp Rating
2027 Engine 350hp / 11L BSFC ( g / kW * hr)
1600
1400
1200
1000
E
Z
o
3
D-
800
o
I—
600
400
200
600 800 1000 1200 1400 1600 1800 2000 2200
Speed ( RPM)
Figure 2-50 2027 Vocational Engine Fuel Map with 350hp Rating

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
2027 Engine 270hp / 7L BSFC ( g / kW * hr)
- 260
800 1000 1200 1400 1600 1800 2000 2200 2400 2600
Speed ( RPM)
Figure 2-51 2027 Vocational Engine Fuel Map with 270hp Rating
2027 Engine 200hp / 7L BSFC ( g / kW * hr)
800 1000 1200 1400 1600 1800 2000 2200 2400
Speed ( RPM)
Figure 2-52 2027 Vocational Engine Fuel Map with 200hp Rating
2.9.2 Defining Baseline Vocational Vehicles
As at proposal, the agencies are subcategorizing the vocational vehicle sector by use of
three gross vehicle weight classes and three distinct test cycles. Also as proposed, these duty
cycles are termed Regional, Multipurpose, and Urban. However, the agencies have made

-------
*** E. O. 12866 Review — Revised - Do Not Cite, Quote, or Release During Review ***
significant changes to these duty cycles as well as changes to the specifications of vehicles that
are considered as part of the baseline for each of these subcategories. For the establishment of
three duty cycle-based subcategories, the agencies are relying on work conducted by the U.S.
Department of Energy at the National Renewable Energy Laboratory (NREL) that grouped
vehicles with similarities of key driving statistics into three clusters of operation. NREL's
methodology and findings are described in a report in the docket for this rulemaking,148
For development and refinement of the certification test cycles, the agencies have
considered NREL's work as well as public comment and engineering judgment. Details on how
the agencies established weightings of the different test cycles for each subcategory are
presented in the RIA Chapter 3.4.3. Figure 2-53 illustrates vehicles in NREL's fleet DNA
database plotted according to similarities in their driving statistics. In this image, the two clusters
identified in a prior exercise are joined by a middle cluster that contains vehicle traces that do not
clearly fall into either the left (slower) or right (faster) cluster. Each point represents one day of
driving in the entire data set. Points are colored according to their optimized cluster placement.
Trace Clustering - 8 Metrics (3 Clusters)
2-
0
Q.
1	o-
o
_a>
Q.
o











pet zerd®
, o «peed std
I
•
speed avg










-7.5
-5
.0
-2.5
0.0
2.5
slo
Principle Component 1
Figure 2-53 Three operational clusters observed by NREL
Consistent with the number of Phase 2 subcategories, nine baseline vocational vehicle
configurations have been developed for those powered by CI engines, plus six configurations for
vocational vehicle powered by SI engines, plus seven custom chassis baseline configurations.
Vocational vehicle attributes set by the agencies in both the baseline and in the executable
version of the GEM include: transmission gear efficiencies, transmission inertia, engine inertia,
axle efficiency, number of axles, axle inertia, axle efficiency, electrical and mechanical

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
accessory power demand, vehicle mass and payload, and aerodynamic cross-section and drag
coefficient. Other vehicle attributes that are available as user inputs for compliance purposes and
for which we have established baseline values include: engine power and displacement (and
multi-point fuel map), axle ratio, transmission type and gear ratios, and tire revs/mile.
In each of our defined baseline configurations, the agencies have not applied any vehicle-
level fuel saving or emission reduction technology beyond what is required to meet the Phase 1
standards. NHTSA and EPA reviewed available information regarding the likelihood that
manufacturers of vocational vehicles would apply technology beyond what is required for Phase
1, and we concluded that the best approach was to analyze a reference case that maintains
technology performance at the Phase 1 level. Thus, the GEM-simulated baseline vocational
vehicle configurations as well as the programmatic vocational vehicle reference case analyzed in
this rule represent what is referred to as a nominally flat baseline.
Tables 4-8, 4-9, and 4-10 in the RIA Chapter 4 present the non-user-adjustable modeling
parameters for HHD, MHD and LHD vocational vehicles, respectively. In addition to those
parameters, to completely define the baseline vehicles, the agencies also selected parameters
shown in Table 2-55 to Table 2-61. These attributes and parameters were selected to represent a
range of performance across this diverse segment, and are intended to represent a reasonable
range of vocational chassis configurations likely to be manufactured in the implementation years
of the Phase 2 program. The tire sizes and axle ratios were selected based on market research of
publically available manufacturer product specifications, as well as some manufacturer-supplied
information about configurations sold in prior model years. The transmission gear ratios were
selected based on the transmissions for which models have been validated in GEM, plus public
comments from transmission suppliers. We received public comments from Allison
recommending close transmission gear ratios for use in coach and transit buses, which we have
programmed as the default GEM transmission for these custom chassis. Considering all of the
above information, the agencies have significantly better defined vocational baselines than at
proposal. A summary of information on which we based these baselines is available in the
docket.149 In general, the trend is that vehicles with higher final drive ratios have been selected
for the subcategories with less weighting of the highway test cycles.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-55 Heavy Heavy-Duty Diesel Modeling Parameters for Vocational Vehicle Baseline
GEM PARAMETER
REGIONAL
REGIONAL
MULTI-
MULTI-
MULTI-
URBAN

(95%)
(5%)
PURPOSE
(80%)
PURPOSE
(10%)
PURPOSE
(10%)

CI Engine
2018 MY 15L
2018 MY
2018 MY 11L
2018 MY
2018 MY
2018 MY 11L

455hp Engine
11L 350 hp
engine
350 hp Engine
15L 455hp
Engine
11L 350 hp
Engine
350 hp Engine
Transmission Type
10-speed MT
6-speed AT
6-speed AT
10-speed MT
10-speed
MT
5-speed AT
Transmission Gears
12.8, 9.25,
3.51, 1.91,
4.6957, 2.213,
12.8, 9.25,
12.8, 9.25,
4.6957, 2.213,

6.76, 4.9,
1.43, 1.0,
1.5291, 1.0,
6.76,4.9,
6.76,4.9,
1.5291, 1.0,

3.58,2.61,
0.74, 0.64
0.7643,
3.58,2.61,
3.58,2.61,
0.7643

1.89, 1.38,

0.6716
1.89, 1.38,
1.89, 1.38,


1.0, 0.73


1.0, 0.73
1.0, 0.73

Torque converter lockup
3
3
3
3
3
3
gear






Drive Axle Gear Ratio
3.76
3.8
4.33
4.33
4.33
5.29
Axle Configuration
6x4
6x4
6x4
6x4
6x4
6x4
Tire Revs/mile
496
515
496
496
496
496
Steer Tires (CRR kg/metric
7.7
7.7
7.7
7.7
7.7
7.7
ton)






Drive Tires (CRR kg/metric
7.7
7.7
7.7
7.7
7.7
7.7
ton)






Electrified Accessories
0
0
0
0
0
0
Tire Pressure System
0
0
0
0
0
0
Idle Reduction
N
N
N
N
N
N
Weight Reduction (lb)
0
0
0
0
0
0

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-56 Vocational MHD SI Baseline Modeling Parameters
GEM PARAMETER
REGIONAL
MULTI-
PURPOSE
URBAN
SI Engine
2018 MY 6.8L, 300 hp engine
Transmission Type
6-speed AT
6-speed AT
5-speed AT
Transmission Gears
3.102, 1.8107, 1.4063, 1.0,0.7117,0.61
3.102, 1.8107,
1.4063, 1.0,
0.7117
Transmission
efficiency
GEM Default
Torque converter
lockup gear
3
3
3
Axle efficiency
GEM Default
Drive Axle Gear Ratio
5.5
5.1
5.1
Axle Configuration
4x2
4x2
4x2
Idle Reduction
No
Tire Revs/mile
517
557
557
Steer Tires (CRR
kg/metric ton)
7.7
7.7
7.7
Drive Tires (CRR
kg/metric ton)
7.7
7.7
7.7
Aerodynamic
Improvement
0
0
0
Electrified Accessories
0
0
0
Tire Pressure System
0
0
0
PTO Improvement
0
0
0
Weight Reduction (lb)
0
0
0

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-57 Vocational MHD Diesel Baseline Modeling Parameters
GEM PARAMETER
REGIONAL
MULTI-
PURPOSE
URBAN
CI Engine
2018 MY 7L, 270 hp Engine
Transmission Type
6-speed AT
6-speed AT
5-speed AT
Transmission Gears
3.102, 1.8107, 1.4063, 1.0,0.7117,0.61
3.102, 1.8107,
1.4063, 1.0,
0.7117
Transmission
efficiency
GEM Default
Torque converter
lockup gear
3
3
3
Axle efficiency
GEM Default
Drive Axle Gear Ratio
5.5
5.29
5.29
Axle Configuration
4x2
4x2
4x2
Idle Reduction
No
Tire Revs/mile
517
557
557
Steer Tires (CRR
kg/metric ton)
7.7
7.7
7.7
Drive Tires (CRR
kg/metric ton)
7.7
7.7
7.7
Aerodynamic
Improvement
0
0
0
Electrified Accessories
0
0
0
Tire Pressure System
0
0
0
PTO Improvement
0
0
0
Weight Reduction (lb)
0
0
0

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-58 SI Light Heavy-Duty Modeling Parameters for Vocational Baseline
GEM PARAMETER
REGIONAL
MULTI-PURPOSE
URBAN
SI Engine
2018 MY 6.8L, 300 hp engine
Transmission Type
6-speed AT
6-speed AT
5-speed AT
Transmission Gears
3.102, 1.8107, 1.4063, 1.0,0.7117,0.61
3.102, 1.8107,
1.4063, 1.0,0.7117
Transmission efficiency
GEM Default
Torque converter lockup
gear
3
3
3
Axle efficiency
GEM Default
Drive Axle Gear Ratio
4.33
4.88
4.88
Axle Configuration
4x2
4x2
4x2
Idle Reduction
No
Tire Revs/mile
680
680
660
Steer Tires (CRR kg/metric
ton)
7.7
7.7
7.7
Drive Tires (CRR kg/metric
ton)
7.7
7.7
7.7
Aerodynamic Improvement
0
0
0
Electrified Accessories
0
0
0
Tire Pressure System
0
0
0
PTO Improvement
0
0
0
Weight Reduction (lb)
0
0
0

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-59 Vocational LHD Diesel Baseline Modeling Parameters
GEM PARAMETER
REGIONAL
MULTI-
PURPOSE
URBAN
CI Engine
2018 MY 7L, 200 hp Engine
Transmission Type
6-speed AT
6-speed AT
5-speed AT
Transmission Gears
3.102, 1.8107, 1.4063, 1.0,
0.7117,0.61
3.102, 1.8107,
1.4063, 1.0,
0.7117
Torque converter lockup gear
3
3
3
Drive Axle Gear Ratio
4.33
4.56
4.56
Axle Configuration
4x2
4x2
4x2
Idle Reduction
No
Tire Revs/mile
670
670
660
Steer Tires (CRR kg/metric ton)
7.7
7.7
7.7
Drive Tires (CRR kg/metric ton)
7.7
7.7
7.7
Aerodynamic Improvement
0
0
0
Electrified Accessories
0
0
0
Tire Pressure System
0
0
0
PTO Improvement
0
0
0
Weight Reduction (lb)
0
0
0
The final baseline configurations for buses shown in Table 2-60 reflect comments from
Allison about close ratio transmission gear spreads that are common for these applications. The
transmission gear ratios for the other three types of HHD custom chassis use the same
transmission as in the HHD Urban primary subcategory. The final baseline configurations for
motor homes and school buses shown in Table 2-61 are identical to the respective baseline
configurations for MHD Regional and MHD Urban vehicles in the primary program.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-60 Custom Chassis HHD Baseline Modeling Parameters
GEM Parameter
Coach Bus
(Regional)
Refuse, Mixer,
Emergency (Urban)
Transit (urban)
CI Engine
2018 MY 11L, 350 hp Engine
Transmission Type
6-speed AT
5-speed AT
5-speed AT
Transmission Gears
3.51, 1.91, 1.43,
1.0, 0.74,0.64
4.69,2.213, 1.5291,
1.0, 0.7643
3.51, 1.91, 1.43, 1.0,
0.74
Torque converter lockup gear
3
3
3
Drive Axle Gear Ratio
4.33
5.29
5.29
Axle Configuration
6x2
6x4
4x2
Idle Reduction
No
No
No
Tire Revs/mile
496
496
517
Steer Tires (CRR kg/metric ton)
7.7
7.7
7.7
Drive Tires (CRR kg/metric ton)
7.7
7.7
7.7
Aerodynamic Improvement
0
0
0
Electrified Accessories
0
0
0
Tire Pressure System
0
0
0
PTO Improvement
0
0
0
Weight Reduction (lb)
0
0
0
Table 2-61 Custom Chassis MHD Baseline Modeling Parameters
GEM Parameter
Motor Homes
School Bus

(Regional)
(Urban)
CI Engine
2018 MY 7L, 270 hp Engine
Transmission Type
6-speed AT
5-speed AT
Transmission Gears
3.102, 1.8107,
3.102, 1.8107,

1.4063, 1.0,
1.4063, 1.0,0.7117

0.7117,0.61

Torque converter lockup gear
3
3
Drive Axle Gear Ratio
5.5
5.29
Axle Configuration
4x2
4x2
Idle Reduction
No
No
Tire Revs/mile
517
557
Steer Tires (CRR kg/metric
ton)
7.7
7.7
Drive Tires (CRR kg/metric
ton)
7.7
7.7
Aerodynamic Improvement
0
0
Electrified Accessories
0
0
Tire Pressure System
0
0
PTO Improvement
0
0
Weight Reduction (lb)
0
0

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.9.2.1 Setting Vocational Vehicle Baselines
The baseline performance of vocational vehicles powered by CI engines as described
above is shown in Table 2-62.
Table 2-62 Baseline Vocational Vehicle Performance with CI Engines
BASELINE EMISSIONS PERFORMANCE IN C02 GRAM/TON-MILE
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty Class
6-7
Heavy Heavy-Duty
Class 8
Urban
482
332
338
Multi-Purpose
420
294
287
Regional
334
249
220
Baseline Fuel Efficiency Performance in gallon per 1,000 ton-mile
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty Class
6-7
Heavy Heavy-Duty
Class 8
Urban
47.3477
32.6130
33.2024
Multi-Purpose
41.2574
28.8802
28.1925
Regional
32.8094
24.4597
21.6110
The baseline performance of vocational vehicles powered by SI engines as described
above is shown in Table 2-63.
Table 2-63 Baseline Vocational Vehicle Performance with SI Engines
BASELINE EMISSIONS PERFORMANCE IN C02 GRAM/TON-
MILE
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7 (and C8
Gasoline)
Urban
502
354
Multi-Purpose
441
314
Regional
357
275
Baseline Fuel Efficiency Performance in ga
Ion per 1,000 ton-mile
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7 (and C8
Gasoline)
Urban
56.4870
39.8335
Multi-Purpose
49.6230
35.3325
Regional
40.1710
30.9441
The baseline performance of the custom chassis configurations described above is shown
in Table 2-64.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-64 Baseline Performance of Custom Chassis
VEHICLE TYPE
EPA
NHTSA
Coach Bus
231
22.6916
Motor Home
249
24.4597
School Bus
332
32.6130
Transit
332
32.6130
Refuse
338
33.2024
Mixer
338
33.2024
Emergency
338
33.2024
2.9.2.2 Assigning Vocational Vehicles to Subcategories
In the NPRM, the agencies proposed criteria by which a vehicle manufacturer would
know in which vocational subcategory - Regional, Urban, or Multipurpose - the vehicle should
be certified. These cut-points were defined using calculations relating engine speed to vehicle
speed. Specifically, we proposed a cutpoint for the Urban duty cycle where a vehicle at 55 mph
would have an engine working above 90 percent of maximum engine test speed for vocational
vehicles powered by diesel engines and above 50 percent for vocational vehicles powered by
gasoline engines. Similarly we proposed a cutpoint for the Regional duty cycle where a vehicle
at 65 mph would have an engine working below 75 percent of maximum engine test speed for
vocational vehicles powered by diesel engines and below 45 percent for vocational vehicles
powered by gasoline engines. We received several comments that identified weaknesses in that
approach. Specifically, Allison explained that vehicles with two shift schedules would need
clarification which top gear to use when calculating the applicable cut-point. Also, Daimler
noted that, to the extent that downspeeding occurs in this sector over the next decade or more,
cutpoints based on today's fleet may not be valid for a future fleet. Allison noted that the
presence of additional top gears could strongly influence the subcategory placement of
vocational vehicles. These comments highlight the possibility of misclassification, and the
potential pitfalls in a mandated classification scheme. Furthermore, the agencies are concerned
that even if cutpoints were set that were viewed as valid in future years, manufacturers would be
able to satisfy the criteria to qualify for the regional subcategory by modifying driveline designs
slightly while maintaining customer satisfaction.
In a regulatory structure where standards for vehicles in different subcategories have
different stringencies, the agencies are inclined to prefer assigning subcategorization based on
regulatory criteria rather than allowing the manufacturers unconstrained choice. The approach to
setting of the final standards is explained in Preamble Section V.C.2.d. Below in Table 2-65 we
present our estimate of the distribution of vocational vehicles we predict will be certified in each
subcategory, as used only for estimating overall programmatic costs and benefits, not as part of
standard-setting. This estimate includes refined population distributions by weight class that
have been adjusted in part in response to comments on the draft NREL report in the NODA as
well as new analysis of telematics data from Ryder lease vehicles.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-65 Vocational Vehicle Types and Population Allocation
VEHICLE TYPE
REGIONAL
MULTI-PURPOSE
URBAN
C4-5 Short Haul Straight Truck
9%
41%
50%
C6-7 Short Haul Straight Truck
15%
50%
35%
C8 Short Haul Straight Truck
20%
60%
20%
Long Haul Straight Truck, Motor
Home, Intercity Bus
100%
0%
0%
School Bus
0%
10%
90%
Transit Bus
0%
0%
100%
Refuse
0%
10%
90%
All Class 4-5
11%
15%
18%
All Class 6-7
10%
11%
16%
All Class 8
5%
8%
6%
2.9.3Costs and Effectiveness of Vocational Vehicle Technologies
The following paragraphs describe the vehicle-level technologies on which the vocational
vehicle standards are predicated, and their projected effectiveness over the test cycles. The
methodology for estimating costs, including indirect cost estimates and learning effects, is
described in RIA Chapter 2.11.1. Certain elements of the cost estimating methodology are the
same as for the Phase 1 program, but certain elements are different including how the agencies
apply the markups, how the markups change with time, and which cost elements are influenced
by learning effects. As a result of different technology complexities, learning effects, and
different short-term and long-term warranty and non-warranty-related indirect costs, some
technology costs identified below may appear higher in MY 2021 than in MY 2027. These
differences are not due to changes in adoption rates, since the costs in Chapter 2.11 and below in
Chapter 2.9.3 to 2.9.4 are for applying a given technology to a single vehicle. Throughout this
chapter, where a dollar cost is given for a technology, note that these are adjusted to be valued as
year 2013 dollars. Average costs for vocational vehicle technology packages, including adoption
rates, are presented below in Chapter 2.9.5. Detailed descriptions of technology packages for SI
engines can be found in the RIA Chapter 2.6. Detailed descriptions of technology packages and
costs for CI engines can be found in the RIA Chapter 2.7.
2.9.3.1 Transmissions
Transmission improvements present a significant opportunity for reducing fuel
consumption and CO2 emissions from vocational vehicles. Transmission efficiency is important
for many vocational vehicles as their duty cycles involve high percentages of driving under
transient operation. The types of transmission improvements the agencies considered for Phase 2
are advanced shift strategy, gear efficiency, torque converter lockup, architectural improvements,
and hybrid powertrain systems.
Of the technologies described above in Chapter 2.4, the agencies are predicating the
vocational vehicle standards in part on performance improvements achieved by use of advanced
transmissions as described in Table 2-66, below. The projected market adoption rates that
inform the technology packages are described in Chapter 2.9.5.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-66 Projected Vocational Transmission Improvements over GEM Baseline
TRANSMISSION
PROJECTED

REGIONAL
MULTI-PURPOSE
URBAN
TECHNOLOGY
IMPROVEMENT OVER
COMPOSITE
COMPOSITE CYCLE
COMPOSITE

TEST CYCLE3

CYCLE

CYCLE
Two More Gears
ARB Transient
1.0%
1.7
1.2
0.9

55 mph Cruise
2.0%




65 mph Cruise
2.0%



Torque Converter
ARB Transient
1-5%
0.7 to 0.9
0.9 to 2.2
0.8 to 3.2
Lockup in 1st Gear (vs
55 mph Cruise
0%



3rd)
65 mph Cruise
0%



Non-Integrated Mild
ARB Transient
14%
3
8
11-12
Hybrid
55 mph Cruise
0%




65 mph Cruise
0%



Integrated Mild
ARB Transient
23-
4-5
14-19
19-25
Hybrid with Stop-Start

26%




55 mph Cruise
0%




65 mph Cruise
0%



Advanced Shift
ARB Transient
7%
3
4-5
5-6
Strategy
55 mph Cruise
2%




65 mph Cruise
2%



Note:
a Technology improvements modeled in GEM are TC lockup and gear number. Hybrids and shift strategy require separate
testing.
2.9.3.1.1 Advanced Shift Strategy
The technology we described at proposal as driveline integration, 80 FR 40296, is now
defined as use of an advanced shift strategy. At proposal the agencies included shift strategy,
aggressive torque converter lockup, and a high efficiency gearbox among the technologies
defined as driveline integration that would only be recognized by use of powertrain testing. The
agencies continue to believe that an effective way to derive efficiency improvements from a
transmission is by optimizing it with the engine and other driveline components to balance both
performance needs and fuel savings. One example of an engine manufacturer partnering with a
transmission manufacturer to achieve an optimized driveline is the SmartAdvantage
powertrain.150 Using engineering calculations to estimate the benefits that can be demonstrated
over the powertrain test, the agencies project that transmission shift strategies, including those
that make use of enhanced communication between engine and driveline, can yield efficiency
improvements ranging from three percent for Regional vehicles to nearly six percent for Urban
vehicles. The calculation is an energy-weighted and cycle-weighted average improvement using
cycle-specific CO2 emissions reported in the GEM output file for baseline vehicles. For the idle
cycles, the development version of GEM provides emissions in grams per hour. For the driving
cycles, GEM provides emissions in grams per ton-mile. By multiplying those values by the
average speed of each cycle and the default payload, all values are converted to grams per hour,
and these are surrogates for the energy expended over those cycles. For example, in the medium
heavy-duty Multipurpose subcategory with a payload of 5.6 tons, the baseline vehicle
configuration has cycle-specific results of 28,000 g CCh/hr for the transient cycle, 59,000 for the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
55 cycle, 85,000 for the 65 cycle, 8,500 for drive idle, and 3,700 for parked idle. By summing
the products of the percent improvement expected over each cycle, the CO2 emitted while
completing the cycle, and the associated composite weighting of the cycle, and dividing by the
sum of the products of the CO2 emitted and cycle weightings, we obtain results shown in Table
2-66. See the RIA Chapter 3.6 for a discussion of the powertrain test procedure.
The agencies have revised the GEM simulation tool to recognize additional transmission
technologies beyond what was possible at proposal. We are adopting a transmission efficiency
test to recognize improved mechanical gear efficiency and reduced transmission friction, where
the test results can be submitted as GEM inputs to override the default efficiency values. The
agencies project that vehicle fuel efficiency can be improved by up to one percent from
improved transmission gear efficiency, which we are projecting to be the same during each of the
driving cycles and (necessarily) zero while idling. Actual test results are likely to show that
some gears have more room for improvement than others, especially where a direct drive gear is
already highly efficient. Using the energy-weighted calculation method described above, the
transmission gear efficiency improvement used in our stringency calculations ranges from 0.82
to 0.97 percent. Final GEM also accepts an input field for torque converter lockup gear. As a
default, GEM simulates automatic transmissions using lockup in third gear. Using the library of
agency transmission files, GEM gives a different effectiveness value in every vocational vehicle
subcategory, because this is influenced by the gear ratios, drive cycle, and torque converter
specifications. Manufacturers will obtain slightly different results with their own driveline
specifications. The observed range of cycle-weighted effectiveness of torque converter lockup is
from less than one percent to three percent, as shown in Table 2-66 above.
Based on use of a sensor, the agencies estimate the total cost to apply an advanced shift
strategy for driveline integration is $87 in MY 2021 and $73 in MY 2027, as described in RIA
2.11.3.7. The agencies have also estimated capital and operational costs associated with building
test cells and conducting testing, as well as research and development costs associated with
designing shift strategies and integrating drivelines. These costs are presented in the RIA Chapter
7.1.1.2 and 7.1.1.3, respectively. The agencies estimate the total cost to apply a high efficiency
gearbox is $315 in MY 2021 and $267 in MY 2027, as described in RIA 2.11.3.5. The agencies
estimate the total cost to apply early torque converter lockup to a vocational vehicle at $31 in
MY 2021 and $26 in MY 2027, as described in RIA 2.11.3.6.
2.9.3.1.2 Architectural Transmission Improvements
One type of architectural improvement the agencies project can reasonably be developed
by manufacturers of all transmission architectures is increased number of gears. The benefit of
adding more gears varies depending on whether the gears are added in the range where most
operation occurs. In some cases additional gears in the low end of the range enhances driving
performance without improving fuel efficiency. The TIAX 2009 report projected that 8-speed
transmissions could incrementally reduce fuel consumption by 2 to 3 percent over a 6-speed
automatic transmission, for Class 3-6 box and bucket trucks, refuse haulers, and transit buses.151
We have run GEM simulations comparing 5-speed, 6-speed, 7-speed, and 8-speed automatic
transmissions where some cases hold the total spread constant, some hold the high end ratio
constant, and some hold the low-end ratio constant, where all cases use a third gear lockup and
axle ratios are held constant. We have observed mixed results, with some improvements over the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
highway cruise cycles as high as six percent, and some cases where additional gears increased
fuel consumption. As proposed, we are allowing GEM to determine the improvement, where
manufacturers will enter the number of gears and gear ratios and the model will simulate the
efficiency over the applicable test cycle. The agencies have revised GEM based on comment,
and we are confident that it fairly represents the fuel efficiency of transmissions with different
gear ratios. Consistent with literature values, we are using engineering calculations to estimate
that two extra gears has an effectiveness of one percent improvement during transient driving
and two percent improvement during highway driving. Weighting these improvements using our
final composite duty cycles (zero improvement at idle) and the energy-weighting method
described above, this technology is estimated to improve vocational vehicle efficiency between
0.9 and 1.7 percent. The agencies estimate the total cost to add two gears to a vocational vehicle
transmission at $504 in MY 2021 and $465 in MY 2027, as described in RIA 2.11.3.1.
Most vocational vehicles currently use torque converter automatic transmissions (AT),
especially in Classes 2b-6. Automatic transmissions offer acceleration benefits over drive cycles
with frequent stops, which can enhance productivity. With the diversity of vocational vehicles
and drive cycles, other kinds of transmission architectures can meet customer needs, including
automated manual transmissions (AMT), dual clutch transmissions (DCT), as well as manual
transmissions (MT).152 As at proposal, dual clutch transmissions may be simulated as AMT's in
GEM. A manufacturer may elect to conduct powertrain testing to obtain specific improvements
for use of a DCT. The RIA Chapter 4.2.2.3 explains the EPA default shift strategy and the losses
associated with each transmission type, and discusses changes that have been made since
proposal. Although the representation of transmissions has improved since proposal, the
differences between AT and AMT are too difficult to isolate for purposes of figuring them into
our stringency calculations. Although we expect manufacturers to have a reasonable model of
transmission behavior for certification purposes, we could not estimate relative improvement
values between AT and AMT for vocational vehicles using any defensible estimation method.
The agencies have not been able to obtain conclusive data that could support a final vocational
vehicle standard, in any subcategory, predicated on adoption of an AMT or DCT with a
predictable level of improvement over an AT. As a result, the only architectural changes on
which the final vocational vehicle standards are based are increasing number of gears and
automation compared with a manual transmission. The final Phase 2 GEM has been calibrated to
reflect a fixed two percent difference between manual transmissions and automated
transmissions during the driving cycles (zero at idle). As in the HHD Regional subcategory
baseline, manual transmissions simulated in GEM perform two percent worse than similarly-
geared AMT. This fixed improvement is discussed further in RIA Chapter 2.4. The agencies
have estimated the cost of upgrading from HHD manual transmissions to AMT at $4,540 in MY
2021 and $3853 in MY 2027, as described in RIA 2.11.3.2.
2.9.3.1.3 Hybrid Drivelines
Hybrid drivelines are included under transmission technologies because, depending on
the design and degree of hybridization, they may either replace a conventional transmission or be
deeply integrated with a conventional transmission. Further, these systems are often
manufactured by companies that also manufacture conventional transmissions.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The agencies are including hybrid powertrains as a technology on which some of the
vocational vehicle standards are predicated, in part.
After considering comments, the agencies are projecting adoption of two types of mild
hybrids, defined using system parameters based on actual systems commercially available in the
market today.153 Some mild hybrid systems will be integrated with an engine sufficient to enable
use of an engine stop-start feature, while some mild hybrids will not be integrated and will only
provide transient benefits related to regenerative braking. We also have reconsidered our
effectiveness estimation method as a result of comments. Instead of relying on previously
published road tests over varying drive cycles, we are applying engineering calculations to
account for defined hybrid system capacities and inefficiencies over our certification test cycle.
We are using a spreadsheet model that calculates the recovered energy of a hybrid system using
road loads of the default baseline GEM vehicles over the ARB Transient test cycle.154
The inputs to this spreadsheet model include maximum hybrid system power, battery
capacity, allowable swing in the battery state of charge, system efficiencies, as well as vehicle
road loads such as tire rolling resistance, vehicle mass and aerodynamic drag area. For stringency
purposes, the system inputs used were 75 kW motor, 8 kWh battery, and 10 percent swing in
SOC. The system efficiencies included 90 percent, 90 percent, 90 percent, 92 percent and 85
percent, for the battery, power electronics, electric motor, axle and transmission, respectively.
The vehicle road loads were identical to those in the baseline GEM vehicle configurations.
Within the system constraints the algorithm stores and releases the available kinetic energy from
the vehicle without any information of engine efficiency through the cycle. The calculations also
take into account the energy that is needed to drive the accessories through the drivetrain when
the vehicle is decelerating. The algorithm is iterative, and the calculations continue until the
battery net energy change is at a value less than one percent of the total fuel energy which is
approximated by 3 times the total tractive work of the cycle.
One simplification in the spreadsheet model is that the effectiveness is assumed to be
zero for the highway cruise cycles. In the real world there are driving conditions on highways
that may present opportunities to capture and re-use energy, including conditions related to road
grade and congestion. However, for this simplified method we are not counting the benefits of
systems that make use of such opportunities. We are not projecting substantially less
effectiveness for heavier vehicles than for lighter vehicles, even though the same systems were
assessed for all weight classes (not scaled up for heavier vehicles). This is due in part to the
assumptions about the fraction of brake energy that enters the hybrid system vs the fraction that
goes entirely to friction braking.
Using this spreadsheet model and system inputs described above, for the non-integrated
mild hybrids, we are estimating a one to 12 percent fuel efficiency improvement over the
powertrain test, depending on the duty cycle (i.e. Regional, Urban, or Multi-purpose) in GEM for
the applicable subcategory. For the integrated mild hybrids, we have projected that the systems
are scaled up for heavier vehicles, and we have combined the effectiveness calculated using the
hybrid spreadsheet model with the GEM effectiveness of stop-start, described below. These
combined effectiveness values range from four to 25 percent for the mild hybrids with stop-start.
Even though the actual improvement from hybrids in Phase 2 will be evaluated using the
powertrain test, because the model uses the same vehicle test cycle and conservative estimates of

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
realistic configurations, the agencies have concluded it is reasonable to use these spreadsheet-
based estimates as a basis for setting stringency in the final rules.
The industry is currently developing many variations of hybrid powertrain systems.
There are a few hybrid systems in the heavy-duty market today and several more under
development, as evidenced by several public comments on this rulemaking. See also Chapter
6.3.3 of the Response to Comments document. In addition, energy storage systems are getting
better.155 Heavy-duty customers are getting used to these systems with the number of
demonstration products on the road. A list of hybrid manufacturers and their products intended
for the vocational market is provided in Table 2-67.
Table 2-67 Examples of Hybrid Manufacturers
MANUFACTURER
PRODUCT
EXAMPLE APPLICATION
Hino
Class 5 cab-over-engine battery-
electric hybrid
Delivery Trucks
Allison
HHD parallel hybrid
Transit Bus
BAE
HHD series or parallel hybrid
Transit Bus
XL
Class 3-4 mild electric hybrid
Shuttle Bus
Crosspoint Kinetics
Class 3-7 mild electric hybrid
Delivery trucks, shuttle buses
Lightning Hybrids
Class 2-5 hydraulic hybrid
Delivery trucks
Parker Hannifin
MHD hydraulic hybrid
Delivery trucks
Freightliner Custom Chassis
MHD hydraulic hybrid
Delivery trucks
Morgan Olson
MHD hydraulic hybrid
Delivery trucks
Autocar-Parker
Runwise hydraulic hybrid
Refuse Trucks
Eaton3
HHD parallel electric hybrid
Trucks and Buses
Odyne
Plug-in electric hybrid, E-PTO
Utility Trucks
Note:
a Currently selling in markets outside the U.S.
The agencies estimate the total cost of a bolt-on, non-integrated mild hybrid system for
any size vocational vehicle at $8,906 in MY 2021 and $6,906 in MY 2027. The agencies
estimate the total cost of an integrated mild hybrid system with stop-start for a LHD vocational
vehicle is $6,320 in MY 2021 and $5,082 in MY 2027. For a MHD vocational vehicle, the total
cost of an integrated system is estimated at $9,934 in MY 2021 and $7,989 in MY 2027. For a
HHD vocational vehicle, the total cost of an integrated system is estimated at $16,590 in MY
2021 and $13,341 in MY 2027, as described in RIA 2.11.7. The estimated higher costs for
heavier vehicles are related to higher power demands and greater energy storage needs. These
estimates assume no engine downsizing in the design of hybrid packages. This is in part to be
conservative in our cost estimates, and in part because in some applications a smaller engine may
not be acceptable if it would risk that performance could be sacrificed during some portion of a
work day.
2.9.3.2 Axles
The agencies are predicating part of the stringency of the final vocational vehicle
standards on performance of two types of axle technologies. The first is advanced low friction
axle lubricants and efficiency as demonstrated using the separate axle test procedure described in

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
the RIA Chapter 3.8 and 40 CFR 1037.560. The agencies received many adverse comments on
the proposal to assign a fixed 0.5 percent improvement for this technology. In consideration of
comments, the agencies are assigning default axle efficiencies to all vocational vehicles.
Manufacturers may submit test data to over-ride these default axle efficiency values in GEM.
Based on comments from axle suppliers as well as other available data, we project the
effectiveness of technologies to improve axle efficiency can achieve between two and three
percent improvement.156 Our cost analysis for the final rulemaking includes maintenance costs
of replacing axle lubricants on a periodic basis. Based on supplier information, some advanced
lubricants have a longer drain interval than traditional lubricants. We are estimating the axle
efficiency & lubricating costs for HHD to be the same as for HHD tractors since those vehicles
likewise typically have three axles. For HHD vocational vehicles (with 3 axles), the agencies
estimate the cost at $200 in MY 2021 and $174 in MY 2027, as described in RIA 2.11.5.4.
However, for LHD and MHD vocational vehicles, we scaled down the cost of this technology to
reflect the presence of a single rear axle. The agencies estimate the total cost of improved axle
efficiency on a LHD or MHD vocational vehicle (with 2 axles) at $134 in MY 2021 and $116 in
MY 2027.
The second axle technology applies only for HHD vocational vehicles, which typically
are built with two rear axles. Part time 6x2 configuration or axle disconnect is a design that
enables one of the rear axles to temporarily disconnect or otherwise behave as if it's a non-driven
axle. The agencies proposed to base the HHD vocational vehicle standard on some use of both
part time and full time 6x2 axles. The agencies received compelling adverse comment on the
application of the permanent 6x2 configuration for vocational vehicles, and in response we are
not basing the final vocational vehicle standards on any adoption of full time 6x2 axles. The
disconnect configuration is one that keeps both drive axles engaged only during some types of
vehicle operation, such as when operating at construction sites or in transient driving where
traction especially for acceleration is vital. Instead of calculating a fixed improvement as at
proposal, the agencies have refined GEM to recognize this configuration as an input, and the
benefit will be actively simulated over the applicable drive cycle. Effectiveness based on
simulations with EPA axle files is projected to be as much as 1.2 percent for HHD Regional
vehicles. Further information about this technology is provided in RIA Chapter 2.4.
The agencies estimate the total cost of part time 6x2 on a vocational vehicle at $121 in
MY 2021 and $117 in MY 2027, as described in RIA 2.11.5.2.
2.9.3.3 Lower Rolling Resistance Tires
Tires are the second largest contributor to energy losses of vocational vehicles, as found
in the energy audit conducted by Argonne National Lab.157 The two most helpful sources of data
in establishing the projected vocational vehicle tire rolling resistance levels for the final Phase 2
standards are the comments from RMA and actual certification data for model year 2014. At
proposal, we projected that all vocational vehicle subcategories could achieve average steer tire
coefficient of rolling resistance (CRR) of 6.4 kg/ton and drive tire CRR of 7.0 kg/ton by MY
2027. These new data have informed our analysis to enable us to differentiate the technology
projections by subcategory. The RMA comments included CRR values for a wide range of
vocational vehicle tires, for rim sizes from 17.5 inches to 24.5 inches, for steer/all position tires
as well as drive tires. The RMA data, while illustrating a range of available tires, are not sales

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
weighted. The 2014 certification data include actual production volumes for each vehicle type,
thus both steer and drive tire population-weighted data are available for emergency vehicles,
cement mixers, school buses, motor homes, coach buses, transit buses, and other chassis cabs.
The certification data are consistent with the RMA assessment of the range of tire CRR currently
available. We also agree with RMA's suggestion to set a future CRR level where a certain
percent of current products can meet future GEM targets. We disagree with RMA that the MY
2027 target should be a level that 50 percent of today's product can meet. With programmatic
averaging, such a level would mean essentially no improvements overall from tire rolling
resistance, because today when manufacturers comply on average, half their tires are above the
target and half are below. Further, with Phase 2 GEM requiring many more vehicle inputs than
tire CRR, manufacturers have many more degrees of freedom (i.e. other available compliance
pathways) to meet the performance standard than they do in Phase 1. In these final rules, the
agencies are generally projecting adoption of LRR tires in MY 2027 at levels currently met by 25
to 40 percent of today's vocational products, on a sales-weighted basis.158 Figure 2-54 and
Figure 2-55 present a summary of the CRR levels of tires fitted on vocational vehicles certified
in the 2014 model year.
MY 2014 Vocational Drive Tires
16 ?
14 j
12 ^
¦ Max a90%-ile I Avg ¦25%-ile BMin
Figure 2-54 Vocational Drive Tire CRR Data Summary

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
MY 2014 Vocational Steer Tires
16 i
14 j
12 ^
¦ Max a90%-ile I Avg ¦25%-ile BMin
Figure 2-55 Vocational Steer Tire CRR Data Summary
The agencies acknowledge there can be tradeoffs when designing a tire for reduced
rolling resistance. These tradeoffs can include characteristics such as wear resistance, cost and
scuff resistance. NHTSA, EPA, and ARB met with stakeholders from the tire industry
(Bridgestone, Continental, Cooper, Goodyear, and Michelin) to discuss the next generation of
lower rolling resistance (LRR) tires for the Phase 2 timeframe for all segments of Class 2b-8
vehicles, including trailers. Manufacturers discussed forecasts for rolling resistance levels and
production availability in the Phase 2 timeframe, as well as their plans for improving rolling
resistance performance while maintaining other performance parameters such as traction,
handling, wear, mass reduction, retreadability, and structural durability.
The meetings included specific discussions of the impacts of the current generation of
LRR tires on vehicle stopping distance and handling. Manufacturers indicated no known safety
detriment in the current on-road fleet from use of LRR tires. While the next generation of tires
may require some tradeoffs in wear performance and costs over the next 10 years to achieve
better tire rolling resistance performance, manufacturers said they will not trade off safety for
performance. They also emphasized that keeping tires inflated (through proper maintenance or
automatic systems) was the best way to assure long term fuel efficiency and safety during
vehicle operation.
In these final rules, we are differentiating the improvement level by weight class and duty
cycle, recognizing that heavier vehicles designed for highway use can generally apply tires with
lower rolling resistance than other vehicle types, and will see a greater benefit during use. In the
Preamble at Section V.C.I, Table V-14, the agencies define five levels of CRR for purposes of
estimating the manufacturing costs associated with applying improved tire rolling resistance to
vocational vehicles. None of the rolling resistance levels projected for adoption in MY 2027 are
lower than the 25th percentile of tire CRR on actual vocational vehicles sold in MY 2014. Thus,

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
we believe the improvements will be achievable without need to develop new tires not yet
available.
As an example of the total vehicle costs to apply LRR tires, the agencies estimate the
total cost to fit a LHD or MHD vocational vehicle with two LRR level 5v steer tires ($57) and
four level 3v drive tires ($107) to be $164 in MY 2021. Detailed tables of LRR tire costs in each
year are provided in RIA Chapter 2.11.8.
As proposed, the agencies will continue the light truck (LT) tire CRR adjustment factor
that was adopted in Phase 1. 80 FR 40299; see generally 76 FR 57172-57174. In Phase 1, the
agencies developed this adjustment factor by dividing the overall vocational test average CRR of
7.7 by the LT vocational average CRR of 8.9. This yielded an adjustment factor of 0.87.
Because the MY 2014 certification data for LHD vocational vehicles may have included some
CRR levels to which this adjustment factor may have already been applied, and because we did
not receive adverse comment on our proposal to continue this, the agencies have concluded that
we do not have a basis to discontinue allowing the measured CRR values for LT tires to be
multiplied by a 0.87 adjustment factor before entering the values in the GEM for compliance.
2.9.3.4 Workday Idle Reduction
The Phase 2 idle reduction technologies considered for vocational vehicles are those that
reduce workday idling, unlike the overnight or driver rest period idling of sleeper cab tractors.
Idle reduction technology is one type of technology that is particularly duty-cycle dependent. In
light of new information, the agencies have learned that our proposal had mischaracterized the
idling operation of vocational vehicles, significantly underestimating the extent of this mode of
operation, and incorrectly calculating it using a drive idle cycle when significant idling also
occurs while parked. As described in Preamble Section V.B.I, in these final rules we have
revised our test cycles to better reflect real world idle operation, including both parked idle and
drive idle test conditions. The RIA Chapter 3.4.2 describes these certification test cycles.
The Phase 1 composite test cycle for vocational vehicles includes a 42 percent weighting
on the ARB Transient test cycle, which comprises nearly 16 percent of idle time. However, no
single idle event in this test cycle is longer than 36 seconds, which is not enough time to
adequately recognize the benefits of idle reduction technologies.1^ In the Phase 2 proposal, we
applied composite test cycle weightings of 10, 20, and 30 percent of a drive idle cycle to the
Regional, Multipurpose and Urban duty cycles, respectively. These weightings were an initial
estimate because the interagency agreement between EPA and DOE-NREL to collaborate to
characterize workday idle among vocational vehicles was not yet complete. As shown in Table
2-68, the average total amount of daily total idle operation per vehicle identified by NREL is 25
percent for vehicles observed in the high speed cluster, 47 percent for vehicles observed in the
slow speed cluster, and 52 percent for vehicles straddling those two clusters. This work was
shared as part of the NODA and supported by commenters. Although some comments indicated
individual fleets log different idle times than those in our test cycles, the final test cycles are
L However, as noted above, emission improvements due to workday idle technology can be recognized under Phase
1 as an innovative credit under 40 CFR 1037.610 and 49 CFR 535.7.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
representative of the range of operation and adequately capture vocational vehicle idle behavior
for purposes of recognizing workday idle reduction technology.
Table 2-68 Summary of Out-of-Gear Idle Behavior
NREL
Cluster
Operating Mode
NREL
Percent of
Workday
Percent
Accounted
for in Final
Transient
Cycle
Final
Weighting of
Parked Idle
Cycle
Final
Weighting
of Drive
Idle Cycle
Sum Of All
Regulatory Idle
Test Weighings
1
Out of Gear Idle
28

25


1
In Gear (Drive Idle)

10

15

1
Zero Speed (both in
gear and out of
gear)
47



50
2
Out of Gear Idle
22

25


2
In Gear (Drive Idle)

8

17

2
Zero Speed (both in
gear and out of
gear)
52



50
3
Out of Gear Idle
25

25


3
In Gear (Drive Idle)

6

0

3
Zero Speed (both in
gear and out of
gear)
25



25
The separate drive idle cycle supplements the drive idle that occurs during the transient
cycle. The time fraction of drive idle represented in the transient cycle is a complex iterative
equation because that is a distance-based cycle. By setting a total target zero-speed time of 50
percent for Multipurpose and Urban vehicles consistent with the recommendations of NREL, the
agencies were able to assign appropriate cycle weightings to the drive idle and parked idle test
cycles for each subcategory. In the final rules, the Regional duty cycle has 25 percent composite
test cycle weighting of parked idle and zero drive idle. The Multi-purpose cycle has 25 percent
of drive and 17 percent parked idle, and the Urban cycle has 15 percent drive idle and 25 percent
parked idle. The final cycle weightings are derived from data summarizing miles accumulated
within 2 mph speed bins for representative vehicles in each cluster. Details on development of
the cycle weightings are found in the RIA Chapter 3.4.3.1 and in the vocational vehicle duty
cycle report by NREL, which is available in the docket.159
At proposal, we identified two types of idle reduction technologies to reduce workday
idle emissions and fuel consumption for vocational vehicles: neutral idle and stop-start. After
considering the new duty cycle information and the many comments received, we are basing our
final vocational vehicle standards in part on the performance of three types of workday idle
reduction technologies: neutral idle, stop-start, and automatic engine shutdown. We believe that
these technologies are effective, feasible, and cost-effective, as discussed further in this section.
Neutral idle is essentially a transmission technology, but it also requires a compatible
engine calibration. Torque converter automatic transmissions traditionally place a load on

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
engines when a vehicle applies the brake while in drive, which we call curb idle transmission
torque (CITT). When an engine is paired with a manual or automated manual transmission, the
CITT is naturally lower than when paired with an automatic, as a clutch disengagement must
occur for the vehicle to stop without stalling the engine. We did not receive adverse comment on
our proposal to include this technology in our standard-setting for vocational vehicles. The
engineering required to program sensors to detect the brake position and vehicle speed, and
enable a smooth re-engagement when the brake pedal is released makes this a relatively low
complexity technology that can be deployed broadly. Navistar commented that idle reduction
strategies must have sufficient engine, aftertreatment and occupant protections in place such that
any fuel cost savings are a net benefit for the owner/operator without compromising safety. We
agree, and for neutral idle we believe an example of an allowable override is if a vehicle is
stopped on a hill. Skilled drivers operating manual transmissions can safely engage a forward
gear from neutral when stopped on upslopes with minimal roll-back. With an AT, the vehicle's
computer would need to handle such situations automatically. In the Phase 2 certification
process, transmission suppliers will attest whether the transmission has this feature present and
active, and certifying entities will be able to enter Yes or No as a GEM input for the applicable
field. The effectiveness of this technology will be calculated using data points collected during
the engine test, and the appropriate fueling over the drive idle cycle and the transient cycle will
be used. Based on GEM simulations using the final vocational vehicle test cycles, the agencies
project neutral idle to provide fuel efficiency improvements ranging from one to seven percent,
depending on the regulatory subcategory. Details are in the docket for this rulemaking.160
Automatic engine shutdown (AES) is an engine technology that is widely available in the
market today, but has seen more adoption in the tractor market than for vocational vehicles.
Although we did not propose to include this technology, we received many comments suggesting
this would be appropriate. Some commenters may have conflated the concept of stop-start with
AES, such as a comment on stop-start asking us to consider the on-board need to power
accessories while the vehicle is in stationary mode. We believe that automatic engine shutdown
is effective and feasible for many different types of vehicles, depending on how significant a
portion of the work day is spent while parked. Most truck operators are aware of the cost of fuel
consumed while idling, and importantly, the wear on the engine due to idling. Engine
manufacturers caution owners to monitor the extent of idling that occurs for each work truck and
to reduce the oil change interval if the idle time exceeds ten percent of the work day.161
Accordingly, many utility truck operators track their oil change intervals in engine hours rather
than in miles.
NTEA provided the agencies with a report with survey results on which work truck fleets
are adopting AES with backup power, and their reasons for doing so.162 The most common
reason given in the survey is to allow an engine to shut down and still have vehicle power
available to run flashing safety lights. Some vocational vehicles also need to conduct work using
a power take-off (PTO) while stationary for hours, such as on a boom truck. The agencies are
adopting an allowable AES over-ride for PTO use. Technologies that can reduce fuel
consumption during this type of high-load idle are discussed below and in the Preamble at
V.C.l.c. We are also adopting an allowable AES over-ride if the battery state of charge drops
below a safe threshold. This would ensure there is sufficient power to operate any engine-off
accessories up to a point where the battery capacity has reached a critical point. Where a
vocational vehicle has such extensive stationary accessory demands that an auxiliary power

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
source is impractical or that an over-ride condition would be experienced frequently, we would
not consider AES to be feasible. In the Phase 2 certification process, engine suppliers will attest
whether this feature is present and tamper-proof, and certifying entities will be able to enter Yes
or No as a GEM input for the applicable field.M As with neutral idle described above, the
effectiveness of AES will be calculated in GEM using data obtained through engine testing. The
appropriate data points over the parked idle cycle will be used for calculating the fueling. Based
on GEM simulations using the final vocational vehicle test cycles, the agencies project AES to
provide fuel efficiency improvements ranging from less than one to seven percent for diesel
vehicles, and from three to eight percent for gasoline vehicles, depending on the regulatory
subcategory. Other overrides are listed in the regulations at 40 CFR 1037.660.
While the primary program does not simulate vocational vehicles over a test cycle that
includes PTO operation, the agencies will continue, with revisions, the hybrid-PTO test option
that was in Phase 1. See 40 CFR 1037.540 and 76 FR 57247. Recall that we will regulate
vocational vehicles at the incomplete stage when a chassis manufacturer may not know at the
time of certification whether a PTO will be installed or how the vehicle will be used. Although
chassis manufacturers will certainly know whether a vehicle's transmission is PTO-enabled, that
is very different from knowing whether a PTO will actually be installed and how it will be used.
Chassis manufacturers may rarely know whether the PTO-enabled vehicle will use this capability
to maneuver a lift gate on a delivery vehicle, to operate a utility boom, or merely as a reserve
item to add value in the secondary market. In cases where a manufacturer can certify that a PTO
with an idle-reduction technology will be installed either by the chassis manufacturer or by a
second stage manufacturer, the hybrid-PTO test cycle may be utilized by the certifying
manufacturer to measure an improvement factor over the GEM duty cycle that would otherwise
apply to that vehicle. In addition, the delegated assembly provisions would apply. See Preamble
Section I.F.2 for a description of the delegated assembly provisions. See RIA Chapter 3.7.4 for a
discussion of the revisions to the hybrid PTO test cycle. In cases where a chassis manufacturer
does not know whether a powertrain that is PTO-enabled will actually have a PTO-using tool
installed, and whether there will be an energy storage system installed to save fuel during PTO
operation, then the agencies do not see a way for the Phase 2 program to recognize hybrid PTO
technology.
Our estimates are that applying neutral idle to a vocational vehicle with an automatic
transmission would cost $118 in MY 2021, decreasing to $114 in MT 2027, as shown in RIA
2.11.6.5. These costs are increased from proposal, based on comments from Allison indicating
hardware may be needed, such as a sensor to detect brake position or road grade. Our estimates
are that applying AES to a vocational vehicle would cost $30 in MY 2021, decreasing to $25 in
MT 2027, as shown in RIA 2.11.6.7. This cost does not include the cost of an auxiliary power
source while the engine is off.
Based on GEM simulations using the final vocational vehicle test cycles, the agencies
project stop-start to provide fuel efficiency improvements ranging from less than one to 14
percent for diesel vehicles, and from one to ten percent for gasoline vehicles, depending on the
regulatory subcategory. Our estimates are that the cost of applying stop-start to a vocational
vehicle will vary by vehicle weight class, because varying amounts of engine and vehicle
M We will consider non-tamper-proof AES as off-cycle technologies for a lesser credit.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
upgrades will be needed. For LHD vocational vehicles, we estimate the total cost would range
from $871 in MY 2021 to $722 in MY 2027. For MHD vocational vehicles, we estimate the
total cost would range from $917 in MY 2021 to $760 in MY 2027. For HHD vocational
vehicles, we estimate the total cost would range from $1,683 in MY 2021 to $1,395 in MY 2027.
These costs, presented in RIA Chapter 2.11.6.6, are derived from costs reported by Tetra Tech
for stop-start, plus costs for electrified accessories derived from values used in the light-duty
GHG program, and scaled up for heavier vehicles.
With either a stop-start engine feature or with a neutral idle transmission calibration, less
fuel is burned at idle. Furthermore, it is expected that SCR catalyst function could be better
managed when an engine shuts off than when it idles. SCR systems are well insulated and can
maintain temperature when an engine is shut off, whereas idling causes relatively cool air to flow
through a catalyst. Therefore, the agencies have reason to believe there may be a NOx co-
benefit to stop-start idle reduction technologies, and possibly also to neutral idle. This would be
true if the NOx reductions from reduced fuel consumption and retained aftertreatment
temperature were greater than any excess NOx emissions due to engine re-starts.
2.9.3.5 Weight Reduction
The agencies are predicating the final vocational vehicle standards in part on use of
material substitution for weight reduction. The method of recognizing this technology is similar
to the method used for tractors. The agencies have created a menu of vocational chassis
components with fixed reductions in pounds that may be entered in GEM when substituting a
component made of a more lightweight material than the base component made of mild steel.
According to the 2009 TIAX report, there are freight-efficiency benefits to reducing weight on
vocational vehicles that carry heavy cargo, and tax savings potentially available to vocational
vehicles that remain below excise tax weight thresholds. This report also estimates that the cost
effectiveness of weight reduction over urban drive cycles is potentially greater than the cost
effectiveness of weight reduction for long haul tractors and trailers. We are adopting as
proposed a GEM allocation of half the weight reduction to payload and half to reduced chassis
weight. We did not receive comment suggesting a different weight allocation. The menu of
components available for a vocational vehicle weight credit in GEM is presented in Table 2-70
and can be found in the regulations at 40 CFR 1037.520. It includes fewer options than
proposed, due to comments from Allison that aluminum transmission cases and clutch housings
are standard for automatic transmissions. The agencies believe there are a number of other
feasible material substitution choices at the chassis level, which could add up to weight savings
of hundreds of pounds. The stringency of the final vocational vehicle standards for custom
chassis transit buses and vehicles in the primary program is based in part on use of aluminum
wheels in 10 positions on 3-axle vocational vehicles (250 lbs) and in 6 wheel positions on 2-axle
vocational vehicles (150 lbs). This is a change from proposal, where we believed application of
lightweight components would be adopted more narrowly. Our projected adoption rate is revised
upward based on the determination that the technology package is smaller (fewer pounds
removed than at proposal) and that aluminum wheels are widely feasible. Based on the default
payloads in GEM, and depending on the vocational vehicle subcategory, the agencies estimate a
reduction of 250 lbs would offer a fuel efficiency improvement of up to one percent for HHD
vehicles, and a reduction of 150 pounds would offer a fuel efficiency improvement up to 0.8
percent for MHD vehicles, and up to 1.5 percent for LHD vehicles, as shown in Table 2-69. The

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
agencies estimate the total cost to apply aluminum wheels to LHD and MHD vocational vehicles
(about 150 pounds) to be $693 in MY 2021 and $587 in MY 2027, as described in RIA
2.11.10.3. We estimate the total cost to apply aluminum wheels to 3-axle vocational vehicles
(about 250 pounds) to be $2495 in MY 2021 and $2204 in MY 2027, as described in RIA
2.11.10.3. This is in the range of $3 to $10 per pound, as reported by TIAX 2009.163
Table 2-69 Estimated Effectiveness of Vocational Weight Reduction

HHD
MHD
LHD
Weight Reduction
250
0
150
0
150
0
Static Test Weight (kg)
18,994
19,051
11,374
11,408
7,223
7,257
Dynamic Test Weight (kg)
19,561
19,618
11,714
11,748
7,563
7,597
Payload (ton)
7.5625
7.5
5.6375
5.6
2.8875
2.85
Effectiveness over Transient
1.0%

0.8%

1.5%

Effectiveness over 55 mph
0.9%
0.7%
1.4%
Effectiveness over 65 mph
0.9%
0.7%
1.4%
Urban Cycle Effectiveness
1.0%
0.8%
1.5%
Multi-Purpose Cycle
Effectiveness
0.9%
0.8%
1.4%
Regional Cycle Effectiveness
0.9%
0.7%
1.4%

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-70 Vocational Weight Reduction Technologies
COMPONENT
MATERIAL
VOCATIONAL VEHICLE CLASS
Class 2b-5
Class 6-7
Class 8
Axle Hubs - Non-Drive
Aluminum
40
40
Axle Hubs - Non-Drive
High Strength Steel
5
5
Axle - Non-Drive
Aluminum
60
60
Axle - Non-Drive
High Strength Steel
15
15
Brake Drums - Non-Drive
Aluminum
60
60
Brake Drums - Non-Drive
High Strength Steel
42
42
Axle Hubs - Drive
Aluminum
40
80
Axle Hubs - Drive
High Strength Steel
10
20
Brake Drums - Drive
Aluminum
70
140
Brake Drums - Drive
High Strength Steel
37
74
Suspension Brackets, Hangers
Aluminum
67
100
Suspension Brackets, Hangers
High Strength Steel
20
30
Crossmember - Cab
Aluminum
10
15
15
Crossmember - Cab
High Strength Steel
2
5
5
Crossmember - Non-Suspension
Aluminum
15
15
15
Crossmember - Non-Suspension
High Strength Steel
5
5
5
Crossmember -Suspension
Aluminum
15
25
25
Crossmember -Suspension
High Strength Steel
6
6
6
Driveshaft
Aluminum
12
40
50
Driveshaft
High Strength Steel
5
10
12
Frame Rails
Aluminum
120
300
440
Frame Rails
High Strength Steel
40
40
87
Wheels - Dual
Aluminum
150
150
250
Wheels - Dual
High Strength Steel
48
48
80
Wheels - Wide Base Single3
Aluminum
294
294
588
Wheels - Wide Base Single3
High Strength Steel
168
168
336
Permanent 6x2 Axle Configuration
Multi
N/A
N/A
300
Note:
" Based on values from Table 6 of 40 CFR 1027.520 and use of four wide base singles on Class 8 vocational
vehicles and two on vehicles with one drive axle.
2.9.3.6 Electrified Accessories
Reducing the mechanical and electrical loads of accessories reduces the power
requirement of the engine and in turn reduces the fuel consumption and CO2 emissions.
Modeling in GEM, as shown in Table 2-71, demonstrates there is a measurable effect of
reducing 1 kW of accessory load for each vocational subcategory.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-71 Effect of Accessory Load Reduction on Vocational CO2 Emissions
VOCATIONAL
SUBCATEGORY
DIESEL (CI) PERCENT C02 PER
KW
GASOLINE (SI) PERCENT
C02 PER KW
HHDR
0.95%
-
HHDM
1.62%
-
HHDU
1.82%
-
MHDR
1.39%
1.28%
MHDM
2.62%
2.14%
MHDU
3.15%
2.48%
LHDR
2.00%
1.87%
LHDM
3.38%
2.91%
LHDU
3.95%
3.44%
Optimization and improved pressure regulation may significantly reduce the parasitic
load of the water, air and fuel pumps. Electrification may result in a reduction in power demand,
because electrically-powered accessories (such as the air compressor or power steering) operate
only when needed if they are electrically powered, but they impose a parasitic demand all the
time if they are engine-driven. In other cases, such as cooling fans or an engine's water pump,
electric power allows the accessory to run at speeds independent of engine speed, which can
reduce power consumption.
Some vocational vehicle applications have much higher accessory loads than is assumed
in the default GEM configurations. In the real world, there may be some vehicles for which
there is a much larger potential improvement available than those listed above, as well as some
for which electrification is not cost-effective. To date, accessory electrification has been
associated only with hybrids, although CalStart commented they are optimistic that accessory
electrification will become more widespread among conventional vehicles in the time frame of
Phase 2.
Electric power steering (EPS) or Electrohydraulic power steering (EHPS) provides a
potential reduction in CO2 emissions and fuel consumption over hydraulic power steering
because of reduced overall accessory loads. This eliminates the parasitic losses associated with
belt-driven power steering pumps which consistently draw load from the engine to pump
hydraulic fluid through the steering actuation systems even when the wheels are not being
turned. EPS is an enabler for all vehicle hybridization technologies since it provides power
steering when the engine is off. EPS is feasible for most vehicles with a standard 12V system.
Some heavier vehicles may require a higher voltage system which may add cost and complexity.
Although we did not propose to allow pre-defined credit for electrified accessories as was
proposed for tractors, we received comment requesting that this be allowed for vocational
vehicles. As discussed in 2.9.3.1 above, the agencies are projecting that some electrified
accessories will be necessary as part of the development of stop-start idle reduction systems for
vocational vehicles. The technology package for vocational stop-start includes costs for high-
efficiency alternator, electric water pump, electric cooling fan, and electric oil pump. However,
because the GEM algorithm for determining the fuel benefit of stop-start does not account for

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
any e-accessories, vehicles certified with stop-start are also eligible to be certified using an
improvement value in the e-accessories column.
Daimler, ICCT, Bendix, Gentherm, Navistar, Odyne, and CARB asked the agencies to
consider electric cooling fans, variable speed water pumps, clutched air compressors, electric air
compressors, electric power steering, electric alternators, and electric AJC compressors. ICCT
cautioned that certain accessories would be recognized over an engine test and credit should not
be duplicated at the vehicle level. Bosch suggested that high-efficiency alternators be
considered, and suggested use of a standard component-level test for alternators to determine
their efficiency, and establishment of a minimum efficiency level that must be attained.
Although there are industry-accepted test procedures for measuring the performance of
alternators, we do not have sufficient information about the baseline level performance of
alternators to define an improved level that would qualify for a benefit at certification. We are
not able to set a fixed improvement for electric cooling fans or clutched accessories due to
similar challenges related to baselines and defining the qualifying technology. In consideration
of ICCT's comment, we are not including water pumps and oil pumps among the components
eligible for a fixed improvement because we believe that our engine test procedure will
recognize improvements that would be seen in the real world from electrifying these parts. Thus,
we believe it is appropriate to offer fixed technology improvements for use of electric power
steering and an electric AJC compressor, as inputs to GEM.
The agencies have combined the GEM results shown in Table 2-71 with information
from comments provided by ICCT, the TIAX 2009 technology report, CARB's Driveline
Optimization report, the 2010 NAS report, and a 2014 article published in IET Electrical
Systems in Transportation to assign fixed improvement values for the defined technologies
shown in Table 2-72.164 These values are consistent with the TIAX study that used 2 to 4
percent fuel consumption improvement for accessory electrification, with the understanding that
electrification of accessories will have more effect in short haul/urban applications and less
benefit in line-haul applications.165
Table 2-72 Effectiveness of Vocational E-Accessories
TECHNOLOGY
EFFECTIVENESS
SUBCATEGORIES
Electric A/C Compressor
0.5%
HHD
1.0%
MHD & LHD
Electric Power Steering
0.5%
Regional
1.0%
Multipurpose & Urban
The improvement value for electric AJC compressors was estimated using a value of 4.7
kW demand from Table 5-11 of the 2010 NAS report, along with an assumption that it runs on
average 40 percent of the time, and that electrification reduces the total load to the engine by 40
percent. Combining these values with the GEM-derived values of percent CO2 per kW reduced
from Table 2-71, the improvement is estimated to be in the range of 0.5 to three percent
depending on the subcategory. The improvement value for electric power steering was estimated
using an average value of 2 kW demand from Table 5-11 of the 2010 NAS report, along with an
assumption that it runs on average 60 percent of the time, and that electrification reduces the
total load to the engine by 40 percent. Combining these values with the GEM-derived values of

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
percent CO2 per kW reduced from Table 2-71, the improvement is estimated to be in the range
of 0.5 to 1.5 percent depending on the subcategory. We have selected conservative values from
these results as fixed technology improvements.
The agencies estimate the total cost to electrify accessories as described above on a LHD
vocational vehicle to be $425 in MY 2021 and $369 in MY 2027. Scaling up, the costs for a
MHD vocational vehicle are estimated at $801 in MY 2021 and $697 in MY 2027, and the costs
for a HHD vocational vehicle are estimated at $1,603 in MY 2021 and $1,393 in MY 2027, as
described in RIA 2.11.10.2.
Manufacturers wishing to obtain credit for technologies that are more effective than we
have projected, or technologies beyond the scope of this defined technology improvement, may
apply for off-cycle credits.
2.9.3.7 Tire Pressure Systems
2.9.3.7.1 TPMS
The agencies did not propose to base the vocational vehicle standards on the performance
of tire pressure monitoring systems (TPMS). However, we received comment that we should
consider this technology. See discussion in Preamble Section III.D. 1 .b. In addition to comments
related to tractors and trailers, RMA commented that TPMS can also apply to the class 2b - 6
vehicles, and if the agencies add TPMS to the list of recognized technologies, that this choice
should also be made available to class 2b-6 vehicles. Bendix commented that TPMS is a proven
product, readily available from a number of truck, bus, and motorcoach OEMs. Autocar
commented that TPMS is useful for refuse truck applications. Tirestamp said that TPMS is ideal
for trucks and buses that are unable to apply ATIS due to difficulties plumbing air lines
externally of the axles. The agencies find these comments to be persuasive. As a result, we are
finalizing vocational vehicle standards that are predicated on the performance of TPMS in all
subcategories, including all custom chassis except emergency vehicles and concrete mixers.
Available information indicates that it is feasible to utilize TPMS on all vocational vehicles,
though systems for heavy vehicles in duty cycles where the air in the tires becomes very hot
must be ruggedized so that the sensors are protected from this heat. Such devices are
commercially available, though they cost more. To account for this in our analysis, we have
projected a lower adoption rate for TPMS in Urban vehicles than for Regional or Multipurpose
vehicles, rather than by increasing the cost and applying an equal adoption rate. We are
assigning a fixed improvement value in GEM for use of this technology in vocational vehicles of
one percent for Regional vehicles including motor coaches and RV's (the same as for tractors
and trailers) and 0.9 percent for Multipurpose, Urban, and other custom chassis vocational
vehicles, recognizing that the higher amount of idle is likely to reduce the effectiveness for these
vehicles. These values will be specified as GEM inputs in the column designated for tire
pressure systems. For HHD vocational vehicles (with 3 axles), the agencies estimate the cost of
TPMS at $583 in MY 2021 and $507 in MY 2027, as described in RIA 2.11.8.9. For LHD and
MHD vocational vehicles, we scaled down the cost of this technology to reflect the presence of a
single rear axle. The agencies estimate the total cost of TPMS on a LHD or MHD vocational
vehicle (with 2 axles) at $307 in MY 2021 and $267 in MY 2027.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.9.3.7.1.1 ATIS
The agencies did not propose to base the vocational vehicle standards on the performance
of automatic tire inflation systems (ATIS), otherwise known as central tire inflation (CTI).
However, we did receive comment indicating that it is feasible on some vocational vehicles,
specifically those which could choose to be certified as custom chassis. Air CTI commented that
central tire inflation is not only feasible but enhances safety on vehicles such as dump trucks and
heavy haul vehicles that need higher tire pressures under certain driving conditions, such as
when loaded, but need lower tire pressures when running empty or operating off-road.
Tirestamp commented that ATIS can be plumbed externally for trucks and buses, but such
systems have a propensity for damage and Autocar has provided information about how much
extra weight this plumbing adds to the chassis. ATA commented that some onboard air pressure
systems may not be able to pressurize tires sufficiently for very heavy vehicles. The primary
vocational vehicle standards are not predicated on any adoption of ATIS because the agencies do
not have sufficient information about which chassis will have an onboard air supply for purposes
of an air suspension or air brakes. ATIS would logically only be adopted for vehicles that
already need an onboard air supply for other reasons. Comments received for custom chassis
were supportive of standards predicated on ATIS for buses with air suspensions. These
comments are again persuasive. As a result, we are basing the optional standards for refuse
trucks, school buses, coach buses, and transit buses in part on the adoption of ATIS. Although
many motor homes have onboard air supply for other reasons making ATIS technically feasible,
it is sufficiently costly that it is not practically feasible. Furthermore, for the same reasons stated
above about the disadvantages of installing external plumbing for ATIS on some trucks and
buses, we have determined it is not feasible for emergency vehicles or concrete mixers.
Nonetheless, we are allowing any vocational vehicle to obtain credit for the performance of
ATIS through a GEM input with a fixed improvement value in GEM for use of this technology
in vocational vehicles of 1.2 percent for Regional vehicles including motor coaches and RV's
(the same as for tractors and trailers) and 1.1 percent for Multipurpose, Urban, and other custom
chassis vocational vehicles, recognizing that the higher amount of idle is likely to reduce the
effectiveness for these vehicles. These values will be specified as GEM inputs in the column
designated for tire pressure systems. See discussion in Preamble Section III.D. 1 .b for our
reasoning behind this effectiveness value. Because ATIS is not projected as a technology in the
basis for the mandatory vocational vehicle standards, we have not estimated detailed costs for
applying this technology on these vehicles. Even so, in RIA 2.11.8.8 (see Table 2-130), the
agencies estimate the cost of ATIS on 3-axle tractors to be $916 in MY 2021 and $796 in
MY2027. We would expect the cost to apply ATIS on a 3-axle vocational vehicle to be
comparable to these costs. Table 2-133 in RIA 2.11.8.8 presents costs the agencies have
estimated to apply ATIS on short van trailers; $481 in MY 2021 and $418 in MY 2027. We
would expect the cost to apply ATIS on a 2-axle vocational vehicle to be comparable to these
costs.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.9.3.8 HFC Leakage
Emissions due to direct refrigerant leakage are significant in all vehicle types. Since the
proposal, EPA has learned that the capacities of vocational vehicle air conditioning systems
range from those that are similar to those of other HD vehicles to some that are much larger.
Even considering these differences, we believe it is appropriate to apply a similar leakage
standard as was applied in the HD Phase 1 program for tractors and HD pickup trucks and vans.
EPA is adopting a 1.50 percent refrigerant leakage per year standard for each air conditioning
system with a refrigerant capacity greater than 733 grams, to assure that high-quality, low-
leakage components are used in the design of these systems. Since refrigerant leakage past the
compressor shaft seal is the dominant source of leakage in belt-driven air conditioning systems,
the agency recognizes that this 1.50 percent leakage standard is not feasible for systems with a
refrigerant capacity of 733 grams or lower, as the minimum feasible leakage rate does not
continue to drop as the capacity or size of the air conditioning system is reduced. The fixed
leakage from the compressor seal and other system devices results in a minimum feasible yearly
leakage rate. EPA does not believe that leakage reducing technologies will be available in MY
2021 to enable lower capacity systems to meet the percent per year standard, so we are adopting
a maximum gram per year leakage standard of 11.0 grams per year for vocational vehicle air
conditioning systems with a refrigerant capacity of 733 grams or lower, as was adopted in the
HD Phase 1 program for tractors and HD pickup trucks and vans.
The standard is derived from the vehicles with the largest system refrigerant capacity
based on the Minnesota GHG Reporting database.166 These are the same data on which the HD
Phase 1 HFC leakage standard was based.167
By requiring that all vocational vehicles achieve the leakage level of 1.50 percent per
year, roughly half of the vehicles in the 2010 data sample would need to reduce their leakage
rates, and an emissions reduction roughly comparable to that necessary to generate direct
emission credits under the light-duty vehicle program would result. See 75 FR at 25426-247.
However, no credits or trading flexibilities are available under this standard for heavy-duty
vocational vehicles. We believe that a yearly system leakage approach assures that high-quality,
low-leakage, components are used in each A/C system design, and we expect that manufacturers
will reduce A/C leakage emissions by utilizing improved, leak-tight components. Some of the
improved components available to manufacturers are low-permeation flexible hoses, multiple o-
ring or seal washer connections, and multiple-lip compressor shaft seals. The availability of low
leakage components in the market is being driven by the air conditioning credit program in the
light-duty GHG rulemaking (which applies to 2012 model year and later vehicles). EPA
believes that reducing A/C system leakage is both highly cost-effective and technologically
feasible. The cooperative industry and government Improved Mobile Air Conditioning (IMAC)
program has demonstrated that new-vehicle leakage emissions can be reduced by 50 percent by
reducing the number and improving the quality of the components, fittings, seals, and hoses of
the A/C system.168 All of these technologies are already in commercial use and exist on some of
today's A/C systems in other heavy-duty vehicles.
EPA is adopting the same compliance method for control of leakage from A/C systems in
vocational vehicles as was adopted for the HD Phase 1 HFC leakage standard. Under this
approach, manufacturers will choose from a menu of A/C equipment and components used in

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
their vehicles in order to establish leakage scores, which characterizes their AJC system leakage
performance and calculates the percent leakage per year as this score divided by the system
refrigerant capacity. The agencies estimate the total cost to apply low leakage AJC components
to a vocational vehicle to be $22 in MY 2021 and $20 in MY 2027, as described in RIA 2.11.4.1.
Consistent with the Light-Duty Vehicle Greenhouse Gas Emissions rulemaking, the
components of vocational vehicle AJC systems are being compared to a set of leakage reduction
technologies that is based closely on that being developed through IMAC and the Society of
Automotive Engineers (as SAE Surface Vehicle Standard J2727, August 2008 version).169 See
generally 75 FR at 25426. The SAE J2727 approach was developed from laboratory testing of a
variety of AJC related components, and EPA believes that the J2727 leakage scoring system
generally represents a reasonable correlation with average real-world leakage in new vehicles.
Like the IMAC approach, our approach associates each component with a specific leakage rate
in grams per year identical to the values in J2727 and then sums together the component leakage
values to develop the total AJC system leakage. As is currently done for other HD vehicles, for
vocational vehicles, the total AJC leakage score will then be divided by the total refrigerant
system capacity to develop a percent leakage per year value.
2.9.4 Other Vocational Vehicle Technologies the Agencies Considered
2.9.4.1 Vocational Aerodynamics
The agencies did not propose to include aerodynamic improvements as a basis for the
Phase 2 vocational vehicle standards. However, we did request comment on an option to allow
credits for use of aerodynamic devices such as fairings on a very limited basis. We received
public comments from AAPC in support of offering this as an optional credit, with a suggestion
to allow this option for a wide range of vehicle sizes, and suggesting that the grams per ton-mile
benefit could be scaled down for larger vehicles. CARB commented in support of a Phase 2
program that would include use of aerodynamic improvements as a basis for the stringency,
suggesting that a large fraction of the vocational vehicle fleet could see real world benefits from
use of aerodynamic devices. Because we do not have sufficient fleet information to establish a
projected application rate for this technology, we are not basing any of the final standards for
vocational vehicles on use of aerodynamic improvements. In consideration of comments, we are
adopting provisions for vocational vehicles to optionally receive an improved GEM result by
certifying use of a pre-approved aerodynamic device.
Based on testing supported by CARB, the agencies have developed a list of specific
aerodynamic devices with pre-defined improvement values (in delta CdA units), as well as
criteria regarding which vehicles are eligible to earn credit in this manner. Manufacturers
wishing to receive credit for other aerodynamic technologies or on other vehicle configurations
may apply for credits using the test procedures at 40 CFR 1037.527.
Table 2-73 shows the vocational aerodynamic technologies that we are adopting as pre-
approved, for which the credit listed is available through GEM. In response to comments, we are
allowing a wide range of vehicles to be eligible to use this option. Vocational vehicles in any
weight class over the Regional duty cycle may use this option, subject to restrictions on the size
of the cargo box (see 40 CFR 1037.520). The agencies have not estimated manufacturing costs

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
for these technologies on vocational vehicles. We project that a manufacturer would only apply
these where it was found to be cost-effective for the specific application. For a description of the
costs estimated for applying aerodynamic technologies to tractors, see the RIA at Chapter
2.11.9.1, where the estimated cost for a Bin2 package on a low roof day cab tractor is shown to
be roughly $1,000.
Table 2-73 Pre-approved Vocational Aerodynamic Technologies
VEHICLE
SKIRT
FRONT
FAIRING
(NOSE CONE)
REAR
FAIRING
(TAIL)
BOTH FRONT
FAIRING AND
SKIRT
Total chassis length at least 36 ft
and frontal area at least 9 m2
0.3
0.3

0.5
Total chassis length at least 23 ft
and frontal area at least 8 m2


0.2

A description of the testing that was conducted in support of the assigned GEM
improvements due to these technologies is presented in the draft report from NREL to CARB.170
The degree of change in CdA for each pre-approved device has been set at conservative values
due to the small number of configurations tested and the large uncertainty inherent in those
results. As an example of the degree of uncertainty, the change in CdA on the class 6 box truck
due to applying a chassis skirt was reported by NREL in Table 8 as being approximately -0.6 m2
with a 95 percent confidence interval of plus or minus -0.6 m2 Manufacturers using this credit
provision may enter the pre-defined delta CdA as an input to GEM, and the simulation will
determine the effectiveness over the applicable duty cycle. Using this approach, we do not need
to set a scaled benefit for different sizes of vehicles. When the vehicle weight class and duty
cycle are specified, a default chassis mass and payload are simulated in GEM. When the pre-
defined delta CdA is entered, the simulation returns a resulting improved performance with
respect to the specified chassis configuration. GEM will logically return a smaller improvement
for heavier vehicles.
The final Regional composite duty cycle in GEM for vocational vehicles has a weighted
average speed of 41.9 mph, increased from the average speed at proposal due to a heftier 56
percent composite weighting of the 65 mph drive cycle. The agencies have learned from the
NREL duty cycle analysis that vocational vehicles with operational behavior of a regional nature
accumulate more miles at highway speeds than previously assumed.
Using GEM simulation results, the agencies estimate the fuel efficiency benefit of
improving the CdA of a Class 6 box truck by 11 percent (0.6 m2 delta CdA off of a default of 5.4
m2) at approximately five percent over the Regional composite test cycle. This same delta CdA
simulated in GEM on a class 8 Regional vocational vehicle results in an overall improvement of
less than four percent because the default CdA in GEM for class 8 vocational vehicles is 6.86 m2
so the change in CdA is only nine percent. Although in actual operation the added weight of
aerodynamic fairings may reduce the operational benefits of these technologies when driving at
low speeds, the agencies are not applying any weight penalty as part of the certification process
for vocational aerodynamic devices.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
As described in the NPRM, we are requiring chassis manufacturers employing this option
to provide assurances to the agencies that these devices will be installed as part of the certified
configuration, even if the installation is completed by another entity. We received many
comments on the requirements for secondary manufacturers as they apply for vocational
aerodynamics as well as other technologies that may be specified by a chassis manufacturer but
installed later. See Preamble Section I.F.2 and Section V.D.2 for further discussion of delegated
assembly issues.
2.9.4.2 E-PTO
Although the primary program does not simulate vocational vehicles over a test cycle that
includes PTO operation, the agencies are adopting a revised hybrid-PTO test procedure. See 76
FR 57247 and 40 CFR 1037.540. Recall that we regulate vocational vehicles at the incomplete
stage when a chassis manufacturer may not know at the time of certification whether a PTO will
be installed or how the vehicle will be used. Chassis manufacturers may rarely know whether
the PTO-enabled vehicle will use this capability to maneuver a lift gate on a delivery vehicle, to
operate a utility boom, or merely to keep it as a reserve item to add value in the secondary
market. For these reasons, it would not be fair to require every vocational vehicle to certify to a
standard test procedure with a PTO cycle in it. Thus, we are not basing the final standards on
use of technology that reduces emissions in PTO mode.
There are products available today that can provide auxiliary power, usually electric, to a
vehicle that needs to work in PTO mode for an extended time, to avoid idling the main engine.
There are different designs of electrified PTO systems on the market today. Some designs have
auxiliary power sources, typically batteries, with sufficient energy storage to power an onboard
tool or device for a short period of time, and are intended to be recharged during the workday by
operating the main engine, either while driving between work sites, or by idling the engine until
a sufficient state of charge is reached that the engine may shut off. Other designs have sufficient
energy storage to power an onboard tool or device for many hours, and are intended to be
recharged as a plug-in hybrid at a home garage. In cases where a manufacturer can certify that a
PTO with an idle-reduction technology will be installed either by the chassis manufacturer or by
a second stage manufacturer, the hybrid-PTO test cycle may be utilized by the certifying
manufacturer to measure an improvement factor over the GEM duty cycle that otherwise applies
to that vehicle. In addition, the delegated assembly provisions will apply (see Section I.F.2).
See RIA Chapter 3.7.4 for a discussion of the revisions to the PTO test cycle.
The agencies will continue the hybrid-PTO test option that was available in Phase 1, with
a few revisions. See the regulations at 40 CFR 1037.540. The calculations recognize fuel
savings over a portion of the test that is determined to be charge-sustaining as well as a portion
that is determined to be charge-depleting for systems that are designed to power a work truck
during the day and return to the garage where recharging from an external source occurs during
off-hours. The agencies requested comment on this idea, and received comment from Odyne
relating to the population and energy storage capacity of plug-in e-PTO systems, for which a
charge-depleting test cycle may be more appropriate. We also partnered with DOE-NREL to
characterize the PTO operation of many vocational vehicles. NREL has characterized the PTO
operation using telematics data from Odyne on over 80 utility trucks with over 1,500 total
operating days, plus telematics data on ten utility trucks from PG&E with hundreds of operating

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
days. Our final regulations include a utility factor table based on these data for use in
determining the effectiveness of a hybrid PTO system. A description of the analysis underlying
the development of this utility factor curve is available in the docket.171 Manufacturers wishing
to conduct testing as specified may apply for off-cycle credits derived from e-PTO or hybrid
PTO technologies.
2.9.4.3 Electric Vehicles
Some heavy-duty vehicles can be powered exclusively by electric motors. Electric
motors are efficient and able to produce high torque, giving e-trucks strong driving
characteristics, particularly in stop-and-go or urban driving situations, and are well-suited for
moving heavy loads. Electric motors also offer the ability to operate with very low noise, an
advantage in certain applications. Currently, e-trucks have some disadvantages over
conventional vehicles, primarily in up-front cost, weight and range. Components are relatively
expensive, and storing electricity using currently available technology is expensive, bulky, and
heavy. However commenters provided information on total cost of ownership for electric trucks,
and some applications may see attractive long term cost scenarios for electric trucks or buses,
considering maintenance savings.
The West Coast Collaborative, a public-private partnership, has estimated the incremental
costs for electric Class 3-6 trucks in the Los Angeles, CA, area.172 Compared to a conventional
diesel, the WCC estimates a battery-electric vehicle system would cost between $70,000 and
$90,000 more than a conventional diesel system. The CalHEAT Technology Roadmap includes
an estimate that the incremental cost for a fully-electric medium- or heavy- duty vehicle would
be between $50,000 and $100,000. In the Draft RIA Chapter 2.12.7.6, the agencies estimated the
cost of a full electric LHD or MHD vocational vehicle at $55,216 in MY 2021 and $52,128 in
MY 2024. The CalHEAT roadmap report also presents several actions that must be taken by
manufacturers and others, before heavy-duty e-trucks can reach what they call Stage 3
Deployment.173
Early adopters of electric drivetrain technology are medium-heavy-duty vocational
vehicles that are not weight-limited and have drive cycles where they don't need to go far from a
central garage. According to CALSTART's report to the NAFA 2014 Institute and Expo, there
is an emerging market of MHD all-electric vocational vehicles, including models from Smith,
EVI, Boulder, AMP, and others. It is a significant stepping stone that we are seeing these
emerging markets, where prototype and demonstration vehicles can be tested and observed in
real world conditions. CalHEAT has published results of a comprehensive performance
evaluation of three battery electric truck models using information and data from in-use data
collection, on road testing and chassis dynamometer testing.174
Given the high costs and the developing nature of this technology, the agencies do not
project fully electric vocational vehicles to be widely commercially available in the time frame
of the final Phase 2 rules. For this reason, the agencies have not based the Phase 2 standards on
adoption of full-electric vocational vehicles. EEI provided information on the total cost of
ownership for electric trucks, where under certain conditions some vehicle applications may see
attractive long term cost scenarios for electric trucks or buses, when considering maintenance
savings. To the extent this technology is able to be brought to market in the time frame of the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Phase 2 program, there is currently a certification path for these chassis from Phase 1, as
described in the Preamble Section V and in the regulations at 40 CFR 1037.150 and 49 CFR
535.8.
2.9.5 Derivation of the Vocational Vehicle Technology Packages
The final standards for vocational vehicles are predicated on the same suite of
technologies in all implementation years of the Phase 2 program. The change in stringency
between those years is a result of different adoption rates of those technologies. Package costs
for each model year are presented following each respective adoption rate discussion.
2.9.5.1 Projected Technology Adoption Rates for Vocational Vehicles
The agencies describe below the extent to which technologies may be adopted by
manufacturers to meet each set of vocational vehicle standards.
2.9.5.1.1 Transmissions
Because we expect that transmission suppliers will be able to conduct a modest amount
of testing that can be valid for a large sales volume of transmissions, the agencies project an
adoption rate of 50 percent in MY 2021, 60 percent in MY 2024, and nearly 70 percent in MY
2027 of transmissions with improved gear efficiencies, with inputs over-riding the GEM defaults
obtained over the separate transmission efficiency test. In response to comments regarding the
diversity of drivelines and the narrow range of products for which powertrain testing is likely to
be conducted, we are projecting an adoption rate of 10 percent in MY 2021, 20 percent in MY
2024, and nearly 30 percent in MY 2027 of advanced shift strategies, with demonstration of
improvements recognized over the separate powertrain test. With additional time and research,
we expect that the adoption of this strategy for improving fuel efficiency will grow.
We are predicating the Phase 2 standards on zero adoption of added gears in the HHD
Regional subcategory, because it is modeled with a 10-speed transmission, and vehicles already
using that number of gears are not expected to see any real world improvement by increasing the
number of available gears. For the Multipurpose and Urban HHD subcategories, the MY 2021
projected adoption of adding gears is 5 percent, increasing to 10 percent for MY 2024 and MY
2027. We are projecting 10 percent of adding two gears in each of the other six subcategories
for MY 2021, increasing to 20 percent for MY 2024 and MY2027. Commenters supported the
inclusion of this technology as part of the basis for the standards. Allison commented that they
have configured an 8-speed vocational transmission. Eaton's new MHD dual clutch transmission
has seven forward gears. There is also a likelihood that suppliers of 8-speed transmissions for
HD pickups and vans may sell some into the LHD vocational vehicle market.
We are also predicating the optional custom chassis standards for school and coach buses
in part on adoption of transmissions with additional gears. In MY 2021, this adoption rate is five
percent, increasing to 10 percent in MY 2024 and 15 percent in MY 2027. Manufacturers who
certify these vehicles to the primary standards will use GEM to model the actual gears and gear
ratios. Manufacturers opting into the custom chassis program will not have this flexibility. The
agencies have estimated the cycle-average benefit of adding an extra gear for school buses

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
(modeled as MHD Urban vehicles) at 0.9 percent and coach buses (with 6 gears in the baseline)
at 1.7 percent, therefore manufacturers using the custom chassis regulatory subcategory
identifiers for these vehicles will be permitted to enter these pre-defined improvement values at
the time of certification.
Based on comment regarding our regulatory baselines, both the HHD Regional and HHD
Multipurpose subcategories now have manual transmissions in the baseline configuration. For
these vehicles, the agencies project upgrades to automated transmissions such as either AMT,
DCT, or automatic, at an adoption rate of 30 percent in MY 2021, 60 percent in MY 2024, and
90 percent in MY 2027 for Regional vehicles. For Multipurpose, beginning with 20 percent
manuals in the baseline, the adoption rate of automated transmissions is five percent in MY 2021
and 20 percent in MY 2024. Consistent with our projections of technology adoption, the
regulations require that any vocational vehicles with manual transmissions must be certified as
Regional in MY 2024 and beyond. This progression of transmission automation is consistent
with the agencies' projection of 10 percent manuals and 90 percent automated transmissions in
the day cab tractor subcategories in MY 2027. See Table III-13 of the Preamble. HHD
vocational vehicles in regional service have many things in common with day cab tractors,
including the same assumed engine size and typical transmission type, and a similar duty cycle.
Thus, it is reasonable for the agencies to make similar projections about the fraction of
automated vs manual transmissions adopted over the next decade among these sectors.
In the seven subcategories (i.e. all of the remaining subcategories) in which automatic
transmissions are the base technology, the agencies project that ten percent of the HHD vehicles
will apply an aggressive torque converter lockup strategy in MY 2021, and 30 percent in the
LHD and MHD subcategories. These adoption rates are projected to increase to 20 percent for
HHD and 40 percent for LHD and MHD in MY 2024. We project adoption of aggressive torque
converter lockup for HHD automatics of 30 percent in MY 2027, and 50 percent for LHD and
MHD. We project these adoption rates to be greater than that of the fully integrated shift strategy
and less than that of the transmission gear efficiency technologies because this is less complex to
apply and may be entered as a GEM input rather than requiring separate test procedures.
In setting the standard stringency, we have projected that non-integrated (bolt-on) mild
hybrids will not have the function to turn off the engine at stop, while the integrated mild hybrids
will have this function. The agencies have estimated the effectiveness of non-integrated mild
hybrids for vehicles certified in the Urban subcategories will achieve as much as 12 percent
improvement, and integrated systems that turn off at stop will see up to 25 percent improvement
in the Urban subcategories. We have also projected zero hybrid adoption rate (mild or
otherwise) by vehicles in the Regional subcategories, expecting that the benefit of hybrids for
those vehicles will be too low to merit use of that type of technology.
There is no fixed hybrid value assigned in GEM. Consequently, any vehicles utilizing
hybrid technology will determine the actual improvement by conducting powertrain testing.
By the full implementation year of MY 2027, the agencies are projecting an overall
vocational vehicle adoption rate of 12 percent mild hybrids, which we estimate will be 14
percent of vehicles certified in the Multi-Purpose and Urban subcategories (six percent integrated
and eight percent non-integrated). We are projecting a low adoption rate in the early years of the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Phase 2 program, zero integrated hybrid systems and two percent of the bolt-on systems in these
subcategories in MY 2021, and three percent integrated mild hybrids in MY 2024 for vehicles
certified in the Multi-Purpose and Urban subcategories, plus 5 percent non-integrated mild
hybrids in MY 2024. Based on our assumptions about the populations of vehicles in different
subcategories, these hybrid adoption rates are about two percent overall in MY 2021 and six
percent overall in MY 2024. With the revised projection of lower cost mild hybrids instead of
strong hybrids and more robust assessment of effectiveness than at proposal, we are confident
that we can project a slightly higher overall adoption rate than we had at proposal.
Navistar commented with concerns that the agencies may be double counting some of the
improvements of deep integration. For example, the addition of a gear to a transmission may
reduce the added benefit of deep integration, as the transmission may already achieve a more
optimal operation state more often due to the greater number of gears. The agencies have been
careful to project adoption rates and effectiveness of transmission technologies in a way that that
avoids over-estimating the achievable reductions. For example, as we developed the packages,
we reduced the adoption rate of advanced shift strategy by the adoption rate of integrated
hybrids, and we reduced the adoption rate of transmission gear efficiency by the amount of non-
integrated hybrids. This means that in the HHD Multipurpose category in MY 2027, the sum of
adoption rates of hybrids, advanced shift strategy, and transmission gear efficiency is 100
percent. Further, instead of summing the combined efficiencies, we combine multiplicatively as
described in Equation 2-2, below. Transmission improvements are central to the Phase 2
vocational vehicle program, second only to idle reduction. We are projecting that many vehicles
will apply more than one technology that improves vehicle performance with respect to the
transmission, which necessarily means that the adoption rate of transmission technologies in
some subcategories sums to greater than 100 percent. For example, with a 50 percent adoption of
torque converter lockup and a 70 percent adoption of high efficiency gearbox for Regional
vehicles in MY 2027, some vehicles may need to - and could reasonably - apply both. However,
we believe we have fairly accounted for dis-synergies of effectiveness where technologies are
applied to a similar vehicle system.
Custom chassis manufacturers have provided compelling comment that the absence of
recognition in the certification process of improved transmission technology will not deter them
from its adoption. Therefore, although some types of improved transmissions are feasible for
some custom chassis, the fact that these vehicles are typically assembled from off-the-shelf parts
in low production volumes makes them much less likely to have access to the most advanced
transmission technologies. Further, for the reasons described above about non-representative
drivelines in the baseline configurations, we believe that allowing these to be certified with a
default driveline is a reasonable program structure. For school buses and others, if a
manufacturer wishes to be recognized beyond the levels described for adopting improved
transmissions, it may optionally certify to the primary standards.
2.9.5.1.2 Axles
The agencies project that 10 percent of vocational vehicles in all subcategories will adopt
high efficiency axles in MY 2021, 20 percent in MY 2024, and 30 percent in MY 2027, and the
standards are predicated on these penetration rates for high efficiency axles. Fuel efficient
lubricant formulations are widespread across the heavy-duty market, though advanced synthetic

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
formulations are currently less popular.N Axle lubricants with improved viscosity and
efficiency-enhancing performance are projected to be widely adopted by manufacturers in the
time frame of Phase 2. Such formulations are commercially available and the agencies see no
reason why they could not be feasible for most vehicles. Nonetheless, we have refrained from
projecting full adoption of this technology. The agencies do not have specific information
regarding reasons why axle manufacturers may specify a specific type of lubricant over another,
and whether advanced lubricant formulations may not be recommended in all cases. The
agencies received adverse comment on allowing fixed credit for use of high efficiency axles,
whether from lubrication or other mechanical designs. In response, we are adopting a separate
axle efficiency test, which can be used as an input to GEM to over-ride default axle efficiency
values. The low overall adoption rate indicates that we expect axle suppliers to only offer high-
efficiency axles for their most high production volume products, especially those that can serve
both the tractor and vocational market. Therefore, we believe it is unlikely that high-efficiency
axles will be adopted in custom chassis applications. Because we are no longer offering a fixed
improvement for this technology as at proposal, this is only available for vocational vehicles that
are certified to the primary program.
The agencies estimate that 10 percent of HHD Regional vocational vehicles and five
percent of HHD Multipurpose vehicles will adopt part time 6x2 axle technology in MY 2021.
This technology is most likely to be applied to Class 8 vocational vehicles (with 2 rear axles) that
are designed for frequent highway trips. The agencies project a 20 percent for HHD Regional
and 15 percent adoption rate for HHD Multipurpose for part time 6x2 axle technologies in MY
2024. In MY 2027, we project 30 percent adoption of part time 6x2 for HHD Regional and 25
percent for HHD Multipurpose. We are establishing a baseline configuration for coach buses
with a 6x2 axle. If a HHD coach bus is sold with a 6x4 or part time 6x2 axle, the manufacturer
must enter the as-built axle configuration as a GEM input. This is true whether the vehicle is in
the primary program or if it is certified to the custom chassis standard.
2.9.5.1.3 Lower Rolling Resistance Tires
The agencies estimate that the per-vehicle average level of rolling resistance from
vocational vehicle tires could be reduced by up to 13 percent for many vehicles by full
implementation of the Phase 2 program in MY 2027, based on broader adoption of vocational
vehicle tires currently available. We estimate this will yield reductions in fuel use and CO2
emissions of up to 3.3 percent for these vehicles. As proposed, the Phase 2 weighting of steer
tire CRR and drive tire CRR is 0.3 times the steer tire CRR and 0.7 times the drive tire CRR,
representing an average weight distribution of the rear axle(s) carrying 2.3 times the weight of
the front axle. The projected adoption rates of tires with improved CRR for chassis in the
primary program are presented in Table 2-74. The levels lv through 5v noted in the table are
defined in Section V.C. 1 .a.iv of the Preamble. By applying the assumed axle load distribution,
the estimated vehicle CRR improvement projected as part of the MY 2021 standards ranges from
5 to 8 percent, which we project will achieve up to 1.9 percent reduction in fuel use and CO2
emissions, depending on the vehicle subcategory. The agencies estimate the vehicle CRR
N April 2014 meeting with Dana.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
improvement in MY 2024 will range from 5 to 13 percent, yielding reductions in fuel use and
CO2 emissions up to 3.2 percent, depending on the vehicle subcategory.
The agencies believe that these tire packages recognize the variety of tire purposes and
performance levels in the vocational vehicle market, and maintain choices for manufacturers to
use the most efficient tires (i.e. those with lowest rolling resistance) only where it makes sense
given these vehicles' differing purposes and applications.
Table 2-74 Projected LRR Tire Adoption Rates

REGIONAL
MULTIPURPOSE
URBAN

Steer
Drive
Steer
Drive
Steer
Drive
2021 HHD
100% LRR
5v
100% LRR
2v
100% LRR 5v
100% LRR 2v
100%
LRR4v
100% LRR lv
2021 MHD
100% LRR
3v
100% LRR
lv
100% LRR 3v
100% LRR lv
100%
LRR 3v
100% LRR lv
2021 LHD
100% LRR
3v
100% LRR
3v
100% LRR 3v
100% LRR 3v
100%
LRR2v
100% LRR 2v
2024 HHD
100% LRR
5v
100% LRR
3v
100% LRR 5v
100% LRR 2v
100%
LRR4v
100% LRR lv
2024 MHD
100% LRR
5v
100% LRR
3v
100% LRR 3v
50% LRR lv,
50% LRR 2v
100%
LRR 3v
100% LRR lv
2024 LHD
100% LRR
5v
100% LRR
3v
100% LRR 3v
100% LRR 3v
100%
LRR2v
100% LRR 2v
2027 HHD
100% LRR
5v
100% LRR
3v
100% LRR 5v
100% LRR 3v
100%
LRR 5v
100% LRR 2v
2027 MHD
100% LRR
5v
100% LRR
3v
100% LRR 5v
100% LRR 3v
100%
LRR 3v
50% LRR lv,
50% LRR 2v
2027 LHD
100% LRR
5v
100% LRR
3v
100% LRR 5v
100% LRR 3v
100%
LRR 3v
50% LRR 2v,
50% LRR 3v
Table 2-75 presents the projected adoption rates of LRR tires for custom chassis. As
noted in Section V.C.I.a of the Preamble, the adoption rates generally represent improvements in
the range of the 25th to 40th percentile using data from actual vehicles in each application that
were certified in MY 2014. A summary of these data is provided in a memorandum to the
docket.175 An exception to this is emergency vehicles. The final emergency vehicle standards
reflect adoption of tires that progress to the 50th percentile by MY 2027, using steer and drive tire
data for certified emergency vehicles. At these adoption rates, manufacturers need not change
any of the tires they are currently fitting on emergency vehicles, and they will comply on
average.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-75 Projected LRR Tire Adoption Rates for Custom Chassis

MY 2021
MY 2024
MY 2027

Steer
Drive
Steer
Drive
Steer
Drive
Coach
100% LRR
4v
100% LRR
4v
100% LRR 5v
100% LRR 5v
100% LRR
5v
100% LRR 5v
RV
100% LRR
5v
100% LRR
5v
100% LRR 5v
100% LRR 5v
100% LRR
5v
100% LRR 5v
School
100% LRR
4v
100% LRR
2v
100% LRR 5v
100% LRR 3v
100% LRR
5v
100% LRR 4v
Transit
100% LRR
lv
100% LRR
lv
100% LRR lv
100% LRR lv
100% LRR
3v
100% LRR 3v
Refuse
100% LRR
lv
100% LRR
lv
100% LRR lv
100% LRR lv
100% LRR
3v
100% LRR 3v
Mixer
100% LRR
2v
100% LRR
lv
100% LRR 3v
100% LRR lv
100% LRR
3v
100% LRR 2v
Emerge
ncy
100% LRR
2v
100% LRR
lv
100% LRR 3v
100% LRR lv
100% LRR
4v
100% LRR lv
2.9.5.1.4 Workday Idle Reduction
In these rules, the adoption rate of AES for HHD Regional vehicles is 40 percent in MY
2021, 80 percent in MY 2024, and 90 percent in MY 2027. This is because these vehicles have
driving patterns with a significant amount of parked idle, and the vast majority have relatively
modest accessory demands such that only a few would have such large demands for backup
power that turning the engine off while parked would not be feasible. For all weight classes of
Regional vehicles except coach buses, the neutral idle and stop start adoption rates remain zero
in all model years because these vehicles have driving patterns with such a small amount of
transient driving that this drive-idle technology would not provide real world benefits. The LHD
and MHD weight class Regional vehicles carry a 30 percent, 60 percent, and 70 percent adoption
rate of AES in MYs 2021, 2024, and 2027 respectively. The adoption rates of idle reduction
technologies for vocational vehicles in MY 2027 are presented in Table 2-76.
Table 2-76 MY 2027 Adoption Rates of Idle Reduction Technologies

Heavy Heavy-Duty
Medium Heavy-Duty
Light Heavy-Duty
Technology
Regional
Multi-
purpose
Urban
Regional
Multi-
purpose
Urban
Regional
Multi-
purpose
Urban
Neutral Idle
0
70
70
0
60
60
0
60
60
Stop-Start
0
20
20
0
30
30
0
30
30
AES
90
70
70
90
70
70
90
70
70
Although it is possible that a vehicle could have both neutral idle and stop-start, our
stringency calculations only consider emissions reductions where a vehicle either has one or the
other of these technologies. The final GEM input file allows users to apply multiple idle
reduction technologies within a single vehicle configuration.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Because we have included costs to maintain engine protection during periods of shut-off,
as well as over-rides to recognize instances where it may not be safe to shut off an engine, we
believe stop-start can safely be applied at the rates described above in the time frames described.
Also, because we have defined two idle cycles where the automatic engine shutoff technology
addresses the condition of being parked with the service brake off, we believe this alleviates
many of the concerns expressed by commenters about stop-start. We believe many commenters
were (erroneously) imagining that stop-start systems would be required to function during
periods of extended parking.
We agree with commenters that stop-start is not feasible for emergency vehicles and
concrete mixers. We further believe that stop-start would not provide any real world benefit for
coach buses or motor homes. However, for school buses, transit buses, and refuse trucks, we
believe stop-start is feasible and likely to result in real world benefits. The only custom chassis
standards reflecting adoption of AES is school buses, because for the others, we believe the
simple shutdown timer would be likely to trigger an over-ride condition frequently enough to
yield a very small benefit from this technology. To make AES practical for a coach or transit
bus, for example, a much larger auxiliary power source would be needed than the one projected
as part of this rulemaking. We have based the school bus standards in part on adoption of AES,
however. Although many school buses have voluntarily adopted idle reduction strategies for
other reasons, we do not believe many have tamper-proof automatic shutdown systems. The
adoption rates of idle reduction technologies for custom chassis are presented in Table 2-77.
Table 2-77 Custom Chassis Workday Idle Adoption Rates
Technology
MY
AES
NI
Stop-Start
Coach
2021
-
40
-
2027
-
70
-
School
2021
30
60
5
2027
70
60
30
Transit
2021
-
60
10
2027
-
70
30
Refuse
2021
-
30
0
2027
-
50
20
As described above, refuse trucks that do not compact waste are ineligible for the
optional custom chassis vocational vehicle standards. We believe trucks that do not compact
waste have sufficiently low PTO operation (usually only while parked) to make application of
drive idle reduction technologies quite feasible. Front-loading refuse collection vehicles tend to
have a relatively low number of stops per day as they tend to collect waste from central locations
such as commercial buildings and apartment complexes. Because these have a relatively low
amount of PTO operation, we expect stop-start will be reasonably effective for these vehicles.
Rear-loading and side-loading neighborhood waste and recycling collection trucks are the refuse
trucks where the largest number of stop-start and neutral idle over-ride conditions are likely to be
encountered. Because chassis manufacturers, even those with small production volumes and
close customer relationships, do not always know whether a refuse truck chassis will be fitted
with a body designed for front loading, rear loading, or side loading, we are applying an adoption

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
rate of 20 percent stop-start in 2027 to refuse trucks certified as custom chassis. Chassis
manufacturers certifying refuse trucks to the optional custom chassis standards may enter Yes in
the input field in GEM for stop-start and the effectiveness will be computed based on the default
350 hp engine with 5-speed HHD automatic transmission. In the case where a chassis
manufacturer certifies a refuse truck to the primary standards under the HHD Urban subcategory,
the MY 2027 adoption rate is also 20 percent, and the stringency assumes a sufficiently capable
stop-start system to not require an excessive use of over-rides. Manufacturers opting to certify
refuse trucks to the primary standards will have an option to be recognized for enhanced stop-
start systems through the powertrain test.
It may take some minor development effort to apply neutral idle to high-torque automatic
transmissions designed for the largest vocational vehicles. Based on stakeholder input, the
designs needed to avoid an uncomfortable re-engagement bump when returning to drive from
neutral may require some engineering refinement as well as some work to enable two-way
communication between engines and transmissions. Nonetheless, this technology should be
available in the near term for many vehicles and is low cost compared to many other
technologies we considered. Commenters asked for over-rides such as when on a steep hill, and
we agree and are adopting this provision.
We see the above idle reduction technologies being technically feasible on the majority
of vocational vehicles, and especially effective on those with the most time in drive-idle in their
workday operation. Although we are not prepared to predict what fraction of vehicles will adopt
stop-start in the absence of Phase 2, the agencies are confident that this technology, which is on
the entry-level side of the hybrid and electrification spectrum, will be widely available in the
Phase 2 time frame.
Based on these projected adoption rates and the effectiveness values described above in
this section, we expect overall GHG and fuel consumption reductions from workday idle on
vocational vehicles to range from one to 13 percent in MY 2027.
2.9.5.1.5 Weight Reduction
As described in the RIA Chapter 2.11.10.3, weight reduction is a relatively costly
technology, at approximately $3 to $10 per pound for a 200-lb package. Even so, for vehicles in
service classes where dense, heavy loads are frequently carried, weight reduction can translate
directly to additional payload. The agencies project that modest weight reduction is feasible for
all vocational vehicles. The agencies are predicating the final standards on adoption of weight
reduction comparable to what can be achieved through use of aluminum wheels. This package is
estimated at 150 pounds for LHD and MHD vehicles, and 250 pounds for HHD vehicles, based
on six and 10 wheels, respectively. In MY 2021, we project an adoption rate of 10 percent, 30
percent in MY 2024, and 50 percent in MY 2027.
The agencies project that manufacturers will have sufficient options of other components
eligible for material substitution so that this level of weight reduction will be feasible even where
aluminum wheels are not selected by customers. Based on comments, we have removed
aluminum transmission cases and aluminum clutch housings from the vocational lookup table in
the regulations at 40 CFR 1037.520.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The only custom chassis standards on which we are predicating the standard on use of
weight reduction is transit buses. In addition to compelling comment from UCS, we considered
information from a 2014 study conducted by the APTA, where researchers found that fewer than
half of all transit bus models comply with a 20,000 pound single axle weight limit when empty
(i.e., at curb weight) and nearly all rear axles on transit buses longer than 35 feet exceed 24,000
pounds. According to APTA, the transit bus manufacturing industry has undertaken significant
research and development activities directed at decreasing the curb weight of transit buses, and
future opportunities to reduce transit bus curb weight include the use of lighter weight materials
and alternative manufacturing techniques, but any weight reductions are expected to be costly for
the manufacturing industry.176 Because overloaded axles is a significant issue for transit buses,
we believe it is appropriate for these rules to recognize it and provide a regulatory driver for
lightweighting in this sector.
We have learned that manufacturers of concrete mixers, refuse trucks, and some high end
buses have already made extensive use of lightweighting technologies in the baseline fleet. We
also received persuasive comment cautioning us not to base the school bus standards on weight
reduction due to potential conflicts with safety standards. In considering this information, we are
allowing all vehicles certified using custom chassis regulatory subcategory identifiers to make
use of weight reduction as a compliance flexibility, but only predicating standard stringency for
transit buses on use of aluminum wheels at the same adoption rate as for the primary program.
2.9.5.1.6	Electrified Accessories
The agencies are predicating the final vocational vehicle standards in part on an adoption
rate of five percent in MY 2021 of an electrified accessory package that achieves one percent
fuel efficiency improvement. The previous discussion in Chapter 2.9.3.6 describes some pre-
defined e-accessory improvements that are available in GEM for all vocational vehicles. In MY
2024 we increase this adoption rate to ten percent, and in MY 2027 the projected adoption rate is
15 percent, applicable in all subcategories excluding custom chassis. Although we believe some
components could be electrified for some custom chassis, we do not have sufficient information
to estimate an incremental cost associated with electrifying the more complex systems on custom
chassis such as buses, or to project a specific adoption rate for this type of improvement.
2.9.5.1.7	Tire Pressure Systems
The agencies are predicating the vocational vehicle standards in part on widespread
adoption of tire pressure monitoring systems. These are readily accepted by fleets as a cost-
effective safety and fuel-saving measure. Because there may be some minor challenges in
applying this technology to some vehicles where the payload and duty cycle lead to very high
tire temperatures and pressures (as described above), we are applying a lower adoption rate to
Urban and Multi-purpose vehicles than to Regional vehicles, as shown in Table 2-78. We are
applying similarly lower adoption rates for refuse trucks and transit buses. We are not
predicating the emergency vehicle or cement mixer standards on adoption of TPMS.
We are predicating the optional school bus, coach bus, transit bus, and refuse truck
standards in part on limited adoption of automatic tire inflation systems (ATIS), as shown in

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-78. These are more costly than TPMS, and require an onboard air supply and sometimes
extensive plumbing of air lines.
Table 2-78 Vocational Tire Pressure System Adoption Rates
Technology
TPMS
ATIS

MY 2021
MY 2024
MY
2027
MY
2021
MY
2024
MY
2027
Regional
60
75
90
-
-
-
Multi-Purpose
50
65
80
-
-
-
Urban
40
55
70
-
-
-
School
70

80
-

20
Coach
50

75
10

25
Transit
40

50
10

20
Refuse
40

50
10

15
Motor Home
60

90
-

-
2.9.5.1.8 HFC L eakage
We project 100 percent adoption rate in all implementation years of the Phase 2 program
for use of low leakage air conditioning system components to reduce direct emissions of HFC
compounds from vocational vehicles.
2.9.6 Vocational Vehicle Standards
The derivation of the vocational vehicle standards incorporates several methods because
some GEM inputs lend themselves to fleet-average values, some are vehicle specific (either on
or off) and some improvements are not directly modeled in GEM. For each model year of
standards, the agencies derived a scenario vehicle for each subcategory using the future model
year engine map with fleet average input values for tire rolling resistance and weight reduction.
For example, the MY 2021 HHD weight reduction input value was derived as follows: 210
pounds times 10 percent adoption yields 21 pounds. Those scenario vehicle performance results
were combined in a post-process method with subcategory-specific improvements from idle
reduction, axle disconnect, torque converter lockup, and transmission automation, using directly
modeled GEM improvements comparing results with these technologies on or off the scenario
vehicle. Subsequently, these performance values were combined with estimated improvement
values of technologies not modeled in GEM, including TPMS, hybrids, and transmission gear
efficiency.
The set of fleet-average inputs for tire CRR and weight reduction for MY 2021, as
modeled in GEM is shown in Table 2-79, along with the respective adoption rates for idle
reduction, axle disconnect, and torque converter lockup. The agencies derived the level of the
MY 2024 standards by using the GEM inputs and adoption rates shown in Table 2-80 below.
The agencies derived the level of the MY 2027 standards by using the GEM inputs and adoption
rates shown in Table 2-81, below. Post-processing improvements for technologies not directly
modeled, including TPMS, e-accessories, hybrids, and axle and transmission improvements are

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
presented as a combined driveline improvement factor in Table 2-82 below. The values in this
table for Si-powered vocational vehicles include improvements due to adoption of SI engine
technology.
After obtaining individual GEM performance values for each of the subcategories, the
agencies conducted fleet-mix averaging described in the Preamble in Section V.C. The resulting
final vocational vehicle standards are presented in Table 2-83 through Table 2-88.
Table 2-79 GEM Inputs Used to Derive MY 2021 Vocational Vehicle Standards
CLASS 2B-5
CLASS 6-7
CLASS 8
Urban
Multi-
purpose
Regional
Urban
Multi-
purpose
Regional
Urban
Multi-Purpose
Regional
SI Engine



2018 MY 6.8L, 300 hp engine



CI Engine
2021 MY 7L, 200 hp Engine
2021 MY 7L, 270 hp Engine
2021 MY
11L, 350
hp
Engine
2021 MY 11L, 350 hp
Engine and 2021 MY 15L
455hp Engine3
Torque Converter Lockup in 1st (adoption rate)
30%
30%
30%
30%
30%
30%
10%
10%
0%
6x2 Disconnect Axle (adoption rate)
0%
0%
0%
0%
0%
0%
0%
5%
10%
AES (adoption rate)
30%
30%
40%
30%
30%
40%
30%
30%
40%
Stop-Start (adoption rate)
10%
10%
0%
10%
10%
0%
0%
0%
0%
Neutral Idle (adoption rate)
50%
50%
0%
50%
50%
0%
50%
50%
0%
Steer Tires (CRR kg/metric ton)
7
6.8
6.8
6.8
6.7
6.7
6.4
6.2
6.2
Drive Tires (CRR kg/metric ton)
7.2
6.9
6.9
7.8
7.5
7.5
7.8
7.5
7.5
Weight Reduction (lb)
15
15
15
15
15
15
25
25
25
Note:
a The Multipurpose and Regional HHD standards are established using averages of configurations with different
engines as described in Table 2-55.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-80 GEM Inputs Used to Derive MY 2024 Vocational Vehicle Standards
CLASS 2B-5
CLASS 6-7
CLASS 8
Urban
Multi-
purpose
Regional
Urban
Multi-
purpose
Regional
Urban
Multi-Purpose
Regional
SI Engine



2018 MY 6.8L, 300 hp engine



CI Engine
2024 MY 7L, 200 hp Engine
2024 MY 7L, 270 hp Engine
2024 MY
11L, 350
hp
Engine
2024 MY 11L, 350 hp
Engine and 2024 MY 15L
455hp Engine3
Torque Converter Lockup in 1st (adoption rate)
40%
40%
40%
40%
40%
40%
20%
20%
0%
6x2 Disconnect Axle (adoption rate)
0%
0%
0%
0%
0%
0%
0%
15%
20%
AES (adoption rate)
60%
60%
80%
60%
60%
80%
60%
60%
80%
Stop-Start (adoption rate)
20%
20%
0%
20%
20%
0%
10%
10%
0%
Neutral Idle (adoption rate)
70%
70%
0%
70%
70%
0%
70%
70%
0%
Steer Tires (CRR kg/metric ton)
7.0
6.8
6.2
6.8
6.7
6.2
6.4
6.2
6.2
Drive Tires (CRR kg/metric ton)
7.2
6.9
6.9
7.8
7.5
6.9
7.8
7.5
6.9
Weight Reduction (lb)
45
45
45
45
45
45
75
75
75
Note:
a The Multipurpose and Regional HHD standards are established using averages of configurations with different
engines as described in Table 2-55.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-81 GEM Inputs Used to Derive MY 2027 Vocational Vehicle Standards
CLASS 2B-5


CLASS 6-7



CLASS


Urban
Multi-
purpose
Regional
Urban
Multi-
purpose
Regional
Urban
Multi-
purpose
Regional
SI Engine
2018 MY 6.8L, 300 hp engine
CI Engine
2027 MY 7L, 200 hp Engine
2027 MY 7L, 270 hp Engine
2027 MY
11L, 350
hp
Engine3
2027 MY 11L, 350 hp
Engine and
2027 MY 15L 455hp
Engine3
Torque Converter Lockup in 1st (adoption rate)
50%
50%
50%
50%
50%
50%
30%
30%
0%
6x2 Disconnect Axle (adoption rate)
0%
0%
0%
0%
0%
0%
0%
25%
30%
AES (adoption rate)
70%
70%
90%
70
70%
90%
70%
70%
90%
Stop-Start (adoption rate)
30%
30%
0%
30%
30%
0%
20%
20%
0%
Neutral Idle (adoption rate)
60%
60%
0%
60%
60%
0%
70%
70%
0%
Steer Tires (CRR kg/metric ton)
6.8
6.2
6.2
6.7
6.2
6.2
6.2
6.2
6.2
Drive Tires (CRR kg/metric ton)
6.9
6.9
6.9
7.5
6.9
6.9
7.5
6.9
6.9
Weight Reduction (lb)
75
75
75
75

75

75
125

125

125
Note:
a The Multipurpose and Regional HHD standards are established using averages of configurations with different
engines as described in Table 2-55.
In applying improvements due to technologies that were directly simulated in GEM but
required post-processing to account for adoption rates less than 100 percent, each improvement
was applied multiplicatively to the performance of the scenario vehicle that already had the
improved tires, weight, and engine. The formula used follows the pattern illustrated in Equation
2-2. Similarly, the improvements due to technologies not modeled in GEM were included in this
equation as noted. As described above in Chapter 2.9.3.1 for applicable technologies, the
agencies used an energy-weighted and cycle-weighted average estimating method using cycle-
specific CO2 emissions reported in the GEM output file for baseline vehicles. For the idle
cycles, the development version of GEM provides emissions in grams per hour. For the driving
cycles, GEM provides emissions in grams per ton-mile. By multiplying those values by the
average speed of each cycle and the default payload, GEM outputs in grams per ton-mile for the
driving cycles are converted to grams per hour, and these are surrogates for the energy expended
over those cycles. For example, in the medium heavy-duty Multipurpose subcategory with a
payload of 5.6 tons, the baseline vehicle configuration has cycle-specific results of 284 g
CCh/ton-mile for the transient cycle, 202 g CCh/ton-mile for the 55 cycle, 243 g CCh/ton-mile
for the 65 cycle, 10,226 g/hr for drive idle, and 5,284 g/hr for parked idle. By summing the
products of the percent improvement expected over each cycle, the CO2 emitted while

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
completing the cycle, and the associated composite weighting of the cycle, and dividing by the
sum of the products of the CO2 emitted and cycle weightings, we obtain subcategory-specific
improvement values for each technology. The complete set of calculations is available in the
docket.177
Equation 2-2: Additional percent improvement beyond engine, tires, weight:
l-((l-DIF)*(l-AESa*AESe)*(l-NIa*NIe)*(l-SSa*SSe)*(l-NMTa*NMTe)*(l-TLa*TLe)*(l-ADa*ADe))
Where:
•	DIF is the driveline improvement factor derived using engineering calculations,
not directly modeled in GEM
•	AESa and AESe are the adoption rate and effectiveness, respectively, in percent,
of automatic engine shutdown, as modeled in GEM
•	NIa and NIe are the adoption rate and effectiveness, respectively, in percent, of
neutral idle, as modeled in GEM
•	SSa and SSe are the adoption rate and effectiveness, respectively, in percent, of
stop-start, as modeled in GEM
•	NMTa and NMTe are the adoption rate and effectiveness, respectively, in percent,
of a non-manual transmission, as modeled in GEM
•	TLa and TLe are the adoption rate and effectiveness, respectively, in percent, of
torque converter lockup in first gear, as modeled in GEM
•	ADa and ADe are the adoption rate and effectiveness, respectively, in percent, of
axle disconnect, as modeled in GEM
Table 2-82 Vocational Driveline Improvement Factors

Class 2b-5
Class 6-7
Class 8

Urban
Multi-
purpose
Regional
Urban
Multi-
purpose
Regional
Urban
Multi-
purpose
Regional
CI 2021
0.019
0.018
0.018
0.019
0.019
0.019
0.019
0.018
0.017
CI 2024
0.041
0.036
0.029
0.041
0.036
0.029
0.040
0.036
0.026
CI 2027
0.061
0.053
0.037
0.061
0.053
0.037
0.060
0.052
0.034
SI 2021
0.027
0.026
0.026
0.028
0.027
0.027



SI 2024
0.048
0.044
0.037
0.049
0.044
0.037



SI 2027
0.067
0.059
0.045
0.068
0.060
0.045



Table 2-83 and Table 2-84 present EPA's CO2 standards and NHTSA's fuel consumption
standards, respectively, for chassis manufacturers of Class 2b through Class 8 vocational
vehicles for the beginning model year of the program, MY 2021.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-83 EPA CO2 Standards for MY2021 Class 2b-8 Vocational Vehicles
EPA Standard For Vehicle With CI Engine Ef
bctive MY2021 (gram C02/ton-mile)
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7
Heavy Heavy-Duty
Class 8
Urban
424
296
308
Multi-Purpose
373
265
261
Regional
311
234
205
EPA Standard for Vehicle with SI Engine Effective MY2021 (gram CCh/ton-mile)
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7 (and C8
Gasoline)

Urban
461
328

Multi-Purpose
407
293

Regional
335
261

Table 2-84 NHTSA Fuel Consumption Standards for MY2021 Class 2b-8 Vocational Vehicles
NHTSA STANDARD FOR VEHICLE WITH CI ENGINE EFFECTIVE MY 2021 (FUEL
CONSUMPTION GALLON PER 1,000 TON-MILE)
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7
Heavy Heavy-Duty
Class 8
Urban
41.6503
29.0766
30.2554
Multi-Purpose
36.6405
26.0314
25.6385
Regional
30.5501
22.9862
20.1375
NHTSA Standard for Vehicle with SI Engine
1,000 ton-mile)
iffective MY 2021 (Fuel Consumption gallon per
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7 (and C8
Gasoline)

Urban
51.8735
36.9078

Multi-Purpose
45.7972
32.9695

Regional
37.6955
29.3687

EPA's vocational vehicle CO2 standards andNHTSA's fuel consumption standards for
the MY 2024 stage of the program are presented in Table 2-85 and Table 2-86, respectively.
These reflect broader adoption rates of vehicle technologies already considered in the technology
basis for the MY 2021 standards. The standards for vehicles powered by CI engines also reflect
that in MY 2024, the separate engine standard would be more stringent, so the vehicle standard
keeps pace with the engine standard.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-85 EPA CO2 Standards for MY2024 Class 2b-8 Vocational Vehicles
EPA STANDARD FOR VEHICLE WITH CI ENGINE EFFECTIVE MY2024 (GRAM
CO2/TON-MILE)
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7
Heavy Heavy-Duty
Class 8
Urban
385
271
283
Multi-Purpose
344
246
242
Regional
296
221
194
EPA Standard for Vehicle with SI Engine Effective MY2024 (gram CC^/ton-mile)
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7 (and C8
Gasoline)

Urban
432
310

Multi-Purpose
385
279

Regional
324
251

Table 2-86 NHTSA Fuel Consumption Standards for MY2024 Class 2b-8 Vocational Vehicles
NHTSA STANDARD FOR VEHICLE WITH CI ENGINE EFFECTIVE MY 2024 (FUEL
CONSUMPTION GALLON PER 1,000 TON-MILE)
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7
Heavy Heavy-Duty
Class 8
Urban
37.8193
26.6208
27.7996
Multi-Purpose
33.7917
24.1650
23.7721
Regional
29.0766
21.7092
19.0570
NHTSA Standard for Vehicle with SI Engine
1,000 ton-mile)
iffective MY 2024 (Fuel Consumption gallon per
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7 (and C8
Gasoline)

Urban
48.6103
34.8824

Multi-Purpose
43.3217
31.3942

Regional
36.4577
28.2435

EPA's vocational vehicle CO2 standards andNHTSA's fuel consumption standards for
the full implementation year of MY 2027 are presented in Table 2-87 and Table 2-88,
respectively. These reflect even greater adoption rates of the same vehicle technologies
considered in the basis for the previous stages of the Phase 2 standards. The MY 2027 standards
for vocational vehicles powered by CI engines reflect additional engine technologies consistent
with those on which the separate MY 2027 CI engine standard is based.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-87 EPA CO2 Standards for MY2027 Class 2b-8 Vocational Vehicles
EPA STANDARD FOR VEHICLE WITH CI ENGINE EFFECTIVE MY2027 (GRAM
CO2/TON-MILE)
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7
Heavy Heavy-Duty
Class 8
Urban
367
258
269
Multi-Purpose
330
235
230
Regional
291
218
189
EPA Standard for Vehicle with SI Engine Effective MY2027 (gram CC^/ton-mile)
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7 (and C8
Gasoline)

Urban
413
297

Multi-Purpose
372
268

Regional
319
247

Table 2-88 NHTSA Fuel Consumption Standards for MY2027 Class 2b-8 Vocational Vehicles
NHTSA STANDARD FOR VEHICLE WITH CI ENGINE EFFECTIVE MY 2027 (FUEL
CONSUMPTION GALLON PER 1,000 TON-MILE)
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7
Heavy Heavy-Duty
Class 8
Urban
36.0511
25.3438
26.4244
Multi-Purpose
32.4165
23.0845
22.5933
Regional
28.5855
21.4145
18.5658
NHTSA Standard for Vehicle with SI Engine
1,000 ton-mile)
iffective MY 2027 (Fuel Consumption gallon per
Duty Cycle
Light Heavy-Duty
Class 2b-5
Medium Heavy-Duty
Class 6-7 (and C8
Gasoline)

Urban
46.4724
33.4196

Multi-Purpose
41.8589
30.1564

Regional
35.8951
27.7934

2.9.6.1 GEM-Based Custom Chassis Standards
Table 2-89 and Table 2-90 present EPA's CO2 standards and NHTSA's fuel consumption
standards, respectively, for custom vocational chassis. These standards may be selected by
custom chassis manufacturers, who retain the option of electing to certify to the primary
standards. (As already noted, these custom chassis vehicles will be required to use engines
meeting the Phase 2 engine standards, and thus, should generally incorporate the same engine
improvements as other vocational vehicles). The agencies have analyzed the technological
feasibility of achieving these optional fuel consumption and CO2 standards, based on projections
of actions manufacturers may take to reduce fuel consumption and emissions to achieve the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
standards, and believe that the standards are technologically feasible throughout the regulatory
useful life of the program.
These custom vehicle-level standards are predicated on a simpler set of vehicle
technologies than the primary Phase 2 standard for vocational vehicles. In developing these
optional standards, the agencies have evaluated the current levels of fuel consumption and
emissions, the kinds of technologies that could be utilized by manufacturers to reduce fuel
consumption and emissions, the associated lead time, the associated costs for the industry, fuel
savings for the owner/operator, and the magnitude of the CO2 reductions and fuel savings that
may be achieved. After examining the possibilities of vehicle improvements, the agencies are
basing the vehicle-level standards for coach buses, motor homes, school buses, transit buses, and
refuse trucks on the performance of workday idle reduction technologies, tire pressure systems,
simplified transmission improvements, and further tire rolling resistance improvements. The
agencies are basing the standards for concrete mixers and emergency vehicles on use of tires
with current average levels of rolling resistance. The EPA-only air conditioning standard is
based on leakage improvements. Of these technologies, we believe that improved tire rolling
resistance, neutral idle, and air conditioning leakage improvements are available today and may
be adopted as early as MY 2021. The vehicle technology that we believe will benefit from more
development time for engine and vehicle integration is stop-start idle reduction.
The MY 2024 standards reflect broader adoption rates of vehicle technologies already
considered in the technology basis for the MY 2021 standards. EPA's custom chassis CO2
standards and NHTSA's fuel consumption standards for the full implementation year of MY
2027 reflect even greater adoption rates of the same vehicle technologies considered as the basis
for the MY 2024 standards.
As with the other regulatory categories of heavy-duty vehicles, NHTSA and EPA are
adopting standards that apply to custom chassis vocational vehicles at the time of production,
and EPA is adopting standards for a specified period of time in use (e.g., throughout the
regulatory useful life of the vehicle).
The optional standards shown below were derived using baseline vehicle models with
many attributes similar to those developed for the primary program, as described above in
Chapter 2.9.2. For better transparency with respect to the incremental difference between the
MY 2021 and MY 2027 vehicle standards, we have modeled a certified MY 2027 engine for
both vehicle model years of optional custom chassis standards. Thus, chassis manufacturers who
do not make their own engines may compare the two model years of standards presented in
Table 2-89 and Table 2-90 and know that any differences are due solely to vehicle-level
technologies.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-89 EPA Emission Standards for Custom Chassis (gram CCh/ton-mile)

MY 2021
MY 2027
Coach Bus
210
205
Motor Home
228
226
School Bus
291
271
Transit
300
286
Refuse
313
298
Mixer
319
316
Emergency
324
319
Table 2-90 NHTSA Fuel Consumption Standards for Custom Chassis (gallon per 1,000 ton-mile)

MY 2021
MY 2027
Coach Bus
20.6287
20.1375
Motor Home
22.3969
22.2004
School Bus
28.5855
26.6208
Transit
29.4695
28.0943
Refuse
30.7466
29.2731
Mixer
31.3360
31.0413
Emergency
31.8271
31.3360
2.9.6.2 Summary of Vocational Vehicle Package Costs
The agencies have estimated the costs of the technologies that could be used to comply
with the final Phase 2 vocational vehicle standards. The estimated costs are shown in Table 2-91
for MY2021, in for MY2024, and for MY 2027. Fleet average costs are shown for light,
medium and heavy HD vocational vehicles in each duty-cycle-based subcategory - Urban,
Multi-Purpose, and Regional. As shown in Table 2-91, in MY 2021 these range from
approximately $900 for MHD and LHD Regional vehicles, up to $2,600 for HHD Regional
vehicles. Those two lower-cost packages reflect zero hybrids, and the higher-cost package
reflects significant adoption of automated transmissions. Many changes have been made to the
cost estimates since proposal. In the RIA Chapter 2.12.2, the agencies present vocational vehicle
technology package costs differentiated by MOVES vehicle type. These costs do not indicate the
per-vehicle cost that may be incurred for any individual technology. For more specific
information about the agencies' estimates of per-vehicle costs, please see the RIA Chapter 2.11.
The engine costs listed represent the cost of an average package of diesel engine technologies as
set out in RIA Chapter 2.7.7. Individual technology adoption rates for engine packages are
described in RIA Chapter 2.9.1.2.2. For gasoline vocational vehicles, the agencies are projecting
adoption of engine improvements that are reflected exclusively in the vehicle standard, see
Chapter 2.9.1.2.1 above) for an estimated $138 added to the average SI vocational vehicle
package cost beginning in MY 2021.
The details behind all these costs are presented in RIA Chapter 2.11, including the
markups and learning effects applied and how the costs shown here are weighted to generate an

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
overall cost for the vocational segment. These estimates have changed significantly from those
presented in the proposal, due to changes in projected technology adoption rates as well as
changes in direct costs that reflect comments received.
Table 2-91 Technology Package Incremental Costs for Vocational Vehicles for MY2021a b (2013$)

LIGHT HD
MEDIUM HD

HEAVY HD

Urban
Multi-
purpose
Regiona
1
Urban
Multi-
purpose
Regiona
1
Urban
Multi-
purpose
Regional
Engine0
$298
$298
$298
$275
$275
$275
$275
$275
$275
Tires
$0
$27
$27
$9
$9
$9
$13
$13
$13
Tire Pressure
Monitoring
$123
$154
$184
$123
$154
$184
$233
$292
$350
Transmission
$217
$217
$217
$217
$217
$217
$186
$413
$1,519
Axle related
$13
$13
$13
$13
$13
$13
$20
$26
$32
Weight
Reduction
$69
$69
$69
$69
$69
$69
$250
$250
$250
Idle reduction
$155
$155
$12
$160
$160
$12
$68
$68
$12
Hybridization
$178
$178
$0
$178
$178
$0
$178
$178
$0
Air
Conditioning11
$22
$22
$22
$22
$22
$22
$22
$22
$22
Other6
$30
$30
$30
$49
$49
$49
$89
$89
$89
Total
$1,106
$1,164
$873
$1,116
$1,146
$851
$1,334
$1,625
$2,562
Notes:
a Costs shown are for the 2021 model year and are incremental to the costs of a vehicle meeting the Phase 1 standards. These
costs include indirect costs via markups along with learning impacts. For a description of the markups and learning impacts
considered in this analysis and how it impacts technology costs for other years, refer to RIA Chapter 2.11.
b Note that values in this table include projected technology penetration rates. Therefore, the technology costs shown reflect the
average cost expected for each of the indicated vehicle subcategories.
c Engine costs shown are for a light HD, medium HD or heavy HD diesel engines. For gasoline-powered vocational vehicles we
are projecting $139 of additional engine-based costs beyond Phase 1.
d EPA's air conditioning standards are presented in Preamble Section V.C.
e Other incremental technology costs include electrified accessories and advanced shift strategy.
Table 2-92 presents estimated incremental costs for MY2024 for light, medium and
heavy HD vocational vehicles in each duty-cycle-based subcategory - Urban, Multi-Purpose,
and Regional. As shown, these range from approximately $1,300 for MHD and LHD Regional
vehicles, up to $4,000 for HHD Regional vehicles. The increased costs above the MY 2021
values reflect increased adoption rates of individual technologies, while the individual
technology costs are generally expected to remain the same or decrease, as explained in the RIA
Chapter 2.11. For example, Chapter 2.11.7 presents MY 2024 hybridization costs that range
from $6,046 to $15,872 per vehicle for vocational vehicles.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-92 Technology Package Incremental Costs for Vocational Vehicles for MY2024a b (2013$)

LIGHT HD
MEDIUM HD
HEAVY HD

Urban
Multi-
purpose
Regional
Urban
Multi-
purpose
Regional
Urban
Multi-
purpose
Regional
Engine0
$446
$446
$446
$413
$413
$413
$413
$413
$413
Tires
$0
$31
$33
$10
$10
$33
$13
$13
$53
Tire Pressure
Monitoring
$155
$183
$211
$155
$183
$211
$294
$347
$401
Transmission
$276
$276
$276
$276
$276
$276
$222
$1,032
$2,193
Axle related
$24
$24
$24
$24
$24
$24
$37
$54
$60
Weight
Reduction
$186
$186
$186
$186
$186
$186
$684
$684
$684
Idle reduction
$248
$248
$21
$256
$256
$21
$242
$242
$21
Hybridization
$550
$550
$0
$653
$653
$0
$844
$844
$0
Air
Conditioning11
$20
$20
$20
$20
$20
$20
$20
$20
$20
Othere
$54
$54
$54
$89
$89
$89
$162
$162
$162
Total
$1,959
$2,018
$1,272
$2,082
$2,110
$1,274
$2,932
$3,813
$4,009
Notes:
a Costs shown are for the 2024 model year and are incremental to the costs of a vehicle meeting the Phase 1 standards. These
costs include indirect costs via markups along with learning impacts. For a description of the markups and learning impacts
considered in this analysis and how it impacts technology costs for other years, refer to RIA Chapter 2.11.
b Note that values in this table include projected technology penetration rates. Therefore, the technology costs shown reflect the
average cost expected for each of the indicated vehicle subcategories.
c Engine costs shown are for a light HD, medium HD or heavy HD diesel engines. For gasoline-powered vocational vehicles we
are projecting $136 of additional engine-based costs beyond Phase 1.
d EPA's air conditioning standards are presented in Preamble Section V.C.
e Other incremental technology costs include electrified accessories and advanced shift strategy.
Table 2-93 presents estimated incremental costs for MY2027 for light, medium and
heavy HD vocational vehicles in each duty-cycle-based subcategory - Urban, Multi-Purpose,
and Regional. As shown, these range from approximately $1,500 for MHD and LHD Regional
vehicles, up to $5,700 for HHD Regional vehicles. Although the Multipurpose and Urban
subcategories are projected to adopt some high-cost technologies such as hybrids, the HHD
Regional package comes out more costly because it reflects 90 percent adoption of automated
transmissions. The engine costs shown represent the average costs associated with the MY 2027
vocational diesel engine standard described in Section II.D of the Preamble. For gasoline
vocational vehicles, the agencies are projecting adoption of engine technologies with an
estimated $125 added to the average SI vocational vehicle package cost in MY 2027. Further
details on how these SI vocational vehicle costs were estimated are provided above in Chapter
2.9.1.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-93 Technology Package Incremental Costs for Vocational Vehicles for MY2027a b (2013$)

LIGHT HD
MEDIUM HD
HEAVY HD

Urban
Multi-
purpose
Regional
Urban
Multi-
purpose
Regional
Urban
Multi-
purpose
Regional
Engine0
$481
$481
$481
$446
$446
$446
$446
$446
$446
Tires
$12
$24
$24
$6
$24
$24
$12
$36
$36
Tire Pressure
Monitoring
$187
$214
$240
$187
$214
$240
$355
$405
$456
Transmission
$271
$271
$293
$271
$271
$293
$220
$990
$3,269
Axle related
$35
$35
$35
$35
$35
$35
$52
$82
$87
Weight
Reduction
$294
$294
$294
$294
$294
$294
$1,102
$1,102
$1,102
Idle
reduction
$303
$303
$23
$314
$314
$23
$365
$365
$23
Hybridization
$857
$857
$0
$1,032
$1,032
$0
$1,353
$1,353
$0
Air
Conditioning
d
$20
$20
$20
$20
$20
$20
$20
$20
$20
Other6
$73
$73
$77
$122
$122
$127
$227
$227
$231
Total
$2,533
$2,571
$1,486
$2,727
$2,771
$1,500
$4,151
$5,025
$5,670
Notes:
a Costs shown are for the 2027 model year and are incremental to the costs of a vehicle meeting the Phase 1 standards. These
costs include indirect costs via markups along with learning impacts. For a description of the markups and learning impacts
considered in this analysis and how it impacts technology costs for other years, refer to RIA Chapter 2.11.
b Note that values in this table include projected technology penetration rates. Therefore, the technology costs shown reflect the
average cost expected for each of the indicated vehicle subcategories.
c Engine costs shown are for a light HD, medium HD or heavy HD diesel engines. For gasoline-powered vocational vehicles we
are projecting $125 of additional engine-based costs beyond Phase 1.
d EPA's air conditioning standards are presented in Preamble Section V.C.
e Other incremental technology costs include electrified accessories and advanced shift strategy.
2.10 Technology Application and Estimated Costs - Trailers
The agencies are adopting standards for trailers specifically designed to be pulled by
Class 7 and 8 tractors. These standards are expressed as CO2 and fuel consumption standards,
and would apply to each trailer with respect to the emissions and fuel consumption that would be
expected for a specific standard type of tractor pulling such a trailer. EPA and NHTSA believe it
is appropriate to establish standards for trailers separately from tractors because they are
separately manufactured by distinct companies which control every aspect of their design and
thus are the appropriate entity to certify compliance; the agencies are not aware of any
manufacturers that currently assemble both the finished tractor and the trailer. The legal basis
for setting separate standards for trailers is discussed in the Preamble in Section I.E. This section
of the RIA describes the analyses performed by the agencies as we developed the trailer
program.
2.10.1 Trailer Subcategories Evaluated
The agencies evaluated several trailer subcategories for these rules. Though many of the
same technologies are available for dry and refrigerated vans, the agencies evaluated these trailer

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
types separately. The transport refrigeration unit (TRU) commonly located at the front of
refrigerated trailers adds weight, has the potential to impact the aerodynamic characteristics of
the trailer, and may limit the types of aerodynamic devices that can be applied. Additionally,
"long box" vans in lengths 50 feet or longer and "short box" vans less than 50 feet in length were
evaluated separately due to differences in both weight and use patterns. We have chosen 53-foot
box vans to represent all long box vans in both compliance modeling and testing. Short box vans
are represented by solo 28-foot vans. The agencies did test other trailer lengths and the results
are presented in this chapter.
The agencies identified a list of work-performing devices that are sometimes added to
standard box vans, which may inhibit the use of some aerodynamic devices. Trailer
manufacturers may designate box vans that are restricted from using aerodynamic devices in one
location on the trailer as "partial-aero" box vans. We believe these trailers have the ability to
adopt single aerodynamic technologies, but do not expect them to be able to meet the same
stringencies as the "full-aero" box vans throughout the program.
Additionally, manufacturers may designate box vans that have work-performing devices
in two locations such that they inhibit the use of all practical aerodynamic devices as "non-aero"
box vans that would not be expected to adopt aerodynamic technologies at any point in the
program. These trailers have standards based on the use of tire technologies only. Similarly, we
recognize the potential for CO2- and fuel consumption reduction from three non-box trailers
(e.g., tankers, flatbeds, and container chassis). Standards for these non-box trailers are also based
on the use of tire technologies and do not reflect the use of aerodynamic technologies.
In summary, the agencies are adopting standards for ten trailer subcategories:
-	Long box (longer than 50 feet) dry vans
-	Long box (longer than 50 feet) refrigerated vans
-	Short box (50 feet and shorter) dry vans
-	Short box (50 feet and shorter) refrigerated vans
-	Partial-aero long box dry vans
-	Partial-aero long box refrigerated vans
-	Partial-aero short box dry vans
-	Partial-aero short box refrigerated vans
-	Non-aero box vans (all lengths of dry and refrigerated vans)
-	Non-box trailers (tanker, platform, container chassis only)
The analysis in the following sections describes our evaluation of the cost and
effectiveness of the technologies used in the design of the Phase 2 trailer program. We conclude
with a description of the development of our GEM-based equation that box van manufacturers
will use for compliance.
2.10.2 Defining the Trailer Technology Packages
The impact of a trailer on the overall fuel efficiency and CO2 emissions of a tractor-
trailer vehicle varies depending on three main characteristics of the trailer: aerodynamic drag,

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
tire rolling resistance, and weight. In this section, we outline the technologies that address these
characteristics and the ones the agencies evaluated for the standards.
2.10.2.1 Aerodynamic Drag Reduction
The rigid, rectangular shape of box vans creates significant aerodynamic drag and makes
them ideal candidates for aerodynamic technologies that can reduce drag and improve fuel
consumption and CO2 emissions. Current aerodynamic technologies for box vans have shown
significant drag reductions, as discussed below. These technologies are designed to create a
smooth transition of airflow from the tractor, around the trailer, and beyond the trailer. Box vans
provide opportunities to address drag at the front, rear, and underside of the trailer, and the
agencies considered several types of aerodynamic devices designed to address drag at all of these
points. Table 2-94 lists common aerodynamic technologies for use on box vans and a
description of their intended impact. Several versions of each of these technologies are
commercially available and have seen increased adoption over the past decade. Performance of
these devices varies based on their design, their location and orientation on the trailer, and the
vehicle speed.
Table 2-94 Common Bolt-on Aerodynamic Technologies for Box Trailers
LOCATION
ON TRAILER
EXAMPLE TECHNOLOGIES
INTENDED IMPACT ON AERODYNAMICS
Front
Front fairings and gap-reducing
fairings
Reduce cross-flow through gap and smoothly
transition airflow from tractor to the trailer
Rear
Rear fairings, boat tails and flow
diffusers
Reduce pressure drag induced by the trailer wake
Underside
Side fairings and skirts, and
underbody devices
Manage flow of air underneath the trailer to reduce
turbulence, eddies and wake
2.10.2.1.1 Comparison of Technology Performance: SmartWay-Verification
and GEM Results
SmartWay-verified technologies are evaluated on 53-foot dry vans. The verified
technologies are grouped into bins that represent one percent, four percent, or five percent fuel
savings relative to a typical long-haul tractor-trailer at 65-mph cruise conditions. Use of verified
aerodynamic devices totaling at least five percent fuel savings, along with verified tires, qualifies
a 53-foot dry van trailer for the "SmartWay Trailer" designation. In 2014, EPA expanded the
program to include refrigerated vans and provided a "SmartWay Elite" designation if fleets adopt
verified tires and aerodynamic equipment providing nine percent or greater fuel savings. To-
date, ten aerodynamic technology packages from six manufacturers have received the SmartWay
Elite designation. We may refer to SmartWay verification levels in this analysis, since the trailer
industry is most familiar with these values as a measure of trailer performance.
It is important to note that the cruise speed results presented in SmartWay do not
necessarily match the results of EPA's Greenhouse gas Emissions Model (GEM), which is the
tool the agencies will use for trailer standard development and compliance evaluation. Figure
2-56 shows a comparison of the CO2 reductions calculated for the three individual drive cycles
simulated in GEM: 65-mph cruise, 55-mph cruise, and a transient cycle. It also shows

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
reductions using a combination of the three GEM cycles with the cycle weightings the agencies
are assigning to represent long-haul and short-haul operation. The long-haul weighting is
calculated as 86 percent 65-mph cruise, 9 percent 55-mph cruise, and 5 percent transient. The
short-haul weighting is 64 percent 65-mph, 17 percent 55-mph, and 19 percent transient. These
percent values are based on the drive cycle weightings used in EPA's Phase 1 tractor program.178
This figure could be used to estimate the difference in performance that can be expected
when comparing a constant, 65-mph cruise test similar to SmartWay's performance tests (solid
black line) to the results from GEM (wide dashes) or to other driving conditions. These results
suggest that the SmartWay Elite target improvement of nine percent would be closer to eight
percent using GEM's long-haul simulation, while tractor-trailers that drive closer to 55-mph
would likely see improvements of 7 percent. It can also be seen that tractor-trailers driving
under highly transient conditions are likely to observe much smaller improvements. These
results are for illustrative purposes only and do not provide an exact correlation between test
results, results from GEM, and real-world results.
16%
14%
12%
10%
8%
o
3
TJ

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
performance compared to SmartWay Verified technologies (e.g., many skirts on 53-foot vans).
Additionally, short box vans (50 feet and shorter in length) are simulated with the GEM's short
haul drive cycle weightings, which results in performance that is up to two percent lower than
expected from constant 65-mph cruise speeds in the aerodynamic drag range considered in this
program.
Similar to the trend shown in Figure 2-56, even short box vans that operate in 100 percent
transient conditions experience a non-zero benefit from the use of aerodynamic devices. While
the benefit is low in these conditions, we expect a majority of short box vans, even those that
consider themselves exclusively "short-haul", will spend some time at highway speeds of 55-
mph or faster, at which time the trailer will achieve CO2 and fuel consumption reductions of at
least one percent.
Skirts+Gap
(28' dry van)
Skirts
(28' dry van}
Transient
Reduction in Aerodynmamic Drag Area, ACdA (m2)
Figure 2-57 GEM Drive Cycles' Impact on Aerodynamic Performance for a 28-Foot Box Dry Van with a
Tire Rolling Resistance Level of 5.0 kg/ton and No Weight Reduction
2.10.2.1.2 Aerodynamic Testing Results
EPA collected aerodynamic test data for many of the technologies mentioned previously
on several tractor-trailer configurations using the four test methods outlined in our test
procedures: coastdown, constant speed, wind tunnel, and CFD. The testing included multiple
tractor models, trailer models (including 53-foot, 48-foot, 33-foot, and 28-foot lengths), and
aerodynamic technologies. The results that follow are from coastdown, wind tunnel and CFD
testing. Detailed descriptions of test setup and generation of these results, including constant
speed, are provided in Chapter 3.2.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
In this rulemaking, the aerodynamic performance of a tractor-trailer vehicle is quantified
by the aerodynamic drag area, CdA (coefficient of drag multiplied by frontal area), which is a
function of both tractor and trailer aerodynamic characteristics. The following sections highlight
the impact of tractor and trailer characteristics, wind, test procedure, and trailer devices on
aerodynamic performance. These results were used to create the aerodynamic bins for trailer
manufacturers to use in compliance.
2.10.2.1.2.1 Evaluation Trailer Model Effects
The aerodynamic performance of basic trailer models does not vary significantly from
one manufacturer to the next. The wind tunnel results shown in Figure 2-58 indicate there is
very little difference in performance between trailer manufacturers for their basic trailer models.
The results shown are an average of six tractor models with each 53-foot trailer in the given
configuration. A maximum variation of 0.2 m2 is observed between trailer models with
combinations of skirts and a tail. The other configurations have variation less than 0.1 m2
These results suggest that the aerodynamic designs of current box vans do not drastically differ
by manufacturer, and the addition of bolt-on technologies is expected to result in similar
aerodynamic improvements from these base configurations.
¦ Trailer 1 ~ Trailer 2 ~ Trailer 3
6.5
io io io
lO lO lO
5" 5.5 H II	^ o
s:: II inn inn
No Control	Skirts	Skirts+Tail
Results measured at zero yaw
Figure 2-58 Variation in Performance of Trailer Devices due to Trailer Manufacturer; Average Absolute
CdA of Six Tractors Pulling each 53-foot Basic Dry Van Model
2.10.2.1.2.2 Evaluation Tractor Model Effects
Figure 2-59 shows that there is more variation in aerodynamic performance when
considering tractor models. All of the tractors shown in the figure are Class 8 high roof sleeper
cabs with similar aerodynamic features, but from four separate manufacturers. The absolute
CdA ranges from 0.2 m2 to 0.3 m2 depending on trailer configuration.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
i Tractor 4 ~ Tractor 8 ~ Tractor 9 ~ Tractor 11
CM
E
6.5
6.0
5.5
r-
to LO
« ui
CD
T3 5.0
O
4.5
4.0
O IT)
lO
CM
o iri
LO
I
No Control
Skirts	Skirts+Tail
Results measured at zero yaw
Figure 2-59 Variation in Aerodynamic Performance of Trailer Devices due to Tractor Manufacturer;
Average Absolute CdA of Three 53-Foot Dry Vans Pulled by each Tractor Model
By subtracting the absolute CdA value of the "Skirts" and "Skirts+Tail" configurations
from their corresponding "No Control" configuration, we obtain a change in CdA (i.e., "delta
CdA") that gives the relative impact of adding devices compared to a no control trailer.
Considering a delta CdA instead of absolute values reduces some of the impact of the tractor
characteristics and consequently reduces the variation by nearly half. Figure 2-60 shows that the
variation observed between tractor models is 0.15 m2 or less when using delta CdA. This
reduction in variation due to vehicle characteristics is one of the reasons the agencies are
choosing a delta CdA approach for the Phase 2 trailer program's aerodynamic testing. The
aerodynamic performance results in the rest of this section will be presented as delta CdA.
i Tractor 4 ~ Tractor 8 ~ Tractor 9 ~ Tractor 11

1.8

1.6

1.4
cn"
1.2
E
1.0
<
0.8
~o

O
0.6
ra
-i—*
0.4

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.10.2.1.2.3 Evaluation of Yaw Effects
As discussed in Chapter 2.8, the tractor program, which is using wind-averaged drag
results, specifies the coastdown test procedures as a reference test method and manufacturers
apply a correction factor to "alternative methods" (i.e., wind tunnel, CFD, or constant speed) in
order to maintain consistency between methods. The trailer program did not propose to require a
reference test, in order to reduce the test burden for manufacturers and allow them to choose an
appropriate test method for their needs and resources. The agencies also proposed standards that
were developed using zero yaw drag results. The agencies recognize that the benefits of
aerodynamic devices for trailers can be better seen when measured considering multiple yaw
angles, but we did not propose to accept wind-averaged drag results. The coastdown procedure
has near-zero wind restrictions and we were concerned that devices that show larger benefits at
greater yaw angles would not be captured in coastdown testing.
Commenters indicated that it was unlikely they would use coastdown testing for
compliance. Instead, they would rely on wind tunnel and CFD. Additional commenters
suggested that we consider wind-averaged results for the trailer program and, accordingly, we
evaluated the coastdown and wind tunnel results again, including new results from tests that
were completed following publication of the NPRM.
To evaluate the effect of wind, we compared the zero yaw and wind-averaged results
from EPA's wind tunnel tests. All wind-average results in this section are calculated from a
fourth-order polynomial fit to the measured yaw curve. As described in Chapter 3, the agencies
found that the average of the results from the equation at positive and negative 4.5 degrees yaw
angles was consistent with the wind-averaged results at 7 degrees and 65 miles per hour vehicle
speed (see Chapter 3.2 of this RIA, and 40 CFR 1037.810).
The results shown in Figure 2-61 compare the delta CdA at zero yaw with the wind-
averaged values for tests of six different tractors pulling three different models of 53-foot dry
vans. Figure 2-62 shows a similar comparison for two sets of tractor-trailers with solo 28-foot
dry vans. The wind-averaged analysis generally results in a narrower range of performance for a
given technology. The gap reducer technology shows minimal benefit under a zero yaw analysis
for the 53-foot vans, but a measurable benefit when yaw angles are considered. Tails also show
a noticeable improvement under yaw conditions. The short van results show larger increases in
delta CdA when wind-averaged results are considered.

-------
' E. O. 12866 Review — Revised - Do Not Cite, Quote, or Release During Review
1.80
1.60
Compare Technology Effectiveness: Zero Yaw & Wind-Averaged Delta CdA
From Wind Tunnel Testing of 53-Foot Dry Vans
Open = Zero Yaw
Solid = Wind-Averaged
1.40
£ 1.20
<
O 1.00
o>
Q 0.80
0.60
0.40
0.20
0.00
3
oo
e i
Ql
Gap
X
X
o
!!
+
o
DTail < Skirts SSkirts+Gap E:Skirts+Tail
Skirts+Tail+Gap
Figure 2-61 Comparison of Zero Yaw and Wind-Averaged Delta CdA for Wind Tunnel Tests of 53-Foot Dry
Vans; Results from Seven Class 8 Sleeper Cab Tractors and Three Dry Van Models
E
<
TJ
o
0)
Q
1.80
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00
Technology Effectiveness: Zero Yaw & Wind-Averaged Delta CdA
From Wind Tunnel Testing of 28-Foot Dry Vans
Open = Zero Yaw
Solid = Wind-Averaged
0
OO
CGap DTail
*
y*
63
X
X
Si
+
+
o Skirts Skirts+Gap xSkirts+Tail + Skirts+Tail+Gap
Figure 2-62 Comparison of Zero Yaw and Wind-Averaged Delta CdA for Wind Tunnel Tests of 28-Foot Dry
Vans; Results from Two Class 7 Day Cab Tractors and Two Dry Van Models

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
In light of trailer manufacturers' preference for wind tunnel and CFD, and the benefit
observed when testing at higher yaw angles, we are adopting standards based on wind-averaged
delta CdA values. The following section describes the variation seen in our testing of the three
test methods, including a comparison of the wind-averaged wind tunnel and CFD results to the
coastdown values at near-zero yaw angles.
2.10.2.1.2.4 Evaluation of Test Procedure Effects
As mentioned previously, EPA evaluated trailer aerodynamic performance using three
test procedures: coastdown, wind tunnel and CFD. EPA performed its wind tunnel testing at
ARC Indy using a l/8th-scale model of several tractor-trailers. We also obtained data from
National Research Council of Canada (NRC) from a 30 percent scale model in their 9-meter
wind tunnel.179 Figure 2-63 compares the coastdown and two wind tunnel facilities. The tractor
and trailer used in the coastdown and two wind tunnels are similar, but are not exact matches in
these tests and we cannot directly compare the numerical results. The coastdown tractor
corresponds to Tractor #3 in the coastdown results of Chapter 3.2 and the ARC wind tunnel
model corresponds to Tractor #5 in the ARC wind tunnel results.180 The NRC model is a generic
tractor developed by NRC. The comparison of trailers with skirts suggest that the coastdown and
wind tunnel methods produce similar results with these devices, and the effect of accounting for
higher yaw does not improve the performance with these devices. The limited yaw effect with
skirts was also observed in Figure 2. The yaw impact does appear to be larger when a tail is
included in the trailer configuration. The two wind tunnel results are within 0.2 m2, but the
coastdown result is much lower than both wind tunnel values.
¦ Coastdown nWind Tunnel, 1/8th Scale nWind Tunnel, 30% Scale
1.8
1.6
1.4
I1-2	n
< 1.0
Q
O 0.8
I
Q 0-4 I — H
I I
0.0	1
Skirt	Skirt + Tail
Figure 2-63 Comparison of Coastdown and Wind Tunnel Test Methods using Similar Tractor-Trailers with
a 53-Foot Dry Van
We also compared CFD results from two separate CFD packages. One is based on
Reynolds Averaged Navier-Stokes and the other is Lattice-Boltzmann-based. The two packages
were tested using the same tractor-trailer model, though there were some differences in grid
generation techniques, and open-road environments with a Reynolds number of 1. Ie6. Figure

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
2-64 compares coastdown and wind tunnel results to those predicted by the CFD models. The
coastdown tractor corresponds to Tractor #1 in the coastdown results of Chapter 3.2 and the wind
tunnel tractor corresponds to Tractor #11 in the ARC wind tunnel results.181 The results show
some difference between the CFD packages in the skirt configuration, but the differences remain
within 0.2 m2 between all methods shown. Similar to zero yaw results in Figure 2-61, the
coastdown results are much lower for the configuration with the tail, and we believe this is more
of a yaw effect than a variability between methods.
¦ Coastdown nWind Tunnel, 1/8th Scale nRANSCFD nL-BCFD
1.8
1.6
ST1-4	_
—12	I	II	
§1-0	¦
I- °'8 	 I
I	II
n fel
™	-L-
Skirt	Skirt + Tail
Figure 2-64 Comparison of Coastdown, Wind Tunnel and CFD Test Methods using Similar Tractor-Trailers
with a 53-Foot Dry Van
In general, Figure 2-63 and Figure 2-64 show that the test methods are reasonably close
for a given tractor-trailer configuration. We believe the lower values from the coastdown tests in
configurations that include tails are likely due to the relatively low yaw angles of that test
method, which was also seen in Figure 2-61 for 53-foot dry vans when comparing the zero yaw
and wind-averaged results of tail configurations.
Figure 2-65 displays all of the aerodynamic test results for used in our analysis of 53-foot
dry vans for the given configurations. Each data point is an individual test and the markers differ
based on test method. You can see that the three test methods (which include two wind tunnel
facilities and two CFD packages) produce similar results for most trailer configurations. With
the exception of the one coastdown data point for the tail configuration, even the coastdown
results at near-zero yaw are grouped relatively close to the results from the other test procedures.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Compare Technology Effectiveness: Wind-Averaged Delta CdA
From Several Test Methods of 53-Foot Dry Vans
ni/8th-Scale Wind Tunnel O30%-Scale Wind Tunnel ACoastdown RANS CFD xL-B CFD
1.80
1.60
1.40
£ 1.20
<
o 1.00
to
9
O 0.80
0.60
0.40
0.20
0.00
Figure 2-65 Technology Effectiveness for Several Devices on 53-Foot Dry Vans using Three Test Methods,
Including Two Wind Tunnel Facilities and Two CFD Packages
2.10.2.1.2.5 Evaluation of Aerodynamic Device Performance
Bolt-on aerodynamic technologies can be used individually or in combination. This
section summarizes our comparison of the performance of devices that were tested individually
and in combination with other devices. EPA evaluated several combinations in its aerodynamic
testing and those results are shown below.
Figure 2-66 shows the performance of three bolt-on devices when installed on three
different l/8th-scale trailer models in the wind tunnel. Each trailer is pulled by the same tractor
(i.e., Tractor #4 from the ARC wind tunnel data). These three devices are often used in
combination and it was of interest to investigate if the performance of these devices was additive
when combined, or if the devices work synergistically to achieve greater reductions in
combination.



















-
»QK








=
A

:
~







A




~
~ A





£
~







-
g	































B
A




Gap	Tail	Skirts Skirts+Gap Skirts+Tail Skirts*
Tail+Gap

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
i Trailer 1 ~ Trailer 2 ~ Trailer 3
1.8
1.6
IN 14
£1.2
< 1.0
o
ns
d>
Q
0.8
0.6
0.4
0.2
0.0
S ° H
S ° °
CO CO (N
q o o
o o o
I
Gap
Tail
Skirts
Figure 2-66 Wind Tunnel Performance of Individual Bolt-On Trailer Devices; Tractor #4 (ARC Wind
Tunnel Data) Pulling Each Trailer
In comparison to the values shown in Figure 2-66, Figure 2-67 shows that the devices are
more effective when combined, compared to the sum of their individual performances. For
example, the sum of the individual performances of the tail and skirts on Trailer #1 is 0.98 m2
and the sum of all three device performances is 1.02 m2 Yet, when tested in combination, they
achieve 1.13 m2 and 1.17 m2, respectively. Trailer #3 has similar levels of improvement for the
combined devices (about 13 percent compared to the sum of the individual performances).
However, the improvement from Trailer #2 is only about four percent. While these results
suggest there may be synergies between these particular device combinations, we would not be
able to predict a consistent improvement across all tractor and trailer models.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
i Trailer 1 ~ Trailer 2 ~ Trailer 3
1.8
1.6
— 1.4

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
different lengths, and concludes with a comparison of solo and tandem configurations. The
agencies are including these results to give a general idea of the relative performance that could
be expected when trailers of different lengths or configurations are used. Manufacturers will
continue to use data from solo 28-foot and 53-foot trailer testing for compliance.
Figure 2-68 compares the performance of four dry van lengths. The day cab (DC) tractor
is the same for the 28-foot, 48-foot, and 53-foot trailers shown. The 33-foot van was modeled
with a MY 2014 sleeper cab in a separate test set. We are including the 33-foot results in the
plot for qualitative assessment. You can see that the individual devices do not show a consistent
trend in performance based on trailer length, but there is a noticeable trend of increased
performance with increased length for combinations of devices.
1.80
1.60
1.40
_ 1.20
£_
§ 1.00
o
« 0.80
0.60
0.40
0.20
0.00
Technology Effectiveness: Wind-Averaged Delta CdA
From Wind Tunnel Testing Performed on Several Van Lengths




























a




CL


0
+




ro
O









£


+
r
CO



CT3
a


+






+


r




a


Skirts
Y—
+

j*:
CO
s
e
JC
CO















Q.
O

TO
O r=-|

a.

Q-
(O





O

O

28-ft 2012 DC
33-ft 2014 SC
48-ft 2012 DC
53-ft 2012 DC
Figure 2-68 Comparison of Aerodynamic Performance of Devices on Several Dry Van Lengths; 2012 DC is a
6x4 Day Cab Tractor, and 2014 SC is a 6x4 Sleeper Cab
It should be noted that the 53-foot van is the only "long box van" in this set of trailers.
The 28-foot, 33-foot, and 48-foot trailers are considered "short box vans" in this trailer program
and are represented by a 28-foot trailer for compliance. These results suggest that the shorter
surrogate test trailer will underestimate performance for the longer trailers in its regulatory
subcategory, providing a conservative measure of potential benefits when the longer trailers are
in use.
EPA also tested the 28-foot and 33-foot van in a tandem configuration. Each van pair
was tested with skirts on the first trailer only, skirts on the second trailer only, skirts on both
trailers, skirts and a gap reducer on both trailers, and skirts and a gap reducers on both trailers

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review **
with a tail on the second trailer. As shown in Figure 2-69, the skirts perform similarly for a
given length van when they are on an individual van only, but provide almost twice the benefit
when installed on both vans. The addition of the tail further improves the performance of the
pair of trailers.
Technology Effectiveness: Wind-Averaged Delta CdA
From Wind Tunnel Testing Performed on Tandem Dry Vans

1.80

1.60

1.40
(X
1.20 --
E


-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review *
these results suggest that there is an added performance benefit if customers were to purchase
and deploy a tail on vans that may be used in tandem.
Technology Effectiveness: Wind-Averaged Delta CdA
From Wind Tunnel Testing Performed on Solo & Tandem Dry Vans
1.80
28-ft 6x4 DC
28-ft 6x4 DC, Tandem
33-ft SC
33-ft SC, Tandem
Figure 2-70 Comparison of Aerodynamic Device Performance on Solo and Tandem Dry Vans; the 28-foot
Van is Pulled by a 6x4 Day Cab Tractor, and the 33-foot Van is Pulled by as a Sleeper Cab Tractor
2.10.2.1.3 Performance Bins for Aerodynamic Technologies
The agencies developed aerodynamic bins based on delta CdA to encompass technologies
that are expected to provide similar improvements in drag, and which are intended to account for
variability due to tractor model, test method, device manufacturer, and trailer manufacturer. The
proposed bins were based on zero yaw test results. For the final rulemaking, we are adopting
standards based on wind-averaged aerodynamic test data, for reasons explained immediately
below. In addition, we completed several test programs after the NPRM. The bins described
here reflect the new test results, and our use of wind-averaged values.
Figure 2-71 overlays the aerodynamic bins that we proposed in the NPRM on our recent
wind-averaged test results. While some of the technologies fit into those bins, many of the same
technologies overlap two or more bins. In addition, when the results are wind-averaged, tails and
skirts have similar performance, suggesting that they should be in the same bin.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***

1.80

1.60

1.40
CN
E
1.20
<

u
1.00
(U
¦*->

O

a
0.80

0.60

0.40

0.20

0.00
Compare Technology Effectiveness: Wind-Averaged Delta CdA
From Several Test Methods of 53-Foot Dry Vans
ni/8th-Scale Wind Tunnel «>30%-Scale Wind Tunnel ACoastdown RANS CFD XL-B CFD
NPRM Bin VIII = 18












I—, NPR
W Bin VII = 1.4





B











A

u
~
NPF
M Bin VI = 1.0





J
i





~
u




tT)
~

NP
iM Bin V= 0.7


n









J




NPF
M Bin IV = 0.5


_
3?













NPF
!M Bin III =0.3







	




NP
?M Bin II =0.1
	






Gap	Tail	Skirts Skirts+Gap Skirts+Tail Skirts-*-
Tail+Gap
Figure 2-71 Wind-Averaged Trailer Aerodynamic Test Results Relative to the NPRM Bins
We adjusted the aerodynamic bins to reflect the additional data and the use of wind-
averaged results, as seen in Figure 2-72. The most notable difference is that we expanded the
lower bins. The Bin II threshold delta CdA remains 0.1 m2 Anything below that threshold is
assigned a value of zero. The NPRM Bins III, IV and V were reduced to two bins, such that
Bins II, III and IV are each a width of 0.3 m2 Technologies that achieve a threshold value of 0.4
m2 or greater, such as most of the skirts and tails tested, are assigned to Bin III. Bin IV, which
has a threshold of 0.7 m2, includes the configurations tested with skirts and gap reducers, and
some of the lower performing skirt and tail combinations. A majority of the skirts and tail
combinations and skirts, tails and gap reducer combinations are in Bin V, which is assigned a
value of 1.0 m2 These combinations represent the highest performing devices that we tested.
Bins V, VI, and VII are identical to the highest bins from the NPRM. The agencies observed one
device combination that presently meets Bin VI, suggesting that this bin can be met with
combinations of existing aerodynamic technologies. The agencies believe that there is ample
lead time to optimize additional existing Bin V combinations such that they can also meet Bin VI
by MY 2027.

-------
1.80
1.60
1.40
£ 1.20
<
o 1.00
(U
9
Q 0.80
0.60
0.40
0.20
r E. O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review
Compare Technology Effectiveness: Wind-Averaged Delta CdA
From Several Test Methods of 53-Foot Dry Vans
ni/8th-Scale Wind Tunnel <0>3O%-Scale Wind Tunnel ACoastdown RANS CFD XL-B CFD
Bin VII = 1.8













BinVI = 1.4





B











A

u
~

Bin V = 1.0





J






~
u
A




A	
£T)
~


Bin IV = 0.7




—
A "*¦






p

-






$

-I




Bin III = 0 4



Zi




	





Bin II =0.1
—






Gap	Tail	Skirts Skirts+Gap Skirts+Tail Skirts*
Tail+Gap
Figure 2-72 Wind-Averaged 53-Foot Dry Van Aerodynamic Test Results Relative to the Aerodynamic Bins
that will be Used for Compliance
Much of our testing focused on 53-foot trailers, but we did test several combinations of
solo 28-foot trailers that will be used to represent all short box vans in compliance testing.
Figure 2-73 shows the wind-averaged results for two 28-foot dry vans in several configurations
from wind tunnel and coastdown testing. Similar to the 53-foot dry van results, the performance
of tails and skirts fit into the same bin. It is interesting to note that these results suggest a 28-foot
dry van with skirts and a gap reducer have similar performance as a skirts and tail combination.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
1.80
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00
Compare Technology Effectiveness: Wind-Averaged Delta CdA
From Wind Tunnel and Coastdown on 28-Foot Dry Vans
~ Wind Tunnel oCoastdown
Bin VII = 1.8




















Bin VI = 1.4



























Bin V = 1.0




















Bin IV = 0.7





~




~
~
~




¦
n

Bin III =0.4


u
¥










o
n




Bin II = 0 1
u






Gap	Tail	Skirts Skirts+Gap Skirts+Tail Skirts*
Tail+Gap
Figure 2-73 Wind-Averaged 28-Foot Dry Van Aerodynamic Test Results Relative to the Aerodynamic Bins
that will be Used for Compliance
While the agencies have chosen to test and regulate 28-foot box vans individually, they
are often pulled in a tandem configuration, which restricts the types of aerodynamic devices that
can be applied on the rear of the trailers. We expect rear devices such as boat tails would not be
practical for 28-foot box vans, since those devices are only deployable when the trailer is in the
rear position. We did not base our standards on the use of rear devices. However, the short box
van subcategories include other trailer lengths (e.g., 40-foot and 48-foot) that would be able to
use rear aerodynamic devices and we do not restrict the use of those devices as a means of
achieving compliance. We presented results from 28-foot configurations that included tails to
demonstrate the level of performance that can be expected when operating with those devices.
Table 2-95 below summarizes the bin structure that the agencies will use as the basis for
compliance. Also included in the table are example aerodynamic packages that the agencies
used for our cost analysis summarized below in Chapter 2.10.4.3 and fully described in in
Chapters 2.11 and 2.12. Note that the same technologies are assumed to work for dry and
refrigerated vans in each length category. We assume manufacturers that wish to achieve bins
where our example packages include gap reducers can have a different, similarly effective
technology installed in a separate location on refrigerated vans without additional cost. In each
set of example technologies, we present packages for bin performance that were not observed in
our testing. We considered these packages to be "Optimized Combinations" and assume their

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
cost to be that of an appropriately sized skirt, tail and gap reducer. The highest bins in each
category is assumed to require changes to the design of the trailer, and we did not estimate a cost
for those bins.
Table 2-95 Aerodynamic Technology Bins used to Evaluate Trailer Benefits and Costs
BIN
DELTA CDA
EXAMPLE TECHNOLOGY PACKAGES
Measured
GEM Input
Value
Long Vans
Short Vans
Bin I
<0.10
0.0
No Aero Devices
No Aero Devices
Bin II
0.10-0.39
0.1
High Performing Gap Reducer
Skirts or Tail
Bin III
0.40-0.69
0.4
Skirts or Tail
Skirts + Gap Reducer
Bin IV
0.70-0.99
0.7
Skirts + Gap Reducer
Optimized Combinations
Bin V
1.00 - 1.39
1.0
Skirts + Tail
Changes to Trailer Design
Bin VI
1.39 - 1.79
1.4
Optimized Combinations

Bin VII
> 1.80
1.8
Changes to Trailer Design

The agencies used EPA's Greenhouse gas Emissions Model (GEM) vehicle simulation
tool to conduct this analysis. Within GEM, the aerodynamic performance of each trailer
subcategory is evaluated by subtracting the delta CdA shown in Table 2-95 from the CdA value
representing a specific standard tractor pulling a trailer with no CO2- or fuel consumption-
reducing technologies (i.e., a "no-control" trailer). EPA's aerodynamic testing of Class 8 high
roof sleeper cab tractors pulling standard 53-foot dry vans in its no-control baseline
configuration (zero aerodynamic trailer technologies) produced an average CdA value of 5.9 m2
in coastdown testing and an average wind-averaged CdA from wind tunnel tests was 6.0 m2 The
average CdA value for the solo 28-foot dry van in its no-control configuration was 5.3 m2 for
coastdown and the average CdA from wind tunnel results were 5.6 m2 when wind-averaged.
The agencies chose to model the no-control long dry van subcategory using a default
CdA value of 6.0 m2 (the mean wind-averaged CdA from EPA's wind tunnel testing) in GEM.
We also chose the wind tunnel result of 5.6 m2 to represent the short dry van subcategory. The
agencies did not test any refrigerated vans, but we assumed a refrigerated van's TRU would
behave similar to a gap reducer. Our test results did not show gap reducer technologies to have a
significant effect on CdA and the agencies assigned the same default CdA to refrigerated and dry
box vans in GEM. The trailer subcategories that have design standards (i.e., non-box and non-
aero box trailers) do not have numerical standards to meet, and thus do not have defaults in
GEM. Table 2-96 illustrates the no-control drag areas (CdA) associated with each trailer
subcategory.
Table 2-96 Default CdA Values Associated with the No Control Trailer Configuration within GEM
TRAILER
DRY VAN
SUBCATEGORY

Long Dry Van
6.0
Short Dry Van
5.6
Long Ref. Van
6.0
Short Ref. Van
5.6

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.10.2.2 Tire Rolling Resistance
2.10.2.2.1	Lower Rolling Resistance Tires
On a typical Class 8 long-haul tractor-trailer, over 40 percent of the total energy loss from
tires is attributed to rolling resistance from the trailer tires.182 Trailer tire rolling resistance
values were collected by the agencies to use in the GEM-simulated tractor-trailer vehicle for
Phase 1. The agencies found that the average coefficient of rolling resistance (CRR) for new
trailer tires at that time was 6.0 kg/ton. This value was applied in GEM for the standard trailer
used for tractor compliance in the Phase 1 tractor program. For Phase 2, the agencies are
adopting the same baseline CRR for trailer tires and consider all box van tires with CRR values
below 6.0 kg/ton to be "lower rolling resistance" (LRR) tires. For reference, a trailer tire that
qualifies as a SmartWay-verified tire must meet a CRR value of 5.1 kg/ton, a 15 percent CRR
reduction from the trailer tire identified in Phase 1. Our research of rolling resistance indicates
an additional CRR reduction of 15 percent or more from the SmartWay verification threshold is
possible with tires that are available in the commercial market today.
Similar to the case of tractor tires, LRR tires are available as either dual or as single wide-
based tires for trailers. Single wide-based tires achieve CRR values that are similar to their dual
counterparts, but have an added benefit of weight reduction, which can be an attractive option for
trailers that frequently maximize cargo weight.
2.10.2.2.2	Performance Levels for LRR Tires
Similar to the Phase 2 tractor and vocational vehicle programs, the trailer program is
based on performance reflecting adoption of lower rolling resistance tires (or, for the non-aero
subcategories, actually adopting such tires). Feedback from several box trailer manufacturers
indicates that the standard tires offered on their new trailers are SmartWay-verified tires (i.e.,
CRR of 5.1 kg/ton or better). An informal survey of members from the Truck Trailer
Manufacturers Association (TTMA) indicates about 85 percent of box vans sold today have
SmartWay tires.183 While some trailers continue to be sold with tires of higher rolling
resistances, the agencies believe most box trailer tires currently achieve the baseline trailer tire
CRR of 6.0 kg/ton or better.
The agencies evaluated two levels of box van tire performance for these rules beyond the
baseline trailer tire with a CRR of 6.0 kg/ton. The first performance level was set at the criteria
for SmartWay-verification for trailer tires, 5.1 kg/ton, which is a 15 percent reduction in CRR
from the baseline. As mentioned previously, several tire models available today achieve rolling
resistance values well below the present SmartWay threshold. The agencies expect that tire
manufacturers will continue to respond to demand for more efficient tires and will offer
increasing numbers of tire models with rolling resistance values significantly better than today's
typical LRR tires. We believe it is reasonable to expect the trailer industry could adopt tires with
rolling resistances at a second performance level early in the program. The agencies prosed
standards based on meeting an additional eight percent reduction in rolling resistance by MY
2024, but, given that such a high fraction of new box vans are already adopting LRR tires, we are
adopting standards based on a CRR performance of 4.7 kg/ton by MY 2021. The agencies

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
evaluated these three tire rolling resistance levels, summarized in Table 2-97, in the feasibility
analysis of the following sections.
We received comment from Michelin supporting the use of 6.0 kg/ton as the box van tire
rolling resistance baseline, but they expressed concern that the SmartWay threshold of 5.1 kg/ton
does not apply for non-box trailers, and could compromise their operation. In addition, the
Rubber Manufacturers Association indicated that a baseline of 6.0 kg/ton does not apply to non-
box trailers. The agencies agree that the baseline tires for non-box trailers should have a higher
roller resistance, but we did not receive any comments that included Crr data. For the analysis
for the final rules, the agencies used 2014 tire rolling resistance information submitted by tractor
and vocational manufacturers for Phase 1 compliance to establish a revised baseline Crr value of
6.5 kg/ton for non-box trailer manufacturer. Table 2-97 summarizes the rolling resistance levels
we evaluated in the Phase 2 trailer program.
Table 2-97 Summary of Trailer Tire Rolling Resistance Levels Evaluated
ROLLING
RESISTANCE LEVEL
CRR (KG/TON)
Level 1 (Non-Box Baseline)
6.5
Level 2 (Box Van Baseline)
6.0
Level 3
5.1
Level 4
4.7
2.10.2.3 Tire Pressure Systems
The inflation pressure of tires also impacts the rolling resistance. Tractor-trailers
operating with all tires under-inflated by 10 psi have been shown to increase fuel consumed by
up to one percent.184 Tires can gradually lose pressure from small punctures, leaky valves or
simply diffusion through the tire casing. Changes in ambient temperature can also affect tire
pressure. Trailers that remain unused for long periods of time between hauls may experience any
of these conditions. A 2003 FMCSA report found that nearly one in five trailers had at least one
tire under-inflated by 20 psi or more. If drivers or fleets are not diligent about checking and
attending to under-inflated tires, the trailer may have much higher rolling resistance and much
higher CO2 emissions and fuel consumption.
2.10.2.3.1 Types of Tire Pressure Systems
Tire pressure monitoring systems (TPMS) and automatic tire inflation systems (ATIS)
are designed to address under-inflated tires. Both systems alert drivers if a tire's pressure drops
below its set point. TPMS simply monitors the tires and require user-interaction to reinflate to
the appropriate pressure. Today's ATIS take advantage of trailers' air brake systems to supply
air back into the tires (continuously or on demand) until a selected pressure is achieved. In the
event of a slow leak, ATIS have the added benefit of maintaining enough pressure to allow the
driver to get to a safe stopping area.185 As described in Chapter 2.4.3.3, the agencies will
recognize both systems in the Phase 2 trailer program.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.10.2.3.2 Performance of Tire Pressure Systems
Estimates of the benefits of tire pressure systems vary depending on the base level of
maintenance already performed by the driver or fleet, as well as the number of miles the trailer
travels. Trailers that are well maintained or that travel fewer miles would experience less
benefits from these systems compared to trailers that often drive with poorly inflated tires or log
many miles. The agencies believe these systems can provide a CO2 and fuel consumption
benefit to most trailers. With ATIS use, trailers that have lower annual vehicle miles traveled
(VMT) due to long periods between uses would be less susceptible to low tire pressures when
they resume activity. Trailers with high annual VMT or frequent changes in ambient conditions
would experience the fuel savings associated with consistent tire pressures. TPM systems would
provide a warning of inappropriate tire pressure and the agencies believe the operators have
sufficient incentive to correct the pressure as soon as possible. Tire inflation systems could
provide a CO2 and fuel consumption savings of 0.5-2.0 percent, depending on the degree of
under-inflation in the trailer system.
Maintaining tire pressure is important to fuel consumption. Tire manufacturers estimate
a tire pressure 10 psi below target results in a 0.9 percent increase in fuel consumption. Two
studies have evaluated truck and trailer tire inflation including FMCSA (2003) and TMC
(2002). 186>187 In the 2003 FMCSA study, tire inflation (psi) was measured in 3,200 tractors and
1,300 trailers. The TMC study measured tire inflation rates in two fleets and found that only 38
percent of sampled trailer tires were within +/- 5 psi of target pressure as prescribed by tire
manufacturers. The study also found that more than 20 percent of tires were 20 psi or more
underinflated and four percent of tires were 50 psi or more underinflated compared to the
target. The FMCSA study found similar results. These figures suggest under inflation of tractor
and trailer tires in the U.S. fleet could result in an increase in fuel consumption of approximately
one to two percent. Most recently, FMCSA (2014) evaluated trailer ATIS on trailers in two test
fleets.188 The study found ATIS on trailers, in conjunction with TPMS use on tractors, improved
fuel consumption 1.4 percent in test trucks as compared to control trucks in those fleets.
NHTSA and EPA recognize the role of proper tire inflation in maintaining optimum tire
rolling resistance during normal trailer operation. For these rules, rather than require
performance testing of tire pressure systems, the agencies will recognize the with a single default
reduction for manufacturers that incorporate ATIS or TPMS into their trailer designs. Based on
information available today, we believe that there is a narrow range of performance among
technologies available and among systems in typical use. We proposed to assign a 1.5 percent
reduction in CO2 and fuel consumption for all trailers that implement ATIS, and no credit for
TPMS due to their inherent dependence on operator interaction.189 Based on comments, we are
assigning a 1.2 percent reduction for ATIS and a 1.0 for TPMS. The discounted TPMS value is
meant to reflect our acceptance that a notification will incentivize an operator to address the
problem, but we cannot ensure that it will be done. We believe the use of these systems can
improve tire pressure maintenance and reduce tire rolling resistance.
2.10.2.4 Weight Reduction
Reduction in trailer tare (or empty) weight can lead to fuel consumption reductions in two
ways. For applications where payload is not limited by weight restrictions, the overall weight of

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
the tractor and trailer would be reduced and would lead to improved fuel efficiency. For
applications where payload is limited by weight restrictions, the lower trailer weight would allow
additional payload to be transported during the truck's trip, so g/ton-mile emissions would
decrease. Weight reduction opportunities in trailers exist in both the structural components and
in the wheels and tires. Manufacturers commonly replace components such as roof posts, bows,
side posts, cross members, floor joists, and floor sections with lighter weight options.
Major lower-weight options are not offered consistently by all trailer manufacturers
across the industry. For example, some manufacturers have already marketed lower-weight
major components for many years, while others to date have not done so. There is no clear
"baseline" for current trailer weight against which lower-weight designs could be compared for
regulatory purposes. Trailer manufacturers do not generally sell a single model. Instead, each
sale is likely to include customer-specified configurations with application-specific components.
For this reason, the agencies do not believe it would be appropriate or fair across the industry to
identify a single trailer as a standard baseline from which to apply overall weight reductions
toward compliance. However, the agencies do believe it would be appropriate to allow a
manufacturer to account for weight reductions that involve substituting very specific,
traditionally heavier components with lower-weight options that are not currently widely adopted
in the industry. This method allows manufacturers to easily identify and install components that
will improve benefit them in compliance.
The agencies recognize that when weight reduction is applied to a trailer, some operators
will replace that saved weight with additional payload. To account for this in the average trailer
represented in the GEM vehicle simulation tool, it is assumed that one-third of the weight
reduction is applied to the payload. For tractor-trailers simulated in GEM, it takes a weight
reduction of nearly 1,000 pounds before a one percent fuel savings is achieved and about a 2,500
pound reduction to reach three percent savings. The component substitutions identified by the
agencies result in weight reductions of less than 500 pounds, yet can cost over $1,000. The
agencies believe that few trailer manufacturers would apply weight reduction solely as a means
of achieving reduced fuel consumption and CO2 emissions, and the standards that can be met
without reducing weight. However, we will offer weight reduction as an option for box trailer
manufacturers who wish to apply it to some of their trailers as part of their compliance strategy.
2.10.2.4.1 Weight Reduction Options Recognized in these Rules
For these rules, the agencies have identified several conventional components with
available lighter-weight substitutes (e.g., substituting conventional dual tires with steel wheels
with single wide-based tires and aluminum wheels). We are adopting values for the associated
weight-related savings that would be applied with these substitutions for compliance. We
believe that the initial cost of these component substitutions is currently substantial enough that
only a relatively small segment of the industry has adopted these technologies today.
In addition to weight reduction associated with replacing standard steel wheels with
aluminum versions, and adopting single wide-based tires in place of dual tires, the agencies have
identified 11 common trailer components that have lighter weight options available.I90-191-192-193
Some of the references include confidential data that outlined weight savings and costs
associated with these material substitutions. Table 2-98 lists the components, and estimates of

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
weight savings and costs obtained by the agencies. The table includes one update to the weight
reduction value assigned to floor cross-members. The Aluminum Association indicated that this
value should be 250 pounds and we adjusted the table accordingly.
Manufacturers that adopt these technologies would sum the associated weight reductions
and apply those values in GEM. Steel wheels can be replaced with aluminum wheels and two
dual tires can be replaced with single wide-based tires on aluminum wheels. Relatively large
weight savings are possible by replacing steel upper coupler assemblies or suspension sub-
frames with aluminum versions, but these substitutions are more expensive and more labor-
intensive to install.
Table 2-98 Weight Reduction Options for Trailers
COMPONENT
MATERIAL
WEIGHT

SUBSTITUTION
REDUCTION (LB)
Hub and Drum (per axle)
Cast Iron to Aluminum
80
Floor
Hardwood to Aluminum
375
Floor
Hardwood to Composite
245
Floor Crossmembers
Steel to Aluminum
250
Landing Gear
Steel to Aluminum
50
Rear Door
Steel to Aluminum
187
Rear Door Surround
Steel to Aluminum
150
Roof Bows
Steel to Aluminum
100
Side Posts
Steel to Aluminum
300
Slider Box
Steel to Aluminum
150
Structure for Suspension Assembly
Steel to Aluminum
280
Upper Coupler Assembly
Steel to Aluminum
430
In addition to these conventional components, manufacturers have the option to evaluate
their own trailer weight reduction through the off-cycle testing provisions outlined in the
regulations. Manufacturers can seek approval of a baseline trailer from their own recent
production, and compare its weight to a new, lighter-weight model through an "A to B" weight
measurement. The difference between these two trailers can be applied in GEM for a weight
reduction value.
2.10.2.5 Effectiveness of Technologies
The final standards for trailers are predicated on four performance parameters:
aerodynamic drag reduction, tire rolling resistance reduction, and the adoption of tire pressure
systems and weight reduction. Table 2-99 summarizes the performance levels for each of these
parameters based on the technology characteristics outlined in Chapter 2.10.2.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-99 Performance Parameters for the Trailer Program
AERODYNAMICS (DELTA CDA, M2)
Bin I
0.0
Bin II
0.1
Bin III
0.4
Bin IV
0.7
Bin V
1.0
Bin VI
1.4
Bin VII
1.8
Tire Rolling Resistance (CRR, kg/ton)
Level 1 (Non-Box Baseline)
6.5
Level 2 (Box Van Baseline)
6.0
Level 3
5.1
Level 4
4.7
Tire Inflation System (% reduction)
ATIS
1.2
TPMS
1.0
Weight Reduction (pounds)
Weight
1/3 added to payload,

remaining reduces overall

vehicle weight
As part of the process of demonstrating compliance, trailer manufacturers will perform an
aerodynamic test and measure a delta CdA. The delta CdA value will determine which Bin value
the manufacturer will supply to GEM (i.e. the GEM equation) for compliance. While
manufacturers are required to use the exact value assigned to the aerodynamic bins, they are free
to use any tire rolling resistance value obtained from tire testing.
These performance parameters have different effects on each trailer subcategory due to
differences in the simulated trailer characteristics. Table 2-100 shows the agencies' estimates of
the effectiveness of each parameter for four box trailer types. Each technology was evaluated in
GEM using the baseline parameter values for the other technology categories. For example, each
aerodynamic bin was evaluated using the Tire Level 1 (6.0 kg/ton) and the Base weight reduction
option (zero pounds). The table shows that aerodynamic improvements offer the largest
potential for CO2 emissions and fuel consumption reductions, making them relatively effective
technologies.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-100 Effectiveness (Percent Reduction in CO2 Emissions and Fuel Consumption) of Technologies for
the Box Van Subcategories
AERODYNAMICS
DELTA CDA (M2)
DRY VAN
REFRIGERATED VAN
Long
Short
Long
Short
Bin I
0.0
0%
0%
0%
0%
Bin II
0.1
1%
1%
1%
1%
Bin III
0.4
3%
3%
3%
3%
Bin IV
0.7
5%
5%
5%
5%
Bin V
1.0
7%
7%
7%
7%
Bin VI
1.4
9%
10%
9%
10%
Bin VII
1.8
12%
13%
12%
13%
Tire Rolling Resistance
CRR (kg/ton)
Dry Van
Refrigerated Van
Long
Short
Long
Short
Level 2 (Baseline)
6.0
0%
0%
0%
0%
Level 3
5.1
2%
1%
2%
1%
Level 4
4.7
3%
2%
3%
2%
Weight Reduction
Weight (lb)
Dry Van
Refrigerated Van
Long
Short
Long
Short
Baseline
0
0%
0%
0%
0%
Option 1
100
0%
0%
0%
0%
Option 2
500
1%
1%
1%
1%
Option 3
1000
1%
2%
1%
2%
Option 4
2000
2%
4%
2%
4%
2.10.3 Defining the Baseline Trailers
2.10.3.1 No-Control Default Tractor-Trailer Vehicles within GEM
The regulatory purpose of EPA's heavy-duty vehicle compliance tool, GEM, is to
combine the effects of trailer technologies through simulation so that they can be expressed as
kg/ton-mile and gal/100 ton-mile and thus avoid the need for direct testing of each trailer model
being certified. The trailer program has separate standards for each trailer subcategory, and a
unique tractor-trailer vehicle was chosen to represent each subcategory for compliance. In the
Phase 2 update to GEM, each trailer subcategory is modeled as a particular trailer being pulled
by a standard tractor depending on the physical characteristics and use pattern of the trailer.
Table 2-101 highlights the relevant vehicle characteristics for the no-control default tractor-
trailer of each subcategory. Level 1 trailer tires are used, and the drag area, which is a function
of the aerodynamic characteristics of both the tractor and trailer, is set to the Bin I values shown
previously in Table 2-96. Weight reduction and tire pressure systems are not applied in these
baselines. In general, long box vans are pulled by sleeper cab tractors, and short box vans are
pulled by 4x2 day cabs.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-101 Characteristics of the No-Control Default Tractor-Trailer Vehicles in GEM

DRY VAN
REFRIGERATED VAN
Trailer Length
Long
Short
Long
Short
Standard Tractor




Class
Class 8
Class 7
Class 8
Class 7
Cab Type
Sleeper
Day
Sleeper
Day
Roof Height
High
High
High
High
Axle Configuration
6x4
4x2
6x4
4x2
Engine
2018 MY
2018 MY
2018 MY
2018 MY

15L, 455 HP
11L,350 HP
15L, 455 HP
11L, 350 HP
Steer Tire RR (kg/ton)
6.54
6.54
6.54
6.54
Drive Tire RR (kg/ton)
6.92
6.92
6.92
6.92
Drag Area, C,iA (m2)
6.0
5.6
6.0
5.6
Number of Trailer Axles
2
1
2
1
Trailer Tire RR (kg/ton)
6.00
6.00
6.00
6.00
Total Weight (kg)
31978
18306
33778
20106
Payload (tons)
19
10
19
10
Tire Pressure System Use
0
0
0
0
Weight Reduction (lb)
0
0
0
0
Drive Cycle Weightings




65-MPH Cruise
86%
64%
86%
64%
55-MPH Cruise
9%
17%
9%
17%
Transient Driving
5%
19%
5%
19%
2.10.3.2 Baseline Tractor-Trailer Vehicles to Evaluate Benefits and Costs
In order to evaluate the benefits and costs of the standards, it is necessary to establish a
reference point for comparison. The trailer technologies described in this section exist in the
market today, and their adoption is driven by available fuel savings as well as by the voluntary
SmartWay Partnership and California's Heavy Duty Greenhouse Gas Emission Reduction
Measure tractor-trailer requirements. To estimate the costs and benefits for these rules, the
agencies identified baseline tractor-trailers for each trailer subcategory based on the technology
adoption rates we project would exist if this trailer program was not implemented.
The agencies received comments suggesting our baseline adoption rates were too low for
several technologies and we made changes to our baseline trailers that in most cases should
address the comments. First, we created separate baselines for box vans that qualify as full-aero,
partial-aero and non-aero. We believe market forces will not significantly drive adoption of
CO2- and fuel-consumption reducing technologies for trailers with work performing equipment
(e.g., lift gates) and we are accordingly projecting zero adoption of the technologies in the
baselines for partial-and non-aero box vans. Similarly, we project zero adoption of these
technologies for the non-box trailers. We updated the baseline tire rolling resistance level for
non-box trailers to reflect the lower 6.5 kg/ton value in response to RMA's comment that these
trailers have different operational characteristics and should not have the same baseline tires as
box vans.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
An informal survey of TTMA members in 2014 indicated that 35 percent of long vans
and less than 2 percent of vans under 53-foot in length include aerodynamic devices, yet over 80
percent have adopted lower rolling resistance tires. The agencies believe the trailers for which
manufacturers have adopted these technologies are likely to be trailers that would qualify as
"full-aero" vans, and we created separate baselines to reflect these values. We project that
aerodynamics will increase to 40 percent adoption for full-aero long vans (dry and refrigerated)
and 5 percent for full-aero short vans by 2018 without this rulemaking. We project adoption of
lower rolling resistance tires (Level 3) to 90 percent and ATIS to 45 percent. We held these
adoption rates constant throughout the timeframe of the rules. Table 2-102 summarizes the
updated baseline trailers for each trailer subcategory.
Table 2-102 Adoption Rates and Average Performance Parameters for the Flat Baseline Trailers
TECHNOLOGY
LONG
SHORT
ALL PARTIAL-AERO,
ALL NON-BOX

VANS
VANS
NON-AERO VANS
TRAILERS
Aerodynamics
Bin I
55%
95%
100%
100%
Bin II

5%


Bin III
40%



Bin IV
5%



Bin V




Bin VI




Bin VII




Average Delta CdA (m2) a
0.2
0.0
0.0
0.0
Tire Rolling Resistance
Level 1



100%
Level 2
10%
10%
100%

Level 3
90%
90%


Level 4




Average Crr (kg/ton) a
5.2
5.2
6.0
6.5
Tire Pressure Systems
ATIS
45%
30%


TPMS




Average % Reduction a
0.5%
0.3%
0.0%
0.0%
Weight Reduction
Weight (lb) h
Notes:
a Combines adoption rates with performance levels shown in Table 2-99
b Weight reduction was not projected for the baseline trailers
Also shown in Table 2-102 are average aerodynamic performance (delta CdA), average
tire rolling resistance (CRR), and average reductions due to use of tire pressure and weight
reduction for each stage of the program. These values indicate the performance of theoretical
average tractor-trailers that the agencies project would be in use if no federal regulations were in
place for trailer CO2 and fuel consumption. These average tractor-trailer vehicles serve as
baselines for each trailer subcategory.
Because the agencies cannot be certain about future trends, we also considered a second
baseline. This dynamic baseline reflects the possibility that absent a Phase 2 regulation, there
will be continuing adoption of aerodynamic technologies in the long box trailer market after
2018 that reduce fuel consumption and CO2 emissions. This case assumes the research funded

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
and conducted by the federal government, industry, academia and other organizations will, after
2018, result in the adoption of additional aerodynamic technologies beyond the levels required to
comply with existing regulatory and voluntary programs. One example of such research is the
Department of Energy Super Truck program which had a goal of demonstrating cost-effective
measures to improve the efficiency of Class 8 long-haul freight trucks by 50 percent by 2015.°
This baseline assumes that by 2040, 75 percent of new full-aero long vans will be equipped with
SmartWay-verified aerodynamic devices. The agencies project that the lower rolling resistance
tires and ATIS adoption will remain constant. Table 2-103 shows the agencies' projected
adoption rates of technologies in the dynamic baseline.
Table 2-103 Projected Adoption Rates and Average Performance Parameters for the Dynamic Baseline for
Long Dry and Refrigerated Vans (all other trailers are the same as Table 2-102)
TECHNOLOGY
LONG DRY AND REFRIGERATED
Model Year
2018
2021
2024
2027
2040
Aerodynamics





Bin I
55%
50%
45%
40%
20%
Bin II





Bin III
40%
45%
50%
55%
75%
Bin IV
5%
5%
5%
5%
5%
Bin V





Bin VI





Bin VII





Average Delta CdA (m2) a
0.2
0.3
0.3
0.3
0.4
Tire Rolling Resistance





Level 1





Level 2
10%
10%
10%
10%
10%
Level 3
90%
90%
90%
90%
90%
Level 4





Average Crr (kg/ton) a
5.2
5.2
5.2
5.2
5.2
Tire Inflation





ATIS
45%
45%
45%
45%
45%
TPMS





Average % Reduction a
0.5%
0.5%
0.5%
0.5%
0.5%
Weight Reduction (lbs)





Weighth





Notes:
A blank cell indicates a zero value
a Combines adoption rates with performance levels shown in Table 2-99
b Weight reduction was not projected for the baseline trailers
The agencies applied the vehicle attributes from Table 2-101 and the average
performance values from Table 2-102 in the Phase 2 GEM vehicle simulation to calculate the
CO2 emissions and fuel consumption performance of the reference tractor-trailers. The results of
these simulations are shown in Table 2-104. We used these CO2 and fuel consumption values to
calculate the relative benefits of the standards. Note that the large difference between the per
0 Daimler Truck North America. SuperTruck Program Vehicle Project Review. June 19, 2014. Docket EPA-HQ-
OAR-2014-0827.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
ton-mile values for long and short trailers is due primarily to the large difference in assumed
payload (19 tons compared to 10 tons) and the differing drive cycles as seen in Table 2-101. The
small difference between the dry and refrigerated vans of the same length is due to the weight
difference between the subcategories. Refrigerated vans have an additional 1800 pounds added
to account for the TRU. The alternative baseline in Table 2-103 impacts the long-term
projections of benefits beyond 2027, which are analyzed in Chapters 5 through 7 of this RIA.
The non-box trailers and non-aero box vans are not included in this baseline analysis, because we
are adopting design standards for these trailers. As such, these trailers would not have standards
to meet. Instead, they would have minimum tire technology requirements.
Table 2-104 CO2 Emissions and Fuel Consumption Results for the Baseline Tractor-Trailers

FULL-AERO
DRY VAN
FULL-AERO
REFRIGERATED
VAN
PARTIAL-AERO
DRY VAN
PARTIAL-AERO
REFRIGERATED VAN
Length
Long
Short
Long
Short
Long
Short
Long
Short
CO2 Emissions
(g/ton-mile)
83.2
126.5
84.9
130.3
86.1
128.5
87.9
132.4
Fuel
Consumption
(gal/1000 ton-
miles)
8.17289
12.42633
8.33988
12.79961
8.45776
12.62279
8.63458
13.00589
2.10.4 Effectiveness and Costs of the Standards
The agencies evaluated several alternatives for the trailer program. The analysis below is
for the alternative we believe reflects the agencies' statutory authorities. This alternative is fully
implemented in model year (MY) 2027.
2.10.4.1 Projected Technology Adoption Rates for the Final Standards
The agencies designed the trailer program to have no averaging in MY 2018 through MY
2026. In those years, all box vans sold must meet the standards using any combination of
available technologies. In MY 2027, when the trailer manufacturers are more comfortable with
compliance and the industry is more familiar with the technologies, the agencies are adopting
averaging provisions to allow additional flexibility for the full-aero box van subcategories that
have the most stringent standards. See Section IV.F(5)(a) of the Preamble to this rulemaking for
additional information about averaging. Table 2-105 through Table 2-107 present sets of
assumed adoption rates for aerodynamic, tire, and tire pressure technologies that a manufacturer
could apply to meet the box van standards. Since the agencies are not adopting averaging for
MY 2018-MY 2026, the adoption rates consist of the combination of a single aerodynamic bin,
tire rolling resistance level, and tire pressure system. As mentioned previously, manufacturers
can choose other combinations to meet the standards.
The adoption rates in Table 2-98 begins with all long box trailers achieving current
SmartWay-level aerodynamics (Bin III) in MY 2018 with a stepwise progression to achieving
Bin V in 2024. The adoption rates for short box trailers assume no adoption of aerodynamic

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
devices in MY 2018, adoption of single aero devices in MY 2021, and combinations of devices
by MY 2024. The shorter lengths of these trailers can restrict the design of aerodynamic
technologies that fully match the SmartWay-like performance levels of long boxes and we don't
assume adoption at the same Bin-levels. We nevertheless expect that trailer and device
manufacturers will continue to innovate skirt, under-body, rear, and gap-reducing devices and
combinations to achieve improved aerodynamic performance on these shorter trailers.
The MY 2027 standards for the full-aero box vans are based on an averaging program.
The gradual increase in assumed adoption of aerodynamic technologies throughout the phase-in
to the MY 2027 standards recognizes that even though many of the technologies are available
today and technologically feasible throughout the phase-period, their adoption on the scale of the
program will likely take time. EPA's aerodynamic testing does not show technologies capable
of achieving Bin VI for long vans or Bin IV for short vans. As a result, we did not assume a
similar step-wise progression to 100 percent adoption of those bins. We do believe that the
interim standards provide an incentive to drive innovation over the 10 years leading up to MY
2027 and that aerodynamic improvements at these highest performance levels will be possible
when the program is fully implemented.
We are aware that there is already a high adoption of SmartWay-verified tires (Level 3)
and we expect most manufacturers will install these tires to meet the standards in MY 2018, and
will adopt even lower rolling resistance tires as they become available. By MY 2021, we project
that adoption of Level 4 tires will be used to meet the standards. The agencies are also assuming
that all box vans will adopt ATIS throughout the program, though manufacturers do have the
option to install TPMS if they would prefer to make up the difference using other technologies.
As mentioned previously, the agencies did not include weight reduction in their technology
adoption projections, but certain types of weight reduction could be used as a compliance
pathway.
The agencies proposed that the partial-aero box vans would track with the full-aero van
standards until MY 2024. Wabash commented that these trailers would not be able to meet
standards after MY 2021. The agencies reconsidered the partial-aero standards and recognize
that it would be difficult to meet the proposed MY 2024 standards without the use of multiple
devices and that partial-aero trailers, by definition, are restricted from using multiple devices.
For these reasons, the agencies redesigned the partial-aero standards, such that trailers with
qualifying work-performing equipment can meet standards that would be achievable with the use
of a single aerodynamic device throughout the program. The partial-aero standards do, however,
increase in stringency slightly in MY 2021 to reflect the use of improved lower rolling resistance
tires.
Similar to our analyses of the baseline cases, the agencies derived a single set of
performance parameters for each subcategory by weighting the performance levels included in
Table 2-99 by the corresponding adoption rates. These performance parameters represent a
compliant vehicle for each trailer subcategory and we present these values in the tables.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-105 Projected Adoption Rates and Average Performance Parameters for Long Box Vans
TECHNOLOGY
LONG BOX
DRY & REFRIGERATED VANS
Model Year
2018
2021
2024
2027
Aerodynamic Technologies
Bin I




Bin II




Bin III
100%



Bin IV

100%


Bin V


100%
30%
Bin VI



70%
Bin VII




Average Delta CdA (m2) a
0.4
0.7
1.0
1.3
Trailer Tire Rolling Resistance
Level 1




Level 2



5%
Level 3
100%



Level 4

100%
100%
95%
Average Crr (kg/ton) "
5.1
4.7
4.7
4.8
Tire Pressure Systems
ATIS
100%
100%
100%
100%
TPMS




Average Reduction (%>) a
1.2%
1.2%
1.2%
1.2%
Weight Reduction
Weight (lb) h
Notes:
A blank cell indicates a zero value
a Combines projected adoption rates with performance levels shown in Table 2-99
b This set of adoption rates did not apply any assumed weight reduction to meet these standards for these trailers

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-106 Projected Adoption Rates and Average Performance Parameters for Short Box Vans
TECHNOLOGY

SHORT BOX


DRY & REFRIGERATED VANS
Model Year
2018
2021
2024
2027
Aerodynamic Technologies
Bin I




Bin II

100%


Bin III


100%
40%
Bin IV



60%
Bin V




Bin VI




Bin VII




Average Delta CdA (m2) h
0.0
0.1
0.4
0.6
Trailer Tire Rolling Resistance
Level 1




Level 2



5%
Level 3
100%



Level 4

100%
100%
95%
Average Crr (kg/ton) b
5.1
4.7
4.7
4.8
Tire Pressure Systems
ATIS
100%
100%
100%
100%
TPMS




Average Reduction (%)c
1.2%
1.2%
1.2%
1.2%
Weight Reduction
Weight (lb) h
Notes:
A blank cell indicates a zero value
a The majority of short box trailers are 28 feet in length. We recognize that they are often operated in tandem, which limits the
technologies that can be applied (for example, boat tails).
b Combines projected adoption rates with performance levels shown in Table 2-99
c This set of adoption rates did not apply any assumed weight reduction to meet these standards for these trailers

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-107 Projected Adoption Rates and Average Performance Parameters for Partial-Aero Box Vans
TECHNOLOGY
PARTIAL-AERO
PARTIAL-AERO

LONG BOX VANS
SHORT BOX VANS
Model Year
2018
2021+
2018
2021+
Aerodynamic Technologies
Bin I




Bin II



100%
Bin III
100%
100%


Bin IV




Bin V




Bin VI




Bin VII




Bin VIII




Average Delta CjA (m2) h
0.4
0.4
0.0
0.1
Trailer Tire Rolling Resistance
Level 1




Level 2




Level 3
100%

100%

Level 4

100%

100%
Average Crr (kg/ton) b
5.1
4.7
5.1
4.7
Tire Pressure Systems
ATIS
100%
100%
100%
100%
TPMS




Average Reduction (%) c
1.2%
1.2%
1.2%
1.2%
Weight Reduction
Weight (lb) h
Notes:
A blank cell indicates a zero value
a Combines projected adoption rates with performance levels shown in Table 2-99
b This set of adoption rates did not apply weight reduction to meet these standards for these trailers
The adoption rates shown in these tables are one set of many possible combinations that
box trailer manufacturers could apply to achieve the same average stringency. If a manufacturer
chose these adoption rates, a variety of technology options exist within the aerodynamic bins,
and several models of LRR tires exist for the levels shown. Alternatively, technologies from
other aero bins and tire levels could be used to comply. It should be noted that van
manufacturers are not limited to specific aerodynamic and tire technologies, since these are
performance-based standards, and manufacturers will not be constrained to adopt any particular
way to demonstrate compliance. Certain types of weight reduction, for example, may be used as
a compliance pathway.
Non-aero box vans with two or more work-related special components, and non-box
trailers (tankers, flatbeds, and container chassis) are not shown in the tables above, because they
have design-based tire standards. These trailers will install tires that meet a specified rolling
resistance and tire pressure systems. A tire-based program significantly reduces the compliance
burden for these manufacturers by reducing the amount of tracking and eliminating the need to
run GEM (or utilize the equation derived from GEM). The agencies are adopting these tire-only
requirements in two stages. In MY 2018, manufacturers would be required to use tires meeting a
rolling resistance of Level 3 or better and install tire pressure systems on all non-aero box vans.
Non-box trailers would also need tire pressure systems, but their tire rolling resistance threshold

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
is Level 2. In model years 2021 and later, these trailers would continue to install tire pressure
systems, but an additional level of rolling resistance is required. At minimum, manufacturers of
non-aero box vans and non-box trailers must install TPMS to comply with the standard;
however, they have the option to install ATIS though they will not receive any additional credit
for doing so. The agencies are assuming, as shown in Table 2-108, that manufacturers of these
trailers would adopt TPMS at all stages of the program.
Table 2-108 Design Standard Tire Technology Requirements for the Non-Aero Box Van and Non-Box
Trailers
TECHNOLOGY
NON-AERO BOX VANS
NON-BOX TRAILERS
Model Year
2018
2021+
2018
2021+
Minimum CRR (kg/ton)
5.1
4.7
6.0
5.1
Tire Pressure System
TPMS or ATIS
TPMS or ATIS
TPMS or ATIS
TPMS or ATIS
2.10.4.2 Derivation of the Standards
The average performance parameters from the previous tables were applied as input
values to the GEM vehicle simulation to derive the HD Phase 2 fuel consumption and CO2
emissions standards for each subcategory of box trailers.
The standards are shown in Table 2-109 and Table 2-110. Over the four stages of the
trailer program, the full-aero box vans longer than 50 feet will reduce their CO2 emissions and
fuel consumption by two percent, five percent, seven percent and nine percent compared to their
flat baselines for each year in Table 2-104. Full-aero box vans 50-feet and shorter will achieve
reductions of one percent, two percent, four percent and six percent compared to their flat
baseline cases. The partial-aero long and short box van standards will reduce CO2 and fuel
consumption by six percent and four percent, respectively, by MY 2021. The design-based tires
standards for non-box trailers and non-aero box vans would provide reductions of two percent in
MY 2018 and three percent in MY 2021 and later.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-109 Standards for Full-Aero Box Vans
MODEL
YEAR
SUBCATEGORY
DRY VAN
REFRIGERATED VAN
Length
Long
Short
Long
Short
2018 -
2020
EPA Standard
(CO 2 Grams per Ton-Mile)
81.3
125.4
83.0
129.1
Voluntary NHTSA Standard
(Gallons per 1,000 Ton-Mile)
7.98625
12.31827
8.15324
12.68173
2021 -
2023
EPA Standard
(CO 2 Grams per Ton-Mile)
78.9
123.7
80.6
127.5
NHTSA Standard
(Gallons per 1,000 Ton-Mile)
7.75049
12.15128
7.91749
12.52456
2024 -
2026
EPA Standard
(CO 2 Grams per Ton-Mile)
77.2
120.9
78.9
124.7
NHTSA Standard
(Gallons per 1,000 Ton-Mile)
7.58350
11.87623
7.75049
12.24951
2027 +
EPA Standard
(CO 2 Grams per Ton-Mile)
75.7
119.4
77.4
123.2
NHTSA Standard
(Gallons per 1,000 Ton-Mile)
7.43615
11.72888
7.60314
12.10216
Table 2-110 Standards for Partial-Aero Box Vans
MODEL
YEAR
SUBCATEGORY
DRY VAN
REFRIGERATED VAN
LENGTH
LONG
SHORT
LONG
SHORT
2018 -2020
EPA Standard
(CO2 Grams per Ton-Mile)
81.3
125.4
83.0
129.1
Voluntary NHTSA Standard
(Gallons per 1,000 Ton-Mile)
7.98625
12.31827
8.15324
12.68173
2021 +
EPA Standard
(CO2 Grams per Ton-Mile)
80.6
123.7
82.3
127.5
NHTSA Standard
(Gallons per 1,000 Ton-Mile)
7.91749
12.15128
8.08448
12.52456
2.10.4.3 Projected Cost of Trailer Standards
The agencies evaluated technology costs for 53-foot dry and refrigerated vans and 28-
foot dry vans, which we believe are representative of the majority of trailers in the long and short
box trailer categories, respectively. Similar tire technology costs were assumed for the non-box
trailer subcategory. We identified costs for each technology package evaluated and projected out
the costs for each year of the program. A summary of the technology costs is included in Table
2-111 through Table 2-114 for the four phases of the trailer program, with additional details
available in RIA Chapter 2.12. Costs shown in the following tables are for the specific model
year indicated and are incremental to the average baseline costs, which includes some level of
adoption of these technologies as shown in Table 2-102. For example, the tire costs for the full-
aero subcategories are $l-$2, because there is already a very high adoption of LRR tires in the
baseline. Therefore, the technology costs in the following tables reflect the average cost
expected for each of the indicated trailer subcategories. Throughout the trailer program
discussion, the non-aero box van subcategory is treated as a single category, because all lengths
of these trailers have identical design standards. However, two costs for this subcategory are

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
shown to reflect the difference in the number of tires expected on the different length trailers
(i.e., long vans are assumed to have two axles and eight tires, while short vans have a single axle
and four tires).
Note that these costs do not represent actual costs for the individual components, because
some fraction of the component costs has been subtracted to reflect some use of these
components in the baseline. These costs include indirect costs via markups along with learning
impacts and also reflect estimated costs of the compliance process. For more on the estimated
technology costs exclusive of adoption rates, refer to Chapter 2.12.
Table 2-111 Trailer Technology Incremental Costs in the 2018 Model Year
(2013$)

LONG
VANS,
FULL
AERO
LONG
VANS,
PARTIAL
AERO
SHORT
VANS,
FULL
AERO
SHORT
VANS,
PARTIAL
AERO
LONG
VANS,
NO
AERO
SHORT
VANS,
NO
AERO
NON-
BOX
Aerodynamics
$367
$742
$0
$0
$0
$0
$0
Tires
$2
$40
$1
$20
$40
$20
$28
Tire inflation
system
$347
$659
$338
$494
$421
$210
$421
Total
$716
$1,441
$339
$514
$461
$231
$448
Table 2-112 Trailer Technology Incremental Costs in the 2021 Model Year
(2013$)

LONG
VANS,
FULL
AERO
LONG
VANS,
PARTIAL
AERO
SHORT
VANS,
FULL
AERO
SHORT
VANS,
PARTIAL
AERO
LONG
VANS,
NO
AERO
SHORT
VANS,
NO
AERO
NON-
BOX
Aerodynamics
$743
$679
$450
$475
$0
$0
$0
Tires
$17
$49
$9
$25
$49
$25
$23
Tire inflation
system
$321
$609
$313
$457
$389
$195
$389
Total
$1,081
$1,337
$772
$957
$438
$219
$412

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-113 Trailer Technology Incremental Costs in the 2024 Model Year
	(2013$)	

LONG
LONG
SHORT
SHORT
LONG
SHORT
NON-

VANS,
VANS,
VANS,
VANS,
VANS,
VANS,
BOX

FULL
PARTIAL
FULL
PARTIAL
NO
NO


AERO
AERO
AERO
AERO
AERO
AERO

Aerodynamics
$899
$645
$879
$451
$0
$0
$0
Tires
$11
$48
$6
$24
$48
$24
$27
Tire inflation







system
$294
$558
$286
$418
$357
$178
$357
Total
$1,204
$1,251
$1,171
$894
$405
$202
$383
Table 2-114 Trailer Technology Incremental Costs in the 2027 Model Year
	(2013$)	

LONG
VANS,
FULL
AERO
LONG
VANS,
PARTIAL
AERO
SHORT
VANS,
FULL
AERO
SHORT
VANS,
PARTIAL
AERO
LONG
VANS,
NO
AERO
SHORT
VANS,
NO
AERO
NON-
BOX
Aerodynamics
$1,069
$623
$921
$436
$0
$0
$0
Tires
$22
$44
$11
$22
$44
$22
$16
Tire inflation
system
$279
$529
$272
$397
$338
$169
$338
Total
$1,370
$1,196
$1,204
$855
$382
$191
$354
2.10.5 Evaluation of Compliance Option using GEM-Based Equation
EPA created the Greenhouse gas Emissions Model (GEM) as a compliance tool for
heavy-duty vehicles. Users provide specific performance parameters to the model and GEM
calculates CO2 emissions and fuel consumption results. As described previously, the Phase 2
GEM is designed to accept four performance variables as trailer inputs: change in drag area
(delta CdA), tire rolling resistance level (TRRL), tire pressure systems, and weight reduction
(WR). The reduction applied when using a tire pressure system is accounted for after the vehicle
simulation is complete. The other performance parameters directly impact the results of the
vehicle simulation, by changing the drag, rolling resistance and weight of the simulated vehicle.
We performed a sensitivity analysis for delta CdA, TRRL and WR to evaluate their effect
on the model's results. In the analysis to follow, all of the calculations are shown in terms of
CO2 emissions; use a conversion of 10,180 grams CO2 per gallon of diesel fuel to calculate the
corresponding fuel consumption values. Figure 2-74 through Figure 2-77 show GEM's CO2
results from the proposal for a parameter sweep of a simulated Class 8 tractor pulling each of the
four box van trailers. It can be seen that each of the three parameters has a linear impact on CO2
emissions. A curve fit was applied to each data set and the equation is displayed on each plot.
The intercept in each parameter sweep data set is the baseline CO2 result considering a no-
control trailer, and this value is consistent for all parameters for a given trailer. The coefficients
indicate the relationship between the assessed parameter and the model's CO2 result. A similar
analysis was repeated with the GEM version that was updated since the NPRM. The coefficients
of the regression curves differ, but the trends remain the same.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Impact of AC DA Parameter
Long Dry Van, TRRL = 6.0 kg/ton, WR = 0 lb
90
¦| 85
¦
c
o
2 80
o
° 75
70
¦	



	
••	
	-



		
"	_
y = -5.8216X
+ 86.066

'¦
0.0
0.5
1.0
Delta CDA (m2)
1.5
2.0
(a)
Impact of ATRRL Parameter
Long Dry Van, ACdA = 0 m2, WR = 0 lb
90
I 85
i
C
o
% 80
U)
oj
8 75
70
¦	















y = -1.667x + 86.072



-1.0 -0.5 0.0 0.5 1.0
Delta TRRL (kg/ton)
1.5
2.0
(b)
Impact ofWR Parameter
Long Dry Van, ACdA = 0 m2, TRRL = 6.0 kg/ton
90
f 85
l
c
o
2 80

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Impact of ACdA Parameter
Long Reefer, TRRL = 6.0 kg/ton, WR = 0 lb
90
¦| 85
¦
c
o
3 80
CM
O
° 75
y = -5.7834X +87.915
70
0.0	0.5	1.0	1.5	2.0
Delta CDA (m2)
(a)
Impact of ATRRL Parameter
Long Reefer, ACDA = 0 m2, WR = 0 lb
90
f 85
C
0
1	80
CM
o
o 75










y = -1.7506x +87.919



70
-1.0 -0.5 0.0 0.5 1.0 1.5 2.0
Delta TRRL (kg/ton)
(b)
Impact of WR Parameter
Long Reefer, ACDA = 0 m2, TRRL = 6.0 kg/ton
90
£ 85
C
o
3 80
CM
O
° 75
70
y = -0.001X + 87.896
0	1000 2000 3000 4000 5000
Weight Reduction (lb)
(c)
Figure 2-75 Impact of (a) Delta CdA, (b) Delta Crr, and (c) Weight Reduction on CO2 Results of a GEM-
Simulated Long Refrigerated Van

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Impact of ACdA Parameter
Short Dry Van, TRRL = 6.0 kg/ton, WR = 0 lb
140
- 130
E
I 120
3
n 110
o
o
100
y = -9.4835X + 130.18
90
0.0	0.5	1.0	1.5	2.0
Delta CDA (m2)
(a)
Impact of ATRRL Parameter
Short Dry Van, ACDA = 0 m2, WR = 0 lb
140
_ 130
I
| 120
D)
-110
o
° 100
B	





















y = -1.7801x + 128.43



90
-1.0 -0.5 0.0 0.5 1.0 1.5 2.0
Delta TRRL (kg/ton)
(b)
Impact of WR Parameter
Short Dry Van, ACDA = 0 m2, TRRL = 6.0 kg/ton
140
=r130
£
i
o 120
O 110
o
100
90
y = -0.0026X + 128.33
0	1000 2000 3000 4000 5000
Weight Reduction (lb)
(c)
Figure 2-76 Impact of (a) Delta CdA, (b) Delta Crr, and (c) Weight Reduction on CO2 Results of a GEM-
Simulated Short Dry Van

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Impact of ACdA Parameter
Short Reefer, TRRL = 6.0 kg/ton, WR = 0 lb

140

130
1

c
0
120
3

CM
110
O

O


100

90
y = -9.3557x + 134.25
0.0	0.5	1.0	1.5	2.0
Delta CDA (m2)
(a)
Impact of ATRRL Parameter
Short Reefer, ACDA = 0 m2, WR = 0 lb
140
130
§ 120
1	






















y = -1.8839x+132.37



110
100
90
-1.0 -0.5 0.0 0.5 1.0 1.5 2.0
Delta TRRL (kg/ton)
(b)
Impact of WR Parameter
Short Reefer, ACDA = 0 m2, TRRL = 6.0 kg/ton
140
^130
E
I 120
g 110
o
100
90
y = -0.0026x + 132.3
0	1000 2000 3000 4000 5000
Weight Reduction (lb)
(c)
Figure 2-77 Impact of (a) Delta CdA, (b) Delta Crr, and (c) Weight Reduction on CO2 Results of a GEM-
Simulated Short Refrigerated Van
Additional GEM simulations were performed for each of the four box trailer
subcategories to assess the combined effect of these parameters. As seen in Figure 2-78 and
Figure 2-79 for the long dry van simulation, the coefficients of the curve fit equations were not

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
significantly changed, indicating that the combined impacts of these parameters on GEM's CO2
results were additive. Similar trends were seen with the simulations for the other trailer
subcategories, though the results are not shown here.
90
— 85

c
o
80
O)
Combined Impact of ACdA and TRRL
(WR = 0 lb)
¦TRRL=7.0, WR=0 OTRRL=6.0, WR=0 ATRRL=4, WR=0
o
A
o
o
75
70
y = -5.8x +87.7
y = -5.8x + 86.1
y = -5.8x + 82.7
0.0
0.5
o
A
1.0
Delta CdA (m2)
o
A
1.5
o
A
2.0
Figure 2-78 Combined Impact of Drag Area and Tire Rolling Resistance Level on CO2 Results of a GEM-
Simulated Long Dry Van with No Weight Reduction
Combined Impact of ACdA and WR
(TRRL = 6.0 kg/ton)
90
C
O
^80
O)
O
0 75
70
¦TRRL=6.0, WR=0 OTRRL=
J*	
6.0, WR=1000 ATRRL=6, WR=2000

1.

y = -5.8x +86.1

	 .
y = -5.8x +85.0
y = -5.7x + 84.0

-	
0.0
0.5
1.0
Delta CdA (m2)
1.5
2.0
Figure 2-79 Combined Impact of Drag Area and Weight Reduction on CO2 Results of a GEM-Simulated
Long Dry Van at a Tire Rolling Resistance Level of 5.1 kg/ton
The results presented Figure 2-78 and Figure 2-79 suggest that these parameters could be
combined into a single equation to calculate CO2 emissions. Equation 2-3 is the result of
combining the updated curve fit equations for long box dry vans.
Equation 2-3 Combination of Curve Fit Equations for Long Dry Van GEM Input Parameters
y = 86.1-1. 7(ATRRL) - 5.8(ACDA) - 0. 0010(Wi?)

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Our regulations specify that TRRL be an absolute measure of a tire's coefficient of
rolling resistance (not a change in rolling resistance). As a result, Equation 2-3 was modified
such that the variables of the equation matched the trailer inputs required by GEM. Equation 2-4
is the resulting equation.
Equation 2-4 Modified Equation for Long Dry Vans to Account for TRRL Input Parameter
y = 16.1 + 1. 7(TRRL) - 5.8(ACDA) - 0. 0010(Wi?)
Each of the trailer subcategories follows the same general format and a generic equation
is shown in Equation 2-5. Table 2-115 summarizes the corresponding constants for each of the
trailer subcategories.
Equation 2-5 General GEM-Based CO2 Equation for Trailer Subcategories
eC02 = Ct + C2(TRRL) + C3(ACdA) + C4(WR)
Table 2-115 Constants for GEM-Based CO2 Equation for Trailer Subcategories (See Equation 2-5)
TRAILER SUBCATEGORY
Ci
c2
c3
c4
Long Dry Van
76.1
1.67
-5.82
-0.00103
Long Refrigerated Van
77.4
1.75
-5.78
-0.00103
Short Dry Van
117.8
1.78
-9.48
-0.00258
Short Refrigerated Van
121.1
1.88
-9.36
-0.00264
Over 100 GEM vehicle simulations were performed for a range of delta CdA, TRRL and
weight reduction values. The results of these simulations were compared to CO2 results
calculated using Equation 2-5 for each trailer subcategory. The following figures show the
equation and GEM have nearly identical CO2 results.

-------
*** E. O. 12866 Review — Revised - Do Not Cite, Quote, or Release During Review ***
Compare GEM and Calculated C02 Results
Long Dry Van
y = 1.02x-1.6274
Ra = 0.9995
« 80
a> 70
CO2 = 76.1 + 1.7 CRR - 5.8 ACdA - 0.0010 WR
o 65
GEM Simulation CQ2 Result (g/ton-mi)
Figure 2-80 Comparison of GEM and Calculated CO2 Results for a Long Dry Van
Compare GEM and Calculated CQ2 Results
Long Refrigeraged Van
y = 1.0198X-1.6407
R! = 0.9995
O 75
CO2 = 77.4 + 1.8 CRR - 5.8 ACdA - 0.0010 WR
iJ 65
95
GEM Simulation C02 Result (g/ton-mi)
Figure 2-81 Comparison of GEM and Calculated CO2 Results for a Long Refrigerated Van

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Compare GEM arid Calculated C02 Results
_	Short Dry Van
| 135
O 130
3 125
3 120
w
« 115
(M 110
0	105
"g 100
1	95
o 90
o 90 95 100 105 110 115 120 125 130 135
GEM Simulation C02 Result (g/ton-mi)
Figure 2-82 Comparison of GEM and Calculated CO2 Results for a Short Dry Van
Compare GEM and Calculated C02 Results
_	Short Refrigerated Van
? 140
| 135
3 130
3 125
| 120
cm 115
o 110
~5 105
« 100
J 95
O 95 100 105 110 115 120 125 130 135 140
Calculated C02 Result (g/ton-mi)
Figure 2-83 Comparison of GEM and Calculated CO2 Results for a Short Refrigerated Van
The comparisons shown in Figure 2-80 through Figure 2-83 suggest that an equation may offer a simplified
approach for trailer manufacturers to calculate CO2 without the use of GEM. Equation 2-6 below is a slight
modification to Equation 2-5. As mentioned previously, the trailer program is also offering the use of tire
pressure systems as a means achieving the standards. This parameter is not considered in Equation 2-5.
Equation 2-6 includes a constant, Cs, to address the use of tire pressure systems. Constant Cs is equal to
unity (1.0) for trailers that do not have tire pressure systems installed, equal to 0.988 (accounting for the 1.2
percent reduction) for trailers that include ATIS, and equal to 0.990 for trailers that include TPMS. As
y = 1.0336X-3.9652
Rs = 0.9991
C02 = 118 + 1.8 CRR-9.5 ACdA - 0.0026 WR
y = 1.0327X -4.0149
R2 = 0.9992
CO2 = 121 + 1.9 CRR - 9.3 ACdA - 0.0027 WR

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
mentioned previously, one can use a conversion factor of 10,180 grams CO2 per gallon of diesel fuel to
calculate the corresponding fuel consumption values.
Table 2-116 summarizes the constants available to manufacturers when using Equation
2-6 for compliance.
Equation 2-6 GEM-Based Compliance Equation for Phase 2 Trailer Program
eC02 = [Ci + C2 ¦ (TRRL) + C3 ¦ (ACDA) + C4 ¦ (WR)] ¦ C5
Table 2-116 Constants for GEM-Based CO2 Equation for Trailer Subcategories (See Equation 2-6)
TRAILER
SUBCATEGORY
Ci
C2
C3
c4
C5
No Tire
Pressure
System
ATIS
Installed
TPMS
Installed
Long Dry Van
76.1
1.67
-5.82
-0.00103
1.000
0.988
0.990
Long Refrigerated Van
77.4
1.75
-5.78
-0.00103
Short Dry Van
117.8
1.78
-9.48
-0.00258
Short Refrigerated Van
121.1
1.88
-9.36
-0.00264
The updates to GEM that were made following the NPRM impacted the trailer model and
resulted in a change to the constants for the GEM-based compliance equation that will be used
by trailer manufacturers. We repeated the process of generating and validating the new
constants, and, similar to the proposal, these updated values accurately recreate the GEM
calculations for each trailer subcategory. Consequently, the agencies are adopting this equation-
based compliance approach with the new constants shown in Table 2-116 for the final Phase 2
trailer program.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.11 Technology Costs
2.11.1 Overview of Technology Cost Methodology Learning Effects on
Technology Costs
Chapter 2.11.1.2 presents the methods used to address indirect costs in this analysis.
Chapter 2.11.1.3 presents the learning effects applied throughout this analysis. In Chapter 2.11.2
through 2.11.10 we present individual technology costs including: the direct manufacturing costs
(DMC), their indirect costs (IC) and their total costs (TC, TC=DMC+IC). Note that we also
present technology penetration rates for most technologies and the resultant total cost as applied
to a technology package (which we have denoted as TCp, where TCp=TC x Adoption Rate).
The tables presented show the adoption rate for, generally, alternatives la and 3 where la
represents the reference case (or the "no action" case) and 3 represents the preferred policy case
(i.e. the standards adopted in this final rule). Note also that some TCp values appear as negative
values in some tables (notably the lower rolling resistance (LRR) tire tables). This is because
certain LRR tires are expected in the reference case but are then expected to be removed in the
policy case and replaced by more aggressive LRR tires. In such cases, the reference case tires
show negative TCp costs since they are being removed and replaced.
2.11.1.1 Direct Manufacturing Costs
The direct manufacturing costs (DMCs) used throughout this analysis are derived from
several sources. Many of the tractor, vocational and trailer DMCs can be sourced to the Phase 1
rules which, in turn, were sourced largely from a contracted study by ICF International for
EPA.194 There was no serious disagreement regarding these estimated costs in the public
comments to the Phase 1 rules. We have updated those costs by converting them to 2012 dollars,
as described in Section IX.B.l.e of the Preamble, and by continuing the learning effects
described in the Phase 1 rules and in Section IX. B.l.c of the Preamble. The new tractor,
vocational and trailer costs can be sourced to a more recent study conducted by Southwest
Research Institute (SwRI) under contract to NHTS A.195 The cost methodology used by SwRI in
that study was to estimate retail costs then work backward from there to derive a DMC for each
technology. The agencies did not agree with the approach used by Tetra Tech to move from
retail cost to DMC as it disagreed with EPA's look at retail price equivalents in the HD engine
and truck industry on which EPA has based the indirect cost markup approach to estimate
indirect costs, as discussed more in Chapter 2.11.1.2. As such, the agencies have used an
approach consistent with past GHG/CAFE/fuel consumption rules by dividing estimated retail
prices by our estimated retail price equivalent markups to derive an appropriate DMC for each
technology. We describe our RPEs in Chapter 2.11.1.2.
For HD pickups and vans, we have relied primarily on the Phase 1 rules and the light-
duty 2017-2025 model year rule since most technologies expected on these vehicles are, in
effect, the same as those used on light-duty pickups. Many of those technology DMCs are based
on cost teardown studies which the agencies consider to be the most robust method of cost
estimation. However, many of the HD versions of those technologies would be expected to be
more costly than their light-duty counterparts because of the heavier HD vehicles and/or the
higher power and torque characteristics of their engines. Therefore, we have scaled upward

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
where appropriate many of the light-duty DMCs for this analysis. We have also used some costs
developed under contract to NHTSA by SwRI (the study mentioned above).196
Importantly, in our methodology, all technologies are treated as being sourced from a
supplier rather than being developed and produced in-house. As such, some portion of the total
indirect costs of making a technology or system—those costs incurred by the supplier for
research, development, transportation, marketing etc.—are contained in the sales price to the
engine and/or vehicle manufacturer (i.e., the original equipment manufacturer (OEM)). That sale
price paid by the OEM to the supplier is the DMC we estimate.
2.11.1.2 Indirect Costs
To produce a unit of output, engine and truck manufacturers incur direct and indirect
costs. Direct costs include cost of materials and labor costs. Indirect costs are all the costs
associated with producing the unit of output that are not direct costs - for example, they may be
related to production (such as research and development [R&D]), corporate operations (such as
salaries, pensions, and health care costs for corporate staff), or selling (such as transportation,
dealer support, and marketing). Indirect costs are generally recovered by allocating a share of
the costs to each unit of a good sold (e.g., an engine, a truck, etc.). Although it is possible to
account for direct costs allocated to each unit of good sold, it is more challenging to account for
indirect costs allocated to a unit of goods sold. To make a cost analysis process more feasible,
markup factors, which relate total indirect costs to total direct costs, have been developed. These
factors are often referred to as retail price equivalent (RPE) multipliers.
Cost analysts and regulatory agencies (including both EPA and NHTSA) have frequently
used these multipliers to predict the resultant impact on costs associated with manufacturers'
responses to regulatory requirements. The best approach, if it were possible, to determining the
impact of changes in direct manufacturing costs on a manufacturer's indirect costs would be to
actually estimate the cost impact on each indirect cost element. However, doing this within the
constraints of an agency's time or budget is not always feasible, or the technical, financial, and
accounting information to carry out such an analysis may simply be unavailable.
RPE multipliers provide, at an aggregate level, the relative shares of revenues (Revenue =
Direct Costs + Indirect Costs + Net Income) to direct manufacturing costs. Using RPE
multipliers implicitly assumes that incremental changes in direct manufacturing costs produce
common incremental changes in all indirect cost contributors as well as net income. However, a
concern in using the RPE multiplier in cost analysis for new technologies added in response to
regulatory requirements is that the indirect costs of vehicle modifications are not likely to be the
same for different technologies. For example, less complex technologies could require fewer
R&D efforts or less warranty coverage than more complex technologies. In addition, some
simple technological adjustments may, for example, have no effect on the number of corporate
personnel and the indirect costs attributable to those personnel. The use of RPEs, with their
assumption that all technologies have the same proportion of indirect costs, is likely to
overestimate the costs of less complex technologies and underestimate the costs of more
complex technologies. Table 2-117 shows the RPE factors used in developing indirect costs in
past, and this, agency analyses.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-117 Industry Retail Price Equivalent (RPE) Factors
INDUSTRY
RPE
Heavy engine manufacturers
1.28
Heavy truck manufacturers
1.36
Light-duty vehicle manufacturers
1.50
To address this concern, modified multipliers have been developed by EPA, working
with a contractor, for use in rulemakings. These multipliers are referred to as indirect cost
multipliers (or ICMs). In contrast to RPE multipliers, ICMs assign unique incremental changes
to each indirect cost contributor as well as net income.
ICM = (direct cost + adjusted indirect cost)/(direct cost)
Developing the ICMs from the RPE multipliers requires developing adjustment factors
based on the complexity of the technology and the time frame under consideration: the less
complex a technology, the lower its ICM, and the longer the time frame for applying the
technology, the lower the ICM. This methodology was used in the cost estimation for the recent
light-duty MYs 2012-2016 and MYs 2017-2025 rulemaking and for the heavy-duty MYs 2014-
2018 rulemaking. There was no serious disagreement with this approach in the public comments
to any of these rulemakings. The ICMs for the light-duty context were developed in a peer-
reviewed report from RTI International and were subsequently discussed in a peer-reviewed
journal article.197 Importantly, since publication of that peer-reviewed journal article, the
agencies have revised the methodology to include a return on capital (i.e., profits) based on the
assumption implicit in ICMs (and RPEs) that capital costs are proportional to direct costs, and
businesses need to be able to earn returns on their investments.
For the heavy-duty pickup truck and van cost projections in this rule, the agencies have
used ICM adjustment factors developed for light-duty vehicles, inclusive of a return on capital,
primarily because the manufacturers involved in this segment of the heavy-duty market are the
same manufacturers that build light-duty trucks.
For the combination tractors, vocational vehicles, and heavy-duty engine cost projections
in this rule, the agencies are again using the ICMs used in the HD Phase 1 rules. Those ICMs
were developed by RTI International under EPA contract to update EPA's methodology for
accounting for indirect costs associated with changes in direct manufacturing costs for heavy-
duty engine and truck manufacturers.198 In addition to the indirect cost contributors varying by
complexity and time frame, there is no reason to expect that the contributors would be the same
for engine manufacturers as for truck manufacturers. The resulting report from RTI provides a
description of the methodology, as well as calculations of the indirect cost multipliers that are
being used as the basis for the markups used in this rule. These indirect cost multipliers were
used, along with calculations of direct manufacturing costs, to provide estimates of the full
additional costs associated with new technologies.
As explained in the Phase 1 final rules, and entirely consistent with the analysis
supporting that program, the agencies have made some changes to both the ICM factors and to
the method of applying those factors relative to the factors developed by RTI and presented in

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
their reports. The first of these changes was done in response to continued thinking among the
agencies about how past ICMs have been developed and about which data sources are the most
appropriate on which to rely in determining the appropriate ICMs. The second change was done
in response to both staff concerns and public feedback suggesting that the agencies were
inappropriately applying learning effects to indirect costs via the multiplicative approach to
applying the ICMs.
Regarding the first change - to the ICM factors themselves - a little background must
first be provided. In the original work done under contract to EPA by RTI International,199 EPA
experts had undergone a consensus approach to determining the impact of specific technology
changes on the indirect costs of a company. Subsequently, EPA experts underwent a blind
survey to make this determination on a different set of technology changes. This subsequent
effort, referred to by EPA as a modified-Delphi approach, resulted in different ICM
determinations. This effort is detailed in a memorandum contained in the docket for this
rulemaking.200 Upon completing this effort, EPA determined that the original RTI values should
be averaged with the modified-Delphi values to arrive at the final ICMs for low and medium
complexity technologies and that the original RTI values would be used for high complexity
level 1 while the modified-Delphi values would be used for high complexity level 2. These final
ICMs were used in the 2012-2016 light-duty GHG/CAFE rulemaking. Subsequent to that, EPA
contracted with RTI to update their light-duty report with an eye to the heavy-duty industry. In
that effort, RTI determined the RPE of both the heavy-duty engine and heavy truck industries,
then applied the light-duty indirect cost factors—those resulting from the averaging of the values
from their original report with the modified-Delphi values—to the heavy-duty RPEs to arrive at
heavy-duty specific ICMs. That effort is described in their final heavy-duty ICM report
mentioned above.201
During development of the Phase 1 heavy-duty final rules, the agencies decided that the
original light-duty RTI values, given the technologies considered for low and medium
complexity, should no longer be used and that we should rely solely on the modified-Delphi
values for these complexity levels. The original light-duty RTI study used low rolling resistance
tires as a low complexity technology example and a dual clutch transmission as a medium
complexity technology. Upon further thought, the technologies considered for the modified
Delphi values (passive aerodynamic improvements for low complexity and turbocharging with
downsizing for medium complexity) were considered by the agencies to better represent the
example technologies. As a result, the modified-Delphi values were to become the working
ICMs for low and medium complexity rather than averaging those values with the original RTI
report values. The agencies have also re-examined the technology complexity categories that
were assigned to each light-duty technology and modified these assignments to better reflect the
technologies that are now used as proxies for each category. This decision impacted the low and
medium complexity heavy-duty ICMs too because the modified-Delphi values alone were to be
applied to the heavy-duty RPEs to arrive at heavy-duty ICMs rather than using the averaged
values developed for the light-duty 2012-2016 rulemaking.
A secondary-level change was also made as part of this ICM recalculation to the light-
duty ICMs and, therefore, to the ICMs used in the Phase 1 HD final rules and again in this
analysis for HD pickups and vans. That change was to revise upward the RPE level reported in
the original RTI report from an original value of 1.46 to 1.5 to reflect the long term average RPE.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The original RTI study was based on 2008 data. However, an analysis of historical RPE data
indicates that, although there is year to year variation, the average RPE has remained roughly
1.5. ICMs are applied to future year's data and therefore the agencies believed and continue to
believe that it is most appropriate to base ICMs on the historical average rather than a single
year's result. Therefore, ICMs were adjusted to reflect this average level. As a result, the High 1
and High 2 ICMs used for HD pickups and vans were changed for the Phase 1 final rules and we
continue to use those changed values here.
Table 2-118 shows the ICM values used in this rule. Near term values are used in early
years, depending on the technology, and account for differences in the levels of R&D, tooling,
and other indirect costs that would be incurred. Once the program has been fully implemented,
some of the indirect costs would no longer be attributable to the standards and, as such, a lower
ICM factor is applied to direct costs in later years.
Table 2-118 Indirect Cost Multipliers Used in this Analysis"
CLASS
COMPLEXITY
NEAR
TERM
LONG
TERM
HD Pickup Trucks and Vans
Low
1.24
1.19
Medium
1.39
1.29
Highl
1.56
1.35
High2
1.77
1.50
Loose diesel engines
Low
1.15
1.13
Medium
1.24
1.18
Highl
1.28
1.19
High2
1.44
1.29
Loose gasoline engines
Low
1.24
1.19
Medium
1.39
1.29
Highl
1.56
1.35
High2
1.77
1.50
Vocational Vehicles,
Combination Tractors and
Trailers
Low
1.18
1.14
Medium
1.30
1.23
Highl
1.43
1.27
High2
1.57
1.37
Note:
a Rogozhin, A., et. al., "Using indirect cost multipliers to estimate the total cost of adding
new technology in the automobile industry," International Journal of Production
Economics (2009); "Documentation of the Development of Indirect Cost Multipliers for
Three Automotive Technologies," Helfand, G., and Sherwood, T., Memorandum dated
August 2009; "Heavy Duty Truck Retail Price Equivalent and Indirect Cost Multipliers,"
Report prepared by RTI International and Transportation Research Institute, University of
Michigan, July 2010.
The second change made to the ICMs during development of the Phase 1 final rules had
to do with the way in which the ICMs were applied. Until that time, we had applied the ICMs,
as done in any analysis that relied on RPEs, as a pure multiplicative factor. This way, a direct
manufacturing cost of, say, $100 would have been multiplied by an ICM of 1.24 to arrive at a
marked up technology cost of $124. However, as learning effects (discussed below) are applied
to the direct manufacturing cost, the indirect costs are also reduced accordingly. Therefore, in
year 2 the $100 direct manufacturing cost might reduce to $97 and the marked up cost would

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
become $120 ($97 x 1.24). As a result, indirect costs have been reduced from $24 to $23. Given
that indirect costs cover many things such as facility-related costs, electricity, etc., it is perhaps
not appropriate to apply the ICM to the learned direct costs, at least not for those indirect cost
elements unlikely to change with learning. The agencies decided that it was more appropriate
only to allow warranty costs to decrease with learning since warranty costs are tied to direct
manufacturing costs (since warranty typically involves replacement of actual parts which should
be less costly with learning).p However, the remaining elements of the indirect costs should
remain constant year-over-year, at least until some of those indirect costs are no longer
attributable to the rulemaking effort that imposed them (such as R&D).
As a result, the ICM calculation became more complex with the analysis supporting the
Phase 1 final rules, and we continue to use that more complex calculation here. We first
establish the year in which the direct manufacturing costs are considered "valid." For example, a
cost estimate might be considered valid today, or perhaps not until high volume production is
reached in some future model year. That year is considered the base year for the estimated cost.
That cost is the cost used to determine the "non-warranty" portion of the indirect costs. For
example, the near term non-warranty portion of the loose diesel engine low complexity ICM is
0.149 (the warranty versus non-warranty portions of the ICMs are shown in Table 2-119). For
the improved water pump technology we have estimated a direct manufacturing cost of $82.66
(2012$) in MY 2014. So the non-warranty portion of the indirect costs would be $12.32 ($82.66
x 0.149). This value would be added to the learned direct manufacturing cost for each year
through 2022 since the near term markup is considered appropriate for that technology through
2022. Beginning in 2023, when long-term indirect costs begin, the additive factor would become
$10.08 ($82.66 x 0.122). Additionally, the $82.66 cost in 2014 would become $80.18 in MY
2015 due to learning ($82.66 x (1-3 percent)). So, while the warranty portion of the indirect
costs would be $0.49 ($82.66 x 0.006) in 2014, they would decrease to $0.48 ($80.18 x 0.006) in
2015 as warranty costs decrease with learning. The resultant indirect costs for the water pump
would be $12.81 ($12.32+$0.49) in MY 2014 and $12.80 ($12.32+$0.48) in MY2015, and so on
for subsequent years.
Importantly, since the bulk of the indirect costs calculated using this methodology are the
non-warranty costs, and since those costs do not change over with learning, one cannot look at
the ICMs shown in Table 2-118 and assume that our HD pickup and van total costs are, in
general, 1.24 or 1.39 times the direct costs (since most technologies considered for application in
HD pickups and vans are low and medium technologies). This can be illustrated by building on
the example presented above for a water pump on a heavy diesel engine. We already calculated
the MY 2014 total cost as $95.46 (2012$, $82.66+$12.32+$0.49). This is an effective markup of
1.155 ($95.46/$82.66). This is expected since the cost is based in 2014 and the near term ICM is
1.155. In MY2022, the final year of near term markups for this technology, the total cost would
be $80.21 since the learned direct cost has reduced to $67.50, the non-warranty indirect costs
(calculated above) remain $12.32, and the warranty indirect costs have become $0.39
($67.50x0.006). So, in MY2022, we now have an effective markup of 1.19 ($80.21/$67.50).
p We note that the labor portion of warranty repairs does not decrease due to learning. However, we do not have
data to separate this portion and so we apply learning to the entire warranty cost. Because warranty costs are a small
portion of overall indirect costs, this has only a minor impact on the analysis.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-119 Warranty and Non-Warranty Portions of ICMs


SHORT-TERM
LONG-TERM
CLASS
COMPLEXITY
WARRANTY
NON-
WARRANTY
WARRANTY
NON-
WARRANTY
HD Pickup and
Vans
Low
0.012
0.230
0.005
0.187
Medium
0.045
0.343
0.031
0.259
Highl
0.065
0.499
0.032
0.314
High2
0.074
0.696
0.049
0.448
Loose diesel
engines
Low
0.006
0.149
0.003
0.122
Medium
0.022
0.213
0.016
0.165
Highl
0.032
0.249
0.016
0.176
High2
0.037
0.398
0.025
0.265
Loose gasoline
engines
Low
0.012
0.230
0.005
0.187
Medium
0.045
0.343
0.031
0.259
Highl
0.065
0.499
0.032
0.314
High2
0.074
0.696
0.049
0.448
Vocational
Vehicles,
Combination
Tractors and
Trailers
Low
0.013
0.165
0.006
0.134
Medium
0.051
0.252
0.035
0.190
Highl
0.073
0.352
0.037
0.233
High2
0.084
0.486
0.056
0.312
The complexity levels and subsequent ICMs applied throughout this analysis for each
technology are shown in Table 2-120. One notable change since the proposal is to waste heat
recovery which used a short term markup through 2025 in the proposal but uses that markup
through 2027 in this final rule.
Table 2-120 Indirect Cost Markups and Near Term/Long Term Cutoffs Used in this Analysis
TECHNOLOGY
APPLIED TO
ICM
NEAR TERM


COMPLEXITY
THRU
Cylinder head improvements 1
LH/MH/HH Engines
Low
2022
Cylinder head improvements 2
LH/MH/HH Engines
Low
2027
Turbo efficiency improvements 1
LH/MH/HH, HD Pickup & Van Engines
Low
2022
Turbo efficiency improvements 2
LH/MH/HH Engines
Low
2027
EGR cooler efficiency improvements 1
LH/MH/HH Engines
Low
2022
EGR cooler efficiency improvements 2
LH/MH/HH Engines
Low
2027
Water pump improvements 1
LH/MH/HH Engines
Low
2022
Water pump improvements 2
LH/MH/HH Engines
Low
2027
Oil pump improvements 1
LH/MH/HH Engines
Low
2022
Oil pump improvements 2
LH/MH/HH Engines
Low
2027
Fuel pump improvements 1
LH/MH/HH Engines
Low
2022
Fuel pump improvements 2
LH/MH/HH Engines
Low
2027
Fuel rail improvements 1
LH/MH/HH Engines
Low
2022
Fuel rail improvements 2
LH/MH/HH Engines
Low
2027
Fuel injector improvements 1
LH/MH/HH Engines
Low
2022
Fuel injector improvements 2
LH/MH/HH Engines
Low
2027
Piston improvements 1
LH/MH/HH Engines
Low
2022
Piston improvements 2
LH/MH/HH Engines
Low
2027
Valve train friction reductions 1
LH/MH/HH Engines
Low
2022
Valve train friction reductions 2
LH/MH/HH Engines
Low
2027
Turbo compounding 1
LH/MH/HH Engines
Low
2022
Turbo compounding 2
LH/MH/HH Engines
Low
2027
Aftertreatment improvements 1
LH/MH/HH Engines
Low
2022

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Aftertreatment improvements 2
LH/MH/HH Engines
Low
2024
Model based control
LH/MH/HH Engines
Low
2022
Waste heat recovery
HH Engines
Medium
2027
Engine friction reduction 1
HD Pickup & Van Engines
Low
2018
Engine friction reduction 2
HD Pickup & Van Engines
Low
2024
Engine changes to accommodate low friction
lubes
HD Pickup & Van Engines
Low
2018
Variable valve timing - coupled
HD Pickup & Van Engines
Low
2018
Variable valve timing - dual
HD Pickup & Van Engines
Medium
2018
Stoichiometric gasoline direct injection
HD Pickup & Van Engines
Medium
2018
Cylinder deactivation
HD Pickup & Van Engines
Medium
2018
Cooled EGR
HD Pickup & Van Engines
Medium
2024
Turbocharging & downsizing
HD Pickup & Van Engines
Medium
2018
"Right sized" diesel engine
HD Pickup & Van vehicles, Tractors
Low
2022
6 speed transmission
HD Pickup & Van vehicles
Medium
2018
8 speed transmission
HD Pickup & Van vehicles, Vocational
Medium
2018
Automated & Automated manual transmission
(AMT)
Vocational, Tractors
Medium
2022
High efficiency gearbox (HEG)
Vocational, Tractors, HD Pickup &
Vans
Low
2022,2024
Early torque converter lockup (TORQ)
Vocational, HD Pickup & Vans
Low
2022,2018
Auto transmission, power-shift
Tractors
Medium
2022
Dual clutch transmission
Tractors
Medium
2022
Driveline integration
Vocational
Low
2022
6x2 axle
Tractors
Low
2022
Axle disconnect
Vocational
Low
2022
Axle downspeed
Tractors
Low
2022
High efficiency axle
Vocational, Tractors
Low
2022
Lower RR tires 1
HD Pickup & Van vehicles
Low
2018
Lower RR tires 2
HD Pickup & Van vehicles
Low
2024
Low drag brakes
HD Pickup & Van vehicles
Low
2018
Electric power steering
HD Pickup & Van vehicles
Low
2018
High efficiency transmission
HD Pickup & Van vehicles
Low
2024
Driveline friction reduction
HD Pickup & Van vehicles
Low
2022
Improved accessories (electrification)
HD Pickup & Van vehicles
Low
2018
Improved accessories (electrification)
Vocational, Tractors
Low
2022
Lower RR tires 1
Vocational, Tractors, Trailers
Low
2022
Lower RR tires 2
Vocational, Tractors, Trailers
Low
2022
Lower RR tires 3
Vocational, Tractors, Trailers s
Medium
2025
Lower RR tires 4
Vocational, Tractors, Trailers
Medium
2028
Lower RR tires 5
Vocational, Tractors, Trailers
Medium
2031
Automated Tire Inflation System (ATIS)
Tractors, Trailers
Low
2022
Tire Pressure Monitoring System
Vocational, Tractors & Trailers
Low
2022
Aero 1
HD Pickup & Van vehicles
Low
2018
Aero 2
HD Pickup & Van vehicles
Medium
2024
Aero Bins 1 thru 4
Tractors
Low
2022
Aero Bin 5 thru 7
Tractors
Medium
2025
Aero Bins 1 thru 8
Trailers
Low
2018
Weight reduction (via single wide tires and/or
aluminum wheels)
Tractors
Low
2022
Weight reduction via material changes
HD Pickup & Van vehicles
Low
2018
Weight reduction via material changes - 200
lbs, 400 lbs
Vocational
Low
2022
Weight reduction via material changes - 1000
lbs
Vocational
Medium
2022
Weight reduction via material changes
Tractors
Low
2022
Auxiliary power unit (APU), battery APU,
APU with DPF
Tractors
Low
2022

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Fuel operated heater (FOH)
Tractors
Low
2022
Air conditioning leakage
Vocational, Tractors
Low
2022
Air conditioning efficiency
Tractors
Low
2022
Neutral idle
Vocational
Low
2022
Stop-start (no regeneration)
HD Pickup & Van vehicles
Medium
2018
Stop-start (with enhancements)
Vocational
Medium
2022
Auto Engine Shutdown System
Vocational, Tractors
Low
2022
Mild hybrid
HD Pickup & Van vehicles
Highl
2024
Mild hybrid
Vocational
Highl
2025
Strong hybrid
HD Pickup & Van vehicles
Highl
2024
Hybrid without stop-start
Vocational
Highl
2022
Advanced cruise control
Tractors
Low
2022
There is some level of uncertainty surrounding both the ICM and RPE markup factors.
The ICM estimates used in this rule group all technologies into three broad categories and treat
them as if individual technologies within each of the three categories (low, medium, and high
complexity) would have the same ratio of indirect costs to direct costs. This simplification
means it is likely that the direct cost for some technologies within a category will be higher and
some lower than the estimate for the category in general. More importantly, the ICM estimates
have not been validated through a direct accounting of actual indirect costs for individual
technologies. RPEs themselves are inherently difficult to estimate because the accounting
statements of manufacturers do not neatly categorize all cost elements as either direct or indirect
costs. Hence, each researcher developing an RPE estimate must apply a certain amount of
judgment to the allocation of the costs. Moreover, RPEs for heavy- and medium-duty trucks and
for engine manufacturers are not as well studied as they are for the light-duty automobile
industry. Since empirical estimates of ICMs are ultimately derived from the same data used to
measure RPEs, this affects both measures. However, the value of RPE has not been measured
for specific technologies, or for groups of specific technologies. Thus, even if we assume that
the examined technology accurately represents the average impact on all technologies in its
representative category, applying a single average RPE to any given technology by definition
overstates costs for very simple technologies, or understates them for more advanced
technologies in that group.
2.11.1.3 Learning Effects on Technology Costs
For some of the technologies considered in this analysis, manufacturer learning effects
would be expected to play a role in the actual end costs. The "learning curve" or "experience
curve" describes the reduction in unit production costs as a function of accumulated production
volume. In theory, the cost behavior it describes applies to cumulative production volume
measured at the level of an individual manufacturer, although it is often assumed—as both
agencies have done in past regulatory analyses—to apply at the industry-wide level, particularly
in industries that utilize many common technologies and component supply sources. Both
agencies believe there are indeed many factors that cause costs to decrease over time. Research
in the costs of manufacturing has consistently shown that, as manufacturers gain experience in
production, they are able to apply innovations to simplify machining and assembly operations,
use lower cost materials, and reduce the number or complexity of component parts. All of these
factors allow manufacturers to lower the per-unit cost of production (i.e., the manufacturing
learning curve).202

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The agencies have a detailed description of the learning effect in the light-duty 2012-
2016 rulemaking. Most studies of the effect of experience or learning on production costs appear
to assume that cost reductions begin only after some initial volume threshold has been reached,
but not all of these studies specify this threshold volume. The rate at which costs decline beyond
the initial threshold is usually expressed as the percent reduction in average unit cost that results
from each successive doubling of cumulative production volume, sometimes referred to as the
learning rate. Many estimates of experience curves do not specify a cumulative production
volume beyond which cost reductions would no longer occur, instead depending on the
asymptotic behavior of the effect for learning rates below 100 percent to establish a floor on
costs.
In past rulemaking analyses, as noted above, both agencies have used a learning curve
algorithm that applied a learning factor of 20 percent for each doubling of production volume.
NHTSA has used this approach in analyses supporting recent CAFE rules. In its analyses, EPA
has simplified the approach by using an "every two years" based learning progression rather than
a pure production volume progression {i.e., after two years of production it was assumed that
production volumes would have doubled and, therefore, costs would be reduced by 20 percent).
In the light-duty 2012-2016 rulemaking, the agencies employed an additional learning
algorithm to reflect the volume-based learning cost reductions that occur further along on the
learning curve. This additional learning algorithm was termed "time-based" learning simply as a
means of distinguishing this algorithm from the volume-based algorithm mentioned above,
although both of the algorithms reflect the volume-based learning curve supported in the
literature.203 To avoid confusion, we now refer to this learning algorithm as the "flat-portion" of
the learning curve. This way, we maintain the clarity that all learning is, in fact, volume-based
learning, and the level of cost reductions depend only on where on the learning curve a
technology's learning progression is. We distinguish the flat-portion of the curve from the steep-
portion of the curve to indicate the level of learning taking place in the years following
implementation of the technology. The agencies have applied the steep-portion learning
algorithm for those technologies considered to be newer technologies likely to experience rapid
cost reductions through manufacturer learning and the flat-portion learning algorithm for those
technologies considered to be mature technologies likely to experience minor cost reductions
through manufacturer learning. As noted above, the steep-portion learning algorithm results in
20 percent lower costs after two full years of implementation {i.e., the 2016 MY costs are 20
percent lower than the 2014 and 2015 model year costs). Once the steep-portion learning steps
have occurred (for technologies having the steep-portion learning algorithm applied), flat-portion
learning at 3 percent per year becomes effective for 5 years. For technologies having the flat-
portion learning algorithm applied), flat-portion learning at 3 percent per year begins in year 2
and remains effective for 5 years. Beyond 5 years of learning at 3 percent per year, 5 years of
learning at 2 percent per year, then 5 at 1 percent per year become effective. There was no
serious disagreement with this approach in the public comments to any of the GHG/fuel
economy/consumption rulemakings.
Learning effects are applied to most but not all technologies because some of the
expected technologies are already used rather widely in the industry and, presumably, learning
impacts have already occurred. The steep-portion learning algorithm was applied for only a
handful of technologies that are considered to be new or emerging technologies. Most

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
technologies have been considered to be more established given their current use in the fleet and,
hence, the lower flat-portion learning algorithm has been applied. The learning algorithms
applied to each technology are summarized in Table 2-121. One change has been made since the
proposal to waste heat recovery which used learning algorithm 12 in the proposal but uses a new
learning algorithm 14 in this final rule.
Table 2-121 Learning Effect Algorithms Applied to Technologies Used in this Analysis
TECHNOLOGY
APPLIED TO
LEARNING
ALGORITHM
LEARNING
FACTOR
"CURVE" A
Cylinder head improvements 1
LH/MH/HH Engines
Flat
2
Cylinder head improvements 2
LH/MH/HH Engines
Flat
13
Turbo efficiency improvements 1
LH/MH/HH, HD Pickup
& Van Engines
Flat
2
Turbo efficiency improvements 2
LH/MH/HH Engines
Flat
13
EGR cooler efficiency improvements 1
LH/MH/HH Engines
Flat
2
EGR cooler efficiency improvements 2
LH/MH/HH Engines
Flat
13
Water pump improvements 1
LH/MH/HH Engines
Flat
2
Water pump improvements 2
LH/MH/HH Engines
Flat
13
Oil pump improvements 1
LH/MH/HH Engines
Flat
2
Oil pump improvements 2
LH/MH/HH Engines
Flat
13
Fuel pump improvements 1
LH/MH/HH Engines
Flat
2
Fuel pump improvements 2
LH/MH/HH Engines
Flat
13
Fuel rail improvements 1
LH/MH/HH Engines
Flat
2
Fuel rail improvements 2
LH/MH/HH Engines
Flat
13
Fuel injector improvements 1
LH/MH/HH Engines
Flat
2
Fuel injector improvements 2
LH/MH/HH Engines
Flat
13
Piston improvements 1
LH/MH/HH Engines
Flat
2
Piston improvements 2
LH/MH/HH Engines
Flat
13
Valve train friction reductions 1
LH/MH/HH Engines
Flat
2
Valve train friction reductions 2
LH/MH/HH Engines
Flat
13
Turbo compounding 1
LH/MH/HH Engines
Flat
2
Turbo compounding 2
LH/MH/HH Engines
Flat
13
Aftertreatment improvements 1 & 2
LH/MH/HH Engines
Flat
2
Model based control
LH/MH/HH Engines
Flat
13
Waste heat recovery
HH Engines
Steep
14
Engine friction reduction 1 & 2
HD Pickup & Van
Engines
None
1
Engine changes to accommodate low
friction lubes
HD Pickup & Van
Engines
None
1
Variable valve timing
HD Pickup & Van
Engines
Flat
8
Stoichiometric gasoline direct injection
HD Pickup & Van
Engines
Flat
7
Cylinder deactivation
HD Pickup & Van
Engines
Flat
8
Cooled EGR
HD Pickup & Van
Engines
Flat
7
Turbocharging & downsizing
HD Pickup & Van
Engines
Flat
7
"Right sized" diesel engine
HD Pickup & Van
vehicles, Tractors
None
1

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
6 speed transmission
HD Pickup & Van
vehicles
Flat
7
8 speed transmission
HD Pickup & Van
vehicles, Vocational
Flat
7
Automated & Automated manual
transmission (AMT)
Vocational, Tractors
Flat
12
High efficiency gearbox (HEG)
Vocational, Tractors, HD
Pickup & Vans
Flat
13,6
Early torque converter lockup (TORQ)
Vocational, HD Pickup &
Vans
Flat
13, 8
Auto transmission, power-shift
Tractors
Flat
12
Dual clutch transmission
Tractors
Flat
12
Driveline integration
Vocational
Flat
13
6x2 axle
Tractors
Flat
12
Axle disconnect
Vocational
None
1
Axle downspeed
Tractors
Flat
12
High efficiency axle
Vocational, Tractors
Flat
12
Lower RR tires 1
HD Pickup & Van
vehicles
None
1
Lower RR tires 2
HD Pickup & Van
vehicles
Steep
11
Low drag brakes
HD Pickup & Van
vehicles
None
1
Electric power steering
HD Pickup & Van
vehicles
Flat
8
High efficiency transmission
HD Pickup & Van
vehicles
Flat
6
Driveline friction reduction
HD Pickup & Van
vehicles
Flat
3
Improved accessories (electrification)
HD Pickup & Van
vehicles
Flat
8
Improved accessories
Tractors
Flat
12
Improved fan
Tractors
Flat
12
Lower RR tires 1
Vocational, Tractors,
Trailers
Flat
2
Lower RR tires 2
Vocational, Tractors,
Trailers
Flat
2
Lower RR tires 3
Vocational, Tractors,
Trailers
Flat
12
Lower RR tires 4
Vocational, Tractors,
Trailers
Flat
13
Lower RR tires 5
Vocational, Tractors,
Trailers

13
Automated Tire Inflation System (ATIS)
Tractors, Trailers
Flat
12
Tire Pressure Monitoring System (TPMS)
Vocational, Tractors,
Trailers
Flat
12
Aero 1 & 2
HD Pickup & Van
vehicles
Flat
8
Aero Bins 1 & 2
Tractors
None
1
Aero Bin 3
Tractors
Flat
2
Aero Bins 4 thru 7
Tractors
Steep
4
Aero Bins 1 thru 8
Trailers
Flat
2
Weight reduction (via single wide tires
and/or aluminum wheels)
Tractors
Flat
2

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Weight reduction via material changes
HD Pickup & Van
vehicles
Flat
6
Weight reduction via material changes
Vocational, Tractors
Flat
13
Auxiliary power unit (APU), battery
APU, APU with DPF
Tractors
Flat
2
Fuel operated heater (FOH)
Tractors
Flat
2
Air conditioning leakage
Vocational, Tractors
Flat
2
Air conditioning efficiency
Tractors
Flat
12
Neutral idle
Vocational
None
1
Stop-start (no regeneration)
HD Pickup & Van
vehicles
Steep
9
Stop-start (with enhancements)
Vocational
Flat
13
Mild hybrid
HD Pickup & Van
vehicles
Flat
6
Mild hybrid
Tractors
Flat
12
Strong hybrid
HD Pickup & Van
vehicles
Steep
11
Hybrid without stop-start
Vocational
Steep
11
Advanced cruise control
Tractors
Flat
12
Note:
" See table and figure below.
The actual year-by-year factors for the numbered curves shown in Table 2-121 are shown
in Table 2-122 and are shown graphically in Figure 2-84.
Table 2-122 Year-by-year Learning Curve Factors for the Learning Curves Used in this Analysis
CURVEA
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
1
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
2
1.000
0.970
0.941
0.913
0.885
0.868
0.850
0.833
0.817
0.800
0.784
0.769
0.761
0.753
3
1.031
1.000
0.970
0.941
0.913
0.894
0.877
0.859
0.842
0.825
0.808
0.792
0.784
0.777
4
1.000
1.000
0.800
0.800
0.640
0.621
0.602
0.584
0.567
0.550
0.533
0.517
0.507
0.497
6
1.096
1.063
1.031
1.000
0.970
0.941
0.913
0.885
0.859
0.842
0.825
0.808
0.792
0.776
7
0.941
0.913
0.885
0.868
0.850
0.833
0.817
0.800
0.784
0.769
0.753
0.738
0.731
0.723
8
1.031
1.000
0.970
0.951
0.932
0.913
0.895
0.877
0.859
0.842
0.825
0.809
0.801
0.793
9
1.250
1.000
1.000
0.970
0.941
0.913
0.885
0.859
0.833
0.808
0.784
0.760
0.745
0.730
11
1.563
1.563
1.563
1.563
1.563
1.250
1.250
1.000
0.970
0.941
0.913
0.885
0.859
0.842
12
1.130
1.096
1.063
1.031
1.000
0.970
0.941
0.913
0.894
0.877
0.859
0.842
0.825
0.808
13
1.238
1.201
1.165
1.130
1.096
1.063
1.031
1.000
0.970
0.941
0.913
0.894
0.877
0.859
14
1.563
1.563
1.563
1.563
1.563
1.250
1.250
1.000
1.000
0.800
0.800
0.640
0.621
0.602
Note:
" Curves 5 and 10 were generated but subsequently not used so are not included in the table.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
1.60
1.40
1.20
1.00
0.80
0.60
0.40
2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025
Figure 2-84 Year-by-year Learning Curve Factors for the Learning Curves used in this Analysis
Importantly, where the factors shown in Table 2-122 and, therefore, the curves shown in
Figure 2-84 equal "1.00" represents the year for which any particular technology's cost is based.
In other words, for example, the cost estimate that we have for cylinder head improvements 2 is
"based" in 2021 (curve 13). Therefore, its learning factor equals 1.00 in 2021 and then decreases
going forward to represent lower costs due to learning effects. Its learning factors are greater
than 1.00 in years before 2021 to represent "reverse" learning, i.e., higher costs than our 2021
estimate since production volumes have, presumably, not yet reached the point where our cost
estimate can be considered valid.
2.11.1.4 Technology Penetration Rates and Package Costs
Determining the stringency of the standards involves a balancing of relevant factors -
chiefly technology feasibility and effectiveness, costs, and lead time. For each of the standards,
the agencies have projected a technology path to achieve the standards reflecting an application
rate of those technologies the agencies consider to be available at reasonable cost in the lead
times provided. The agencies do not expect each of the technologies for which costs have been
developed to be employed by all engines and vehicles across the board. Further, many of today's
vehicles are already equipped with some of the technologies and/or are expected to adopt them
by MY2018 to comply with the HD Phase 1 standards. Estimated penetration rates in both the
reference and control cases are necessary for each vehicle category. The penetration rates for

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
many technologies are zero in the reference case; however, for some technologies—notably aero
and tire technologies—the reference case penetration rate is not always zero. These reference
and control case penetration rates are then applied to the technology costs with the result being a
package cost for each vehicle category. As such, package costs are rarely if ever a simple sum of
all the technology costs since each technology would be expected to be adopted at different rates.
For HD pickups and vans, the CAFE model predicts the technology penetration rates that
most cost effectively meet the standards being adopted. Similar to vocational vehicles, tractors
and trailers, package costs are rarely if ever a simple sum of all the technology costs since each
technology would be expected to be adopted at different rates. The methods for estimating
technology penetration rates and resultant costs (and other impacts) for HD pickups and vans are
discussed in Chapter 10 of this RIA.
2.11.1.5 Conversion of Technology Costs to 2013 U.S. Dollars
As noted above in Section IX C. 1, the agencies are using technology costs from many
different sources. These sources, having been published in different years, present costs in
different year dollars (i.e., 2009 dollars or 2010 dollars). For this analysis, the agencies sought to
have all costs in terms of 2013 dollars to be consistent with the dollars used by AEO in its 2015
Annual Energy Outlook.204 While the factors used to convert from 2009 dollars (or other) to
2013 dollars are small, the agencies prefer to be overly diligent in this regard to ensure
consistency across our benefit-cost analysis. The agencies have used the GDP Implicit Price
Deflator for Gross Domestic Product as the converter, with the actual factors used as shown in
Table 2-123.205
Table 2-123 Implicit Price Deflators and Conversion Factors for Conversion to 2013$
CALENDAR YEAR
2005
2006
2007
2008
2009
2010
2011
2012
2013
Price index for GDP
91.988
94.814
97.337
99.246
100
101.221
103.311
105.214
106.929
Factor applied for
2013$
1.162
1.128
1.099
1.077
1.069
1.056
1.035
1.016
1.000
The sections above describe the technologies expected to be used to enable compliance
with the standards and the penetration rates we estimate to be possible. Here we present the cost
of each technology, the markups used for each, the learning effect applied, etc. The tables here
present the direct manufacturing cost (DMC) we have estimated for each technology, the indirect
costs (IC) associated with that technology, and the resultant total cost (TC) of each (where
TC=DMC+IC). Each table also presents, where appropriate, the expected adoption rate of each
technology in both the reference case (i.e., alternative la or the "no new controls" case) and the
policy case (the standards). For most technologies, the reference case adoption rate will be
shown as 0 percent (or blanks in the tables) since the Phase 2 technologies are expected to be in
limited or no use in the regulatory timeframe. However, for some technologies—notably tire and
aero technologies—there is expected to considerably adoption of Phase 2 technologies in the
reference case. The final row(s) of the tables shown here include the penetration rates applied to
the technology costs to arrive at a total cost of each technology as it is applied to the ultimate
package (noted as TCp). In Chapter 2.12 of this RIA, we sum these costs (the TCp costs) into
total cost applied to the packages presented later in Chapter 7 of this RIA. We also describe how
we moved from the total cost applied to the packages developed for the regulatory classes (i.e.,

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Class 8 Sleeper cab, LH vocational medium-speed, etc.) to the MOVES sourcetypes (i.e., transit
bus, refuse truck, combination long haul, etc.) in order to develop program costs. This final
step—moving from regulatory classes to MOVES sourcetypes, was necessary because MOVES
populations, sales, inventory calculations, etc., are based on sourcetypes, not regulatory classes,
and to allow for a more granular look at payback as presented in Chapter 7.2.4 of this RIA.
Note that the text surrounding the tables presented here refer to low/medium/high
complexity ICMs and to learning curves used. We discuss both the ICMs and the learning
effects used in this analysis in Chapter 2.11.1.2 and 2.11.1.3 of this RIA, respectively.
We received some comments on our technology costs, both direct and indirect costs, and
on learning impacts. We address those comments in Section 11.3 of the Response to Comments
document.
2.11.2 Costs of Engine Technologies
2.11.2.1 Aftertreatment Improvements
We have estimated the cost of aftertreatment improvements based on the aftertreatment
improvements technology discussed in the Phase 1 rules. That technology was estimated at $25
(DMC, 2008$, in 2014) for each percentage improvement in fuel consumption, or $100 (DMC,
2008$, in 2014) for the 4 percent improvement expected as a result of that program. In Phase 2,
we are expecting only a 0.6 percent improvement in fuel consumption resulting from
aftertreatment improvements. Therefore, the cost in Phase 2 including updates to 2013$ is $16
(DMC, 2013$, in 2014). We consider this technology to be on the flat portion of the learning
curve (curve 2) and have applied a low complexity ICM with short term markups through 2024.
The resultant technology costs, penetration rates and total cost applied to the package are shown
below.
Table 2-124 Costs of Aftertreatment Improvements - Level 2
Light/Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Aftertreatment
improvements - level 2
DMC
$14
$14
$14
$13
$13
$13
$13
$12
$12
$12
Aftertreatment
improvements - level 2
IC
$2
$2
$2
$2
$2
$2
$2
$2
$2
$2
Aftertreatment
improvements - level 2
TC
$17
$17
$16
$16
$16
$15
$15
$15
$15
$15
Aftertreatment
improvements - level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aftertreatment
improvements - level 2
Alt 3
0%
0%
0%
50%
50%
50%
90%
90%
90%
100%
Aftertreatment
improvements - level 2
TCp
$0
$0
$0
$8
$8
$8
$14
$13
$13
$15
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-125 Costs of Aftertreatment Improvements - Level 2
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Aftertreatment
improvements - level 2
DMC
$14
$14
$14
$13
$13
$13
$13
$12
$12
$12
Aftertreatment
improvements - level 2
IC
$2
$2
$2
$2
$2
$2
$2
$2
$2
$2
Aftertreatment
improvements - level 2
TC
$17
$17
$16
$16
$16
$15
$15
$15
$15
$15
Aftertreatment
improvements - level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aftertreatment
improvements - level 2
Alt 3
0%
0%
0%
45%
45%
45%
95%
95%
95%
100%
Aftertreatment
improvements - level 2
TCp
$0
$0
$0
$7
$7
$7
$14
$14
$14
$15
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.2 Cylinder Head Improvements
We have estimated the cost of cylinder head improvements based on the cylinder head
improvements technology discussed in the Phase 1 rules. That technology was estimated at $9
(DMC, 2008$, in 2014) for light HDD engines and at $5 (DMC, 2008$, in 2014) for medium and
heavy HDD engines. In Phase 2, we are estimating equivalent costs for an additional level of
cylinder head improvements. With updates to 2013$, we estimate the costs at $10 (DMC,
2013$, in 2021) for light HDD engines and at $6 (DMC, 2013$, in 2021) for medium and heavy
HDD engines. We consider this technology to be on the flat portion of the learning curve (curve
13) and have applied a low complexity ICM with short term markups through 2027. The
resultant technology costs, penetration rates and total cost applied to the package are shown
below.
Table 2-126 Costs for Cylinder Head Improvements - Level 2
Light HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Cylinder head
improvements - level 2
DMC
$11
$11
$10
$10
$10
$10
$9
$9
$9
$9
Cylinder head
improvements - level 2
IC
$2
$2
$2
$2
$2
$2
$2
$2
$2
$2
Cylinder head
improvements - level 2
TC
$13
$12
$12
$12
$11
$11
$11
$11
$10
$10
Cylinder head
improvements - level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Cylinder head
improvements - level 2
Alt 3
0%
0%
0%
50%
50%
50%
90%
90%
90%
100%
Cylinder head
improvements - level 2
TCp
$0
$0
$0
$6
$6
$6
$10
$10
$9
$10
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-127 Costs for Cylinder Head Improvements - Level 2
Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Cylinder head
improvements - level 2
DMC
$6
$6
$6
$6
$6
$6
$5
$5
$5
$5
Cylinder head
improvements - level 2
IC
$1
$1
$1
$1
$1
$1
$1
$1
$1
$1
Cylinder head
improvements - level 2
TC
$7
$7
$7
$7
$7
$6
$6
$6
$6
$6
Cylinder head
improvements - level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Cylinder head
improvements - level 2
Alt 3
0%
0%
0%
50%
50%
50%
90%
90%
90%
100%
Cylinder head
improvements - level 2
TCp
$0
$0
$0
$3
$3
$3
$6
$6
$5
$6
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-128 Costs for Cylinder Head Improvements - Level 2
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Cylinder head
improvements - level 2
DMC
$6
$6
$6
$6
$6
$6
$5
$5
$5
$5
Cylinder head
improvements - level 2
IC
$1
$1
$1
$1
$1
$1
$1
$1
$1
$1
Cylinder head
improvements - level 2
TC
$7
$7
$7
$7
$7
$6
$6
$6
$6
$6
Cylinder head
improvements - level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Cylinder head
improvements - level 2
Alt 3
0%
0%
0%
45%
45%
45%
95%
95%
95%
100%
Cylinder head
improvements - level 2
TCp
$0
$0
$0
$3
$3
$3
$6
$6
$6
$6
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.3 Turbocharger Efficiency Improvements
We have estimated the cost of turbo efficiency improvements based on the turbo
efficiency improvements technology discussed in the Phase 1 rules. That technology was
estimated at $16 (DMC, 2008$, in 2014) for all HDD engines. In Phase 2, we are estimating
equivalent costs for an additional level of turbo efficiency improvements. With updates to
2013$, we estimate the costs at $17 (DMC, 2013$, in 2021) for all HDD engines. We consider
this technology to be on the flat portion of the learning curve (curve 13) and have applied a low
complexity ICM with short term markups through 2027. The resultant technology costs,
penetration rates and total cost applied to the package are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-129 Costs for Turbocharger Efficiency Improvements - Level 2
Light/Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Turbo efficiency
improvements - level 2
DMC
$18
$18
$17
$17
$16
$16
$15
$15
$15
$14
Turbo efficiency
improvements - level 2
IC
$3
$3
$3
$3
$3
$3
$3
$3
$3
$3
Turbo efficiency
improvements - level 2
TC
$21
$21
$20
$19
$19
$18
$18
$18
$17
$17
Turbo efficiency
improvements - level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Turbo efficiency
improvements - level 2
Alt 3
0%
0%
0%
50%
50%
50%
90%
90%
90%
100%
Turbo efficiency
improvements - level 2
TCp
$0
$0
$0
$10
$9
$9
$16
$16
$16
$17
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-130 Costs for Turbocharger Efficiency Improvements - Level 2
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Turbo efficiency
improvements - level 2
DMC
$18
$18
$17
$17
$16
$16
$15
$15
$15
$14
Turbo efficiency
improvements - level 2
IC
$3
$3
$3
$3
$3
$3
$3
$3
$3
$3
Turbo efficiency
improvements - level 2
TC
$21
$21
$20
$19
$19
$18
$18
$18
$17
$17
Turbo efficiency
improvements - level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Turbo efficiency
improvements - level 2
Alt 3
0%
0%
0%
45%
45%
45%
95%
95%
95%
100%
Turbo efficiency
improvements - level 2
TCp
$0
$0
$0
$9
$9
$8
$17
$17
$16
$17
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
For HD diesel pickups and vans, we are estimating use of the Phase 1 level of turbo
efficiency improvements, or $17 (DMC, 2012$, in 2014). We consider this technology to be on
the flat portion of the learning curve (curve 2) and have applied a low complexity ICM with
short term markups through 2022. The resultant technology costs are shown below.
Table 2-131 Costs for Turbocharger Efficiency Improvements - Level 1
HD Pickups & Vans (2012$)
TECHNOLOGY

2021
2022
2023
2024
2025
2026
2027
Turbo efficiency improvements - level 1
DMC
$14
$14
$13
$13
$13
$13
$12
Turbo efficiency improvements - level 1
IC
$3
$3
$2
$2
$2
$2
$2
Turbo efficiency improvements - level 1
TC
$16
$16
$15
$15
$15
$15
$15
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.11.2.4 Turbo Compounding
We have estimated the cost of turbo compounding based on the turbo compounding
technology discussed in the Phase 1 rules. That technology was estimated at $813 (DMC,
2008$, in 2014) for all HDD tractor engines. In Phase 2, we are estimating equivalent costs for
an additional level of turbo compounding improvements. With updates to 2013$, we estimate
the costs at $875 (DMC, 2013$, in 2021) for all HDD tractor engines. We consider this
technology to be on the flat portion of the learning curve (curve 13) and have applied a low
complexity ICM with short term markups through 2027. The resultant technology costs,
penetration rates and total cost applied to the package are shown below.
Table 2-132 Costs for Turbocharger Compounding - Level 2
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Turbo compounding
- level 2
DMC
$959
$930
$902
$875
$849
$824
$799
$783
$767
$752
Turbo compounding
- level 2
IC
$136
$136
$136
$136
$135
$135
$135
$135
$135
$135
Turbo compounding
- level 2
TC
$1,095
$1,066
$1,038
$1,011
$985
$959
$934
$918
$902
$887
Turbo compounding
- level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Turbo compounding
- level 2
Alt 3
0%
0%
0%
5%
5%
5%
10%
10%
10%
10%
Turbo compounding
- level 2
TCp
$0
$0
$0
$51
$49
$48
$93
$92
$90
$89
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.5 Valve Actuation
We have estimated the cost of valve actuation based on the dual cam phasing cost
estimate used in the 2017-2025 light-duty rule. In that analysis, we estimated costs at $151
(DMC, 2010$, in 2015) for a large V8 engine. In this HD Phase 2 program, we are estimating
equivalent costs for this technology. With updates to 2013$, we estimate the costs at $160
(DMC, 2013$, in 2015) for all HDD engines. We consider this technology to be on the flat
portion of the learning curve (curve 8) and have applied a medium complexity ICM with short
term markups through 2018. The resultant technology costs, penetration rates and total cost
applied to the package are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-133 Costs for Valve Actuation
Light/Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Valve actuation
DMC
$149
$146
$143
$140
$137
$135
$132
$129
$128
$127
Valve actuation
IC
$61
$46
$46
$46
$46
$46
$45
$45
$45
$45
Valve actuation
TC
$210
$192
$189
$186
$183
$180
$177
$175
$173
$172
Valve actuation
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Valve actuation
Alt 3
0%
0%
0%
50%
50%
50%
90%
90%
90%
100%
Valve actuation
All
$0
$0
$0
$93
$92
$90
$160
$157
$156
$172
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-134 Costs for Valve Actuation
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Valve actuation
DMC
$149
$146
$143
$140
$137
$135
$132
$129
$128
$127
Valve actuation
IC
$61
$46
$46
$46
$46
$46
$45
$45
$45
$45
Valve actuation
TC
$210
$192
$189
$186
$183
$180
$177
$175
$173
$172
Valve actuation
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Valve actuation
Alt 3
0%
0%
0%
45%
45%
45%
95%
95%
95%
100%
Valve actuation
All
$0
$0
$0
$84
$82
$81
$169
$166
$165
$172
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
For HD pickups and vans, we have estimated the costs of dual cam phasing based on the
DMC, IC and TC presented above in Table 2-133.
For discrete variable valve lift (DVVL), we have again used the 2017-2025 light-duty
FRM values updated to 2012$ to arrive at a cost of $259 (DMC, 2012$, in 2015). We consider
this technology to be on the flat portion of the learning curve (curve 8) and have applied medium
complexity markups with short term markups through 2024. The resultant costs are presented
below.
Table 2-135 Costs for Discrete Variable Valve Lift (DWL)
Gasoline HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Discrete variable
valve lift (DWL)
DMC
$227
$223
$218
$214
$210
$207
$205
Discrete variable
valve lift (DWL)
IC
$74
$74
$74
$74
$74
$73
$73
Discrete variable
valve lift (DWL)
TC
$301
$297
$292
$288
$283
$281
$279
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
2.11.2.6 EGR
We have estimated the cost of EGR cooler improvements based on the EGR cooler
improvements technology discussed in the Phase 1 rules. That technology was estimated at $3

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
(DMC, 2008$, in 2014) for all HDD engines. In Phase 2, we are estimating equivalent costs for
an additional level of EGR cooler improvements. With updates to 2013$, we estimate the costs
at $3 (DMC, 2013$, in 2021) for all HDD engines. We consider this technology to be on the flat
portion of the learning curve (curve 13) and have applied a low complexity ICM with short term
markups through 2027. The resultant technology costs, penetration rates and total cost applied to
the package are shown below.
Table 2-136 Costs for EGR Cooler Improvements - Level 2
Light/Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
EGR cooler - level 2
DMC
$4
$4
$3
$3
$3
$3
$3
$3
$3
$3
EGR cooler - level 2
IC
$1
$1
$1
$1
$1
$1
$1
$1
$1
$1
EGR cooler - level 2
TC
$4
$4
$4
$4
$4
$4
$4
$4
$3
$3
EGR cooler - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
EGR cooler - level 2
Alt 3
0%
0%
0%
50%
50%
50%
90%
90%
90%
100%
EGR cooler - level 2
TCp
$0
$0
$0
$2
$2
$2
$3
$3
$3
$3
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-137 Costs for EGR Cooler Improvements - Level 2
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
EGR cooler - level 2
DMC
$4
$4
$3
$3
$3
$3
$3
$3
$3
$3
EGR cooler - level 2
IC
$1
$1
$1
$1
$1
$1
$1
$1
$1
$1
EGR cooler - level 2
TC
$4
$4
$4
$4
$4
$4
$4
$4
$3
$3
EGR cooler - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
EGR cooler - level 2
Alt 3
0%
0%
0%
45%
45%
45%
95%
95%
95%
100%
EGR cooler - level 2
TCp
$0
$0
$0
$2
$2
$2
$3
$3
$3
$3
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
For HD pickups and vans, we have estimated the costs of adding cooled EGR to a
gasoline engine based on the values used in the 2017-2025 light-duty FRM. We have scaled
upward the light-duty value by 25 percent and converted to 2012$ to arrive at a cost of $317
(DMC, 2012$, in 2012). We consider this technology to be on the flat portion of the learning
curve (curve 7) and have applied medium complexity markups with near term markups through
2024. The resultant costs are presented below.
Table 2-138 Costs for Cooled EGR
Gasoline HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Cooled EGR
DMC
$253
$248
$243
$239
$234
$231
$229
Cooled EGR
IC
$120
$120
$119
$119
$89
$89
$89
Cooled EGR
TC
$373
$368
$363
$358
$323
$321
$318
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.11.2.7 Water Pump Improvements
We have estimated the cost of water pump improvements based on the water pump
improvements technology discussed in the Phase 1 rules. That technology was estimated at $78
(DMC, 2008$, in 2014) for all HDD engines. In Phase 2, we are estimating equivalent costs for
an additional level of water pump improvements. With updates to 2013$, we estimate the costs
at $84 (DMC, 2013$, in 2021) for all HDD engines. We consider this technology to be on the
flat portion of the learning curve (curve 13) and have applied a low complexity ICM with short
term markups through 2027. The resultant technology costs, penetration rates and total cost
applied to the package are shown below.
Table 2-139 Costs for Water Pump Improvements - Level 2
Light/Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Water pump - level 2
DMC
$92
$89
$87
$84
$82
$79
$77
$75
$74
$72
Water pump - level 2
IC
$13
$13
$13
$13
$13
$13
$13
$13
$13
$13
Water pump - level 2
TC
$105
$103
$100
$97
$95
$92
$90
$88
$87
$85
Water pump - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Water pump - level 2
Alt 3
0%
0%
0%
60%
60%
60%
90%
90%
90%
100%
Water pump - level 2
TCp
$0
$0
$0
$58
$57
$55
$81
$79
$78
$85
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-140 Costs for Water Pump Improvements - Level 2
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Water pump - level 2
DMC
$92
$89
$87
$84
$82
$79
$77
$75
$74
$72
Water pump - level 2
IC
$13
$13
$13
$13
$13
$13
$13
$13
$13
$13
Water pump - level 2
TC
$105
$103
$100
$97
$95
$92
$90
$88
$87
$85
Water pump - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Water pump - level 2
Alt 3
0%
0%
0%
45%
45%
45%
95%
95%
95%
100%
Water pump - level 2
TCp
$0
$0
$0
$44
$43
$41
$85
$84
$82
$85
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.8 Oil Pump Improvements
We have estimated the cost of oil pump improvements based on the oil pump
improvements technology discussed in the Phase 1 rules. That technology was estimated at just
under $4 (DMC, 2008$, in 2014) for all HDD engines. In Phase 2, we are estimating equivalent
costs for an additional level of oil pump improvements. With updates to 2013$, we estimate the
costs at just over $4 (DMC, 2013$, in 2021) for all HDD engines. We consider this technology
to be on the flat portion of the learning curve (curve 13) and have applied a low complexity ICM
with short term markups through 2027. The resultant technology costs, penetration rates and
total cost applied to the package are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-141 Costs for Oil Pump Improvements - Level 2
Light/Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Oil pump - level 2
DMC
$5
$4
$4
$4
$4
$4
$4
$4
$4
$4
Oil pump - level 2
IC
$1
$1
$1
$1
$1
$1
$1
$1
$1
$1
Oil pump - level 2
TC
$5
$5
$5
$5
$5
$5
$4
$4
$4
$4
Oil pump - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Oil pump - level 2
Alt 3
0%
0%
0%
60%
60%
60%
90%
90%
90%
100%
Oil pump - level 2
TCp
$0
$0
$0
$3
$3
$3
$4
$4
$4
$4
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-142 Costs for Oil Pump Improvements - Level 2
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Oil pump - level 2
DMC
$5
$4
$4
$4
$4
$4
$4
$4
$4
$4
Oil pump - level 2
IC
$1
$1
$1
$1
$1
$1
$1
$1
$1
$1
Oil pump - level 2
TC
$5
$5
$5
$5
$5
$5
$4
$4
$4
$4
Oil pump - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Oil pump - level 2
Alt 3
0%
0%
0%
45%
45%
45%
95%
95%
95%
100%
Oil pump - level 2
TCp
$0
$0
$0
$2
$2
$2
$4
$4
$4
$4
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.9 Fuel Pump Improvements
We have estimated the cost of fuel pump improvements based on the fuel pump
improvements technology discussed in the Phase 1 rules. That technology was estimated at just
under $4 (DMC, 2008$, in 2014) for all HDD engines. In Phase 2, we are estimating equivalent
costs for an additional level of fuel pump improvements. With updates to 2013$, we estimate the
costs at just over $4 (DMC, 2013$, in 2021) for all HDD engines. We consider this technology
to be on the flat portion of the learning curve (curve 13) and have applied a low complexity ICM
with short term markups through 2027. The resultant technology costs, penetration rates and
total cost applied to the package are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-143 Costs for Fuel Pump Improvements - Level 2
Light/Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Fuel pump - level 2
DMC
$5
$4
$4
$4
$4
$4
$4
$4
$4
$4
Fuel pump - level 2
IC
$1
$1
$1
$1
$1
$1
$1
$1
$1
$1
Fuel pump - level 2
TC
$5
$5
$5
$5
$5
$5
$4
$4
$4
$4
Fuel pump - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Fuel pump - level 2
Alt 3
0%
0%
0%
60%
60%
60%
90%
90%
90%
100%
Fuel pump - level 2
TCp
$0
$0
$0
$3
$3
$3
$4
$4
$4
$4
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-144 Costs for Fuel Pump Improvements - Level 2
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Fuel pump - level 2
DMC
$5
$4
$4
$4
$4
$4
$4
$4
$4
$4
Fuel pump - level 2
IC
$1
$1
$1
$1
$1
$1
$1
$1
$1
$1
Fuel pump - level 2
TC
$5
$5
$5
$5
$5
$5
$4
$4
$4
$4
Fuel pump - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Fuel pump - level 2
Alt 3
0%
0%
0%
45%
45%
45%
95%
95%
95%
100%
Fuel pump - level 2
TCp
$0
$0
$0
$2
$2
$2
$4
$4
$4
$4
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.10 Fuel Rail Improvements
We have estimated the cost of fuel rail improvements based on the fuel rail improvements
technology discussed in the Phase 1 rules. That technology was estimated at $10 (DMC, 2008$,
in 2014) for LHDD engines and just under $9 (DMC, 2008$, in 2014) for MHDD and HHDD
engines. In Phase 2, we are estimating equivalent costs for an additional level of fuel rail
improvements. With updates to 2013$, we estimate the costs at $11 (DMC, 2013$, in 2021) for
LHDD and at just over $9 (DMC, 2013$, in 2021) for MHDD and HHDD engines. We consider
this technology to be on the flat portion of the learning curve (curve 13) and have applied a low
complexity ICM with short term markups through 2027. The resultant technology costs,
penetration rates and total cost applied to the package are shown below.
Table 2-145 Costs for Fuel Rail Improvements - Level 2
Light HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Fuel rail - level 2
DMC
$12
$12
$11
$11
$11
$10
$10
$10
$10
$9
Fuel rail - level 2
IC
$2
$2
$2
$2
$2
$2
$2
$2
$2
$2
Fuel rail - level 2
TC
$14
$13
$13
$13
$12
$12
$12
$11
$11
$11
Fuel rail - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Fuel rail - level 2
Alt 3
0%
0%
0%
60%
60%
60%
90%
90%
90%
100%
Fuel rail - level 2
TCp
$0
$0
$0
$8
$7
$7
$11
$10
$10
$11
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-146 Costs for Fuel Rail Improvements - Level 2
Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Fuel rail - level 2
DMC
$10
$10
$10
$9
$9
$9
$8
$8
$8
$8
Fuel rail - level 2
IC
$1
$1
$1
$1
$1
$1
$1
$1
$1
$1
Fuel rail - level 2
TC
$12
$11
$11
$11
$10
$10
$10
$10
$10
$9
Fuel rail - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Fuel rail - level 2
Alt 3
0%
0%
0%
60%
60%
60%
90%
90%
90%
100%
Fuel rail - level 2
TCp
$0
$0
$0
$6
$6
$6
$9
$9
$9
$9
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-147 Costs for Fuel Rail Improvements - Level 2
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Fuel rail - level 2
DMC
$10
$10
$10
$9
$9
$9
$8
$8
$8
$8
Fuel rail - level 2
IC
$1
$1
$1
$1
$1
$1
$1
$1
$1
$1
Fuel rail - level 2
TC
$12
$11
$11
$11
$10
$10
$10
$10
$10
$9
Fuel rail - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Fuel rail - level 2
Alt 3
0%
0%
0%
45%
45%
45%
95%
95%
95%
100%
Fuel rail - level 2
TCp
$0
$0
$0
$5
$5
$5
$9
$9
$9
$9
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.11 Fuel Injector Improvements
We have estimated the cost of fuel injector improvements based on the fuel injector
improvements technology discussed in the Phase 1 rules. That technology was estimated at $13
(DMC, 2008$, in 2014) for LHDD engines and $9 (DMC, 2008$, in 2014) for MHDD and
HHDD engines. In Phase 2, we are estimating equivalent costs for an additional level of fuel
injector improvements. With updates to 2013$, we estimate the costs at $13 (DMC, 2012$, in
2021) for LHDD and at $10 (DMC, 2013$, in 2021) for MHDD and HHDD engines. We
consider this technology to be on the flat portion of the learning curve (curve 13) and have
applied a low complexity ICM with short term markups through 2027. The resultant technology
costs, penetration rates and total cost applied to the package are shown below.
Table 2-148 Costs for Fuel Injector Improvements - Level 2
Light HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Fuel injectors - level 2
DMC
$15
$14
$14
$13
$13
$13
$12
$12
$12
$12
Fuel injectors - level 2
IC
$2
$2
$2
$2
$2
$2
$2
$2
$2
$2
Fuel injectors - level 2
TC
$17
$16
$16
$16
$15
$15
$14
$14
$14
$14
Fuel injectors - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Fuel injectors - level 2
Alt 3
0%
0%
0%
50%
50%
50%
90%
90%
90%
100%
Fuel injectors - level 2
TCp
$0
$0
$0
$8
$8
$7
$13
$13
$12
$14
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-149 Costs for Fuel Injector Improvements - Level 2
Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Fuel injectors - level 2
DMC
$11
$11
$10
$10
$10
$10
$9
$9
$9
$9
Fuel injectors - level 2
IC
$2
$2
$2
$2
$2
$2
$2
$2
$2
$2
Fuel injectors - level 2
TC
$13
$12
$12
$12
$11
$11
$11
$11
$10
$10
Fuel injectors - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Fuel injectors - level 2
Alt 3
0%
0%
0%
50%
50%
50%
90%
90%
90%
100%
Fuel injectors - level 2
TCp
$0
$0
$0
$6
$6
$6
$10
$10
$9
$10
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-150 Costs for Fuel Injector Improvements - Level 2
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Fuel injectors - level 2
DMC
$11
$11
$10
$10
$10
$10
$9
$9
$9
$9
Fuel injectors - level 2
IC
$2
$2
$2
$2
$2
$2
$2
$2
$2
$2
Fuel injectors - level 2
TC
$13
$12
$12
$12
$11
$11
$11
$11
$10
$10
Fuel injectors - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Fuel injectors - level 2
Alt 3
0%
0%
0%
45%
45%
45%
95%
95%
95%
100%
Fuel injectors - level 2
TCp
$0
$0
$0
$5
$5
$5
$10
$10
$10
$10
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.12 Piston Improvements
We have estimated the cost of piston improvements based on the piston improvements
technology discussed in the Phase 1 rules. That technology was estimated at just over $2 (DMC,
2008$, in 2014) for all HDD engines. In Phase 2, we are estimating equivalent costs for an
additional level of fuel pump improvements. With updates to 2013$, we estimate the costs at $3
(DMC, 2013$, in 2021) for all HDD engines. We consider this technology to be on the flat
portion of the learning curve (curve 13) and have applied a low complexity ICM with short term
markups through 2027. The resultant technology costs, penetration rates and total cost applied to
the package are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-151 Costs for Piston Improvements - Level 2
Light/Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Piston improvements -
level 2
DMC
$3
$3
$3
$3
$2
$2
$2
$2
$2
$2
Piston improvements -
level 2
IC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Piston improvements -
level 2
TC
$3
$3
$3
$3
$3
$3
$3
$3
$3
$3
Piston improvements -
level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Piston improvements -
level 2
Alt 3
0%
0%
0%
50%
50%
50%
90%
90%
90%
100%
Piston improvements -
level 2
TCp
$0
$0
$0
$1
$1
$1
$2
$2
$2
$3
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-152 Costs for Piston Improvements - Level 2
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Piston improvements -
level 2
DMC
$3
$3
$3
$3
$2
$2
$2
$2
$2
$2
Piston improvements -
level 2
IC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Piston improvements -
level 2
TC
$3
$3
$3
$3
$3
$3
$3
$3
$3
$3
Piston improvements -
level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Piston improvements -
level 2
Alt 3
0%
0%
0%
45%
45%
45%
95%
95%
95%
100%
Piston improvements -
level 2
TCp
$0
$0
$0
$1
$1
$1
$3
$3
$2
$3
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.13 Valvetrain Friction Reduction
We have estimated the cost of valvetrain friction reduction based on the valvetrain
friction reduction technology discussed in the Phase 1 rules. That technology was estimated at
$94 (DMC, 2008$, in 2014) for LHDD engines and $70 (DMC, 2008$, in 2014) for MHDD and
HHDD engines. In Phase 2, we are estimating equivalent costs for an additional level of fuel
injector improvements. With updates to 2013$, we estimate the costs at $101 (DMC, 2013$, in
2021) for LHDD and at $76 (DMC, 2013$, in 2021) for MHDD and HHDD engines. We
consider this technology to be on the flat portion of the learning curve (curve 13) and have
applied a low complexity ICM with short term markups through 2027. The resultant technology
costs, penetration rates and total cost applied to the package are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-153 Costs for Valvetrain Friction Improvements - Level 2
Light HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Valvetrain friction
reduction - level 2
DMC
$111
$107
$104
$101
$98
$95
$92
$90
$89
$87
Valvetrain friction
reduction - level 2
IC
$16
$16
$16
$16
$16
$16
$16
$16
$16
$16
Valvetrain friction
reduction - level 2
TC
$126
$123
$120
$117
$114
$111
$108
$106
$104
$102
Valvetrain friction
reduction - level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Valvetrain friction
reduction - level 2
Alt 3
0%
0%
0%
60%
60%
60%
90%
90%
90%
100%
Valvetrain friction
reduction - level 2
TCp
$0
$0
$0
$70
$68
$66
$97
$95
$94
$102
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-154 Costs for Valvetrain Friction Improvements - Level 2
Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Valvetrain friction
reduction - level 2
DMC
$83
$81
$78
$76
$73
$71
$69
$68
$66
$65
Valvetrain friction
reduction - level 2
IC
$12
$12
$12
$12
$12
$12
$12
$12
$12
$12
Valvetrain friction
reduction - level 2
TC
$95
$92
$90
$87
$85
$83
$81
$79
$78
$77
Valvetrain friction
reduction - level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Valvetrain friction
reduction - level 2
Alt 3
0%
0%
0%
60%
60%
60%
90%
90%
90%
100%
Valvetrain friction
reduction - level 2
TCp
$0
$0
$0
$52
$51
$50
$73
$71
$70
$77
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-155 Costs for Valvetrain Friction Improvements - Level 2
HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Valvetrain friction
reduction - level 2
DMC
$83
$81
$78
$76
$73
$71
$69
$68
$66
$65
Valvetrain friction
reduction - level 2
IC
$12
$12
$12
$12
$12
$12
$12
$12
$12
$12
Valvetrain friction
reduction - level 2
TC
$95
$92
$90
$87
$85
$83
$81
$79
$78
$77
Valvetrain friction
reduction - level 2
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Valvetrain friction
reduction - level 2
Alt 3
0%
0%
0%
45%
45%
45%
95%
95%
95%
100%
Valvetrain friction
reduction - level 2
TCp
$0
$0
$0
$39
$38
$37
$77
$75
$74
$77
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.14 "Right-sized" Diesel Engine
We have estimated the cost of a slightly smaller diesel engine at a $500 savings (DMC,
2013$, in any year) for all HDD tractor engines. We believe this represents an opportunity for
lower costs because smaller diesel engines contain less materials and are, generally, less costly to
produce than a larger diesel engine. As this cost is considered applicable in any year, we have
not applied learning effects (curve 1). We have applied a low complexity ICM with short term
markups through 2022. The resultant technology costs, penetration rates and total cost applied to
the package are shown below. For HD pickups and vans, we estimated the right-sized diesel
engine cost as cost neutral to any reference case diesel engine and limited the technology to
diesel vans. We have not included any costs associated with lost utility of the smaller diesel
engine. We believe that the smaller engine would be attractive to some buyers, but not all, and
that those buyers would not be concerned by any possible lost utility. For that reason, we have
used a limited application rate for this technology. Note that, for HD pickups and vans, we have
considered this technology to be cost neutral.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-156 Costs for "Right-sized" HDD Tractor Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Right-sized
diesel engine
DMC
-$500
-$500
-$500
-$500
-$500
-$500
-$500
-$500
-$500
-$500
Right-sized
diesel engine
IC
$89
$89
$89
$89
$89
$89
$89
$89
$89
$89
Right-sized
diesel engine
TC
-$411
-$411
-$411
-$411
-$411
-$411
-$411
-$411
-$411
-$411
Right-sized
diesel engine
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Right-sized
diesel engine
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
30%
Right-sized
diesel engine
TCp
$0
$0
$0
-$41
-$41
-$41
-$82
-$82
-$82
-$123
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.15 Waste Heat Recovery
In the proposal, we estimated the cost of waste heat recovery based on the estimate from
Tetra Tech showing it at $12,000 (retail, 2013$). Using that $12,000 estimate and dividing by a
1.36 RPE (see Chapter 2.11.1.2 of this RIA) and converting to 2012$, we arrived at our
estimated DMC of $8,692 (DMC, 2012$, in 2018). For this final rule, we have updated our cost
of waste heat recovery based on new understanding of this technology. For this final rule, we
have chosen to start with one specific source considered by TetraTech in developing their cost
estimate. That source is the NESCCAF/ICCT/TIAX work which estimated the cost of the
technology at $15,100 having used an RPE of 2.0.206 Using the description of the technology by
NESCCAF, et al., TetraTech estimated the bill of materials (BOM) costs as shown below. Using
that BOM, along with updated understanding of more recent and future waste heat recovery
systems, EPA eliminated some of the items as unnecessary for the type of system and
effectiveness values that we envision (see Chapter 2.3 and 2.7 of this RIA). As shown in the
table below, EPA estimates the costs of waste heat recovery at $5463 (DMC, 2013$, in 2021)
and has considered this to be an applicable cost for MY2021.
Table 2-157 Direct Manufacturing Costs (DMC) for Waste Heat Recovery

MY2015 Cost


estimated by
EPA updates

TetraTech
(2013$)

(2009$)

Turbine generator & flywheel
$2160
$2309
Condenser
$550
$588
EGR boiler
$400
$428
Stack boiler
$1000
Not needed
Packaging, assembly, labor
$2000
$2138
Controls
$400
Not needed
Power electronics
$900
Not needed
Energy storage
$150
Not needed
Subtotal (direct mfg cost DMC)
$7560
$5463
RPE (2x subtotal)
$15120


-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
We consider this technology to be on the steep portion of the learning curve and have
generated a new learning curve in the final rule to accommodate this reworked cost estimate
(curve 14). We have applied a medium complexity ICM with short term markups through 2027.
The resultant technology costs, penetration rates and total cost applied to the package are
shown below.
Table 2-158 Costs for Waste Heat Recovery (WHR)
HDD Tractor Engines (2013$)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
WHR
DMC
$8,536
$6,829
$6,829
$5,463
$5,463
$4,370
$4,370
$3,496
$3,391
$3,290
WHR
IC
$1,807
$1,721
$1,721
$1,652
$1,652
$1,596
$1,596
$1,552
$1,547
$1,541
WHR
TC
$10,343
$8,550
$8,550
$7,115
$7,115
$5,967
$5,967
$5,048
$4,938
$4,831
WHR
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
WHR
Alt 3
0%
0%
0%
1%
1%
1%
5%
5%
5%
25%
WHR
TCp
$0
$0
$0
$71
$71
$60
$298
$252
$247
$1,208
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.16 Model-based Control
We have estimated the cost of model-based controls at $100 (DMC, 2013$, in 2021). We
consider this technology to be on the flat portion of the learning curve (curve 13) and have
applied a low complexity ICM with short term markups through 2022. The resultant technology
costs, penetration rates and total cost applied to the package are shown below.
Table 2-159 Costs for Model Based Controls
Light/Medium/Heavy HDD Vocational Engines (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Model-based control
DMC
$110
$106
$103
$100
$97
$94
$91
$89
$88
$86
Model-based control
IC
$16
$16
$15
$15
$15
$15
$15
$15
$15
$15
Model-based control
TC
$125
$122
$119
$115
$112
$110
$107
$105
$103
$101
Model-based control
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Model-based control
Alt 3
0%
0%
0%
25%
25%
25%
30%
30%
30%
40%
Model-based control
TCp
$0
$0
$0
$29
$28
$27
$32
$31
$31
$41
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.2.17 Engine Friction Reduction and Accommodating Low Friction
Lubes
We have based the costs for accommodating low friction lubes (LUB) on the costs used
in the light-duty 2017-2025 FRM but have scaled upward that cost by 50 percent to account for
the larger HD engines. Using that cost ($3 DMC, 2006$, in any year) and converting to 2012$
results in a cost of $5 (DMC, 2012$, in any year). We consider this technology to be beyond
learning (curve 1) and have applied low complexity markups with near term markups through
2018. The resultant costs for HD pickups and vans are shown in are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-160 Costs for Accommodating Low Friction Lubes
Gasoline & Diesel HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Engine friction reduction - level 1
DMC
$5
$5
$5
$5
$5
$5
$5
Engine friction reduction - level 1
IC
$1
$1
$1
$1
$1
$1
$1
Engine friction reduction - level 1
TC
$6
$6
$6
$6
$6
$6
$6
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
We have based the costs for engine friction reduction level 1 (EFR1) on the costs used in
the light-duty 2017-2025 FRM. That cost is based on an original estimate of $11/cylinder
(DMC, 2006$, in any year). Using that cost for an 8 cylinder engine and converting to 2012$
results in a cost of $97 (DMC, 2012$, in any year). We consider this technology to be beyond
learning (curve 1) and have applied low complexity markups with near term markups through
2018. The resultant costs for HD pickups and vans are shown in are shown below.
Table 2-161 Costs for Engine Friction Reduction - Level 1
Gasoline & Diesel HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Engine friction reduction - level 1
DMC
$97
$97
$97
$97
$97
$97
$97
Engine friction reduction - level 1
IC
$19
$19
$19
$19
$19
$19
$19
Engine friction reduction - level 1
TC
$116
$116
$116
$116
$116
$116
$116
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
For engine friction reduction level 2 (EFR2, which includes costs for accommodating low
friction lubes) we have used the same approach as used in the light-duty 2017-2025 rule in that
we have doubled the DMC associated with LUB and EFR1. As with those technologies, we
consider EFR2 to be beyond learning (curve 1) and have applied low complexity markups but
have applied near term markups through 2024. The resultant costs for gasoline HD pickups and
vans are shown below.
Table 2-162 Costs for Engine Friction Reduction - Level 2
Gasoline HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Engine friction reduction - level 2
DMC
$205
$205
$205
$205
$205
$205
$205
Engine friction reduction - level 2
IC
$50
$50
$50
$50
$39
$39
$39
Engine friction reduction - level 2
TC
$254
$254
$254
$254
$244
$244
$244
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
For diesel HD pickups and vans, we have used the above costs for EFR level 2 and added
to that costs associated with improvements to other parasitic loads on the engine. For that latter
portion of the cost, we have used the light HDD engine DMCs for improved water pump level 1,
improved oil pump level 1, improved fuel pump level 1, improved fuel injectors level 1 and
valvetrain friction reduction level 1, which together result in a cost of $193 (DMC, 2012$, in and
year). We consider this combined set of technologies to be beyond the effects of learning (curve
1) and have applied low complexity markups with near term markups through 2022. The
resultant costs for diesel HD pickups and vans are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-163 Costs for Engine Friction Reduction & Improvements to Other Parasitics
Diesel HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Engine friction reduction - diesel
DMC
$397
$397
$397
$397
$397
$397
$397
Engine friction reduction - diesel
IC
$96
$96
$87
$87
$77
$77
$77
Engine friction reduction - diesel
TC
$494
$494
$484
$484
$474
$474
$474
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
2.11.2.18 Cylinder Deactivation
For cylinder deactivation on HD pickups and vans, we have based the costs on values
presented in the light-duty 2017-2025 FRM with updates to 2012$ to arrive at a cost of $169
(DMC, 2012$, in 2015). We consider this technology to be on the flat portion of the learning
curve (curve 8) and have applied medium complexity markups with near term markups through
2018. The resultant costs are presented below.
Table 2-164 Costs for Cylinder Deactivation
Gasoline HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Cylinder deactivation
DMC
$148
$145
$142
$139
$137
$135
$134
Cylinder deactivation
IC
$48
$48
$48
$48
$48
$48
$48
Cylinder deactivation
TC
$196
$193
$190
$187
$185
$183
$182
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
2.11.2.19 Stoichiometric Gasoline Direct Injection (SGDI)
For gasoline direct injection on HD pickups and vans, we have based the costs on values
presented in the light-duty 2017-2025 FRM with updates to 2012$ to arrive at a cost of $417
(DMC, 2012$, in 2012). We consider this technology to be on the flat portion of the learning
curve (curve 7) and have applied medium complexity markups with near term markups through
2018. The resultant costs are presented below.
Table 2-165 Costs for Direct Injection
Gasoline HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Gasoline direct injection
DMC
$333
$327
$320
$314
$307
$304
$301
Gasoline direct injection
IC
$118
$118
$118
$117
$117
$117
$117
Gasoline direct injection
TC
$451
$445
$438
$431
$425
$422
$418
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost.
2.11.2.20 Turbocharging & Downsizing
For turbocharging and downsizing (TDS) on HD pickups and vans, we have based the
costs on values presented in the light-duty 2017-2025 FRM with updates to 2012$. For the twin
turbo configuration expected on a V6 engine (downsized from a V8), we estimate the cost at
$735 (DMC, 2012$, in 2012). We consider this technology to be on the flat portion of the
learning curve (curve 7) and have applied medium complexity markups with near term markups
through 2018. For downsizing from an overhead valve (OHV) V8 to an overhead cam (OHC)

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
V6 valvetrain, we have estimated the cost at $340 (DMC, 2012$, in 2017). We consider this
technology to be on the flat portion of the learning curve (curve 6) and have applied medium
complexity markups with near term markups through 2018. For downsizing from an OHC V8 to
an OHC V6, we have estimated the cost at -$295 (DMC, 2012$, in 2012). We consider this
technology to be on the flat portion of the learning curve to arrive at a cost of $417 (DMC,
2012$, in 2012). We consider this technology to be on the flat portion of the learning curve
(curve 7) and have applied medium complexity markups with near term markups through 2024.
The resultant costs for the turbocharging system and for downsizing from an OHV V8 to an
OHC V6 are shown below, and downsizing from an OHC V8 to an OHC V6 are also shown
below.
Table 2-166 Costs for Adding Twin Turbos
Gasoline HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Adding twin turbos
DMC
$588
$576
$565
$553
$542
$537
$531
Adding twin turbos
IC
$208
$208
$208
$207
$207
$207
$207
Adding twin turbos
TC
$796
$784
$772
$761
$749
$744
$738
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
Table 2-167 Costs for Downsizing from an OHV V8 to an OHC V6
Gasoline HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Downsizing from OHV
V8 to OHC V6
DMC
$301
$292
$286
$280
$275
$269
$264
Downsizing from OHV
V8 to OHC V6
IC
$97
$97
$97
$97
$96
$96
$96
Downsizing from OHV
V8 to OHC V6
TC
$398
$389
$383
$377
$371
$365
$360
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
Table 2-168 Costs for Downsizing from an OHC V8 to an OHC V6
Gasoline HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Downsizing from OHC
V8 to OHC V6
DMC
-$236
-$232
-$227
-$223
-$218
-$216
-$214
Downsizing from OHC
V8 to OHC V6
IC
$112
$112
$111
$111
$83
$83
$83
Downsizing from OHC
V8 to OHC V6
TC
-$125
-$120
-$116
-$111
-$135
-$133
-$131
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
2.11.3 Transmissions
2.11.3.1 Adding Additional Gears (Vocational)
We have estimated the cost of adding 2 additional gears for vocational vehicles
(light/medium HD, heavy HD urban/multipurpose) based on the light-duty cost for an 8 speed

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
automatic transmission relative to a 6 speed automatic of $78 (DMC, 2010$, in 2012).Q We
have scaled that value by typical torque values of 2000 foot-pounds for vocational and 332 for a
light-duty truck. With updates to 2013$, this DMC for vocational vehicles becomes $495
(DMC, 2013$, in 2012). We consider this technology to be on the flat portion of the learning
curve (curve 7) and have applied a medium complexity ICM with short term markups through
2018. The resultant technology costs, penetration rates and total cost applied to the package are
shown below.
Table 2-169 Costs for Adding 2 Gears to an Automatic Transmission
Vocational Light/Medium HD Urban/Multipurpose/Regional Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Adding additional gears
DMC
$421
$413
$404
$396
$388
$380
$373
$365
$362
$358
Adding additional gears
IC
$146
$109
$109
$108
$108
$108
$107
$107
$107
$107
Adding additional gears
TC
$567
$521
$513
$504
$496
$488
$480
$473
$469
$465
Adding additional gears
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Adding additional gears
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
20%
Adding additional gears
TCp
$0
$0
$0
$50
$50
$49
$96
$95
$94
$93
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-170 Costs for Adding 2 Gears to an Automatic Transmission
Vocational Heavy HD Urban/Multipurpose Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Adding additional gears
DMC
$421
$413
$404
$396
$388
$380
$373
$365
$362
$358
Adding additional gears
IC
$146
$109
$109
$108
$108
$108
$107
$107
$107
$107
Adding additional gears
TC
$567
$521
$513
$504
$496
$488
$480
$473
$469
$465
Adding additional gears
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Adding additional gears
Alt 3
0%
0%
0%
5%
5%
5%
10%
10%
10%
10%
Adding additional gears
TCp
$0
$0
$0
$25
$25
$24
$48
$47
$47
$47
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.3.2 Automated/Automated Manual Transmissions (AMT)
We have estimated the cost of an AMT transmission, relative to a manual transmission,
based on an estimate by Tetra Tech of $5,100 (retail, 2013$). Using that estimate, we divided by
an RPE of 1.36 to arrive at an estimated cost of $3750 (DMC, 2013$, in 2018). We consider this
technology to be on the flat portion of the learning curve (curve 12) and have applied a medium
complexity ICM with short term markups through 2022. The resultant technology costs,
penetration rates and total cost applied to the package are shown below.
Q This cost was updated by FEV in early 2013. We are using the updated cost here, not the value used in the light-
duty 2017-2025 final rule.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-171 Costs for an Automated Transmission
Vocational Heavy HD & Heavy HD Multipurpose Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Manual to AMT
DMC
$3,750
$3,638
$3,528
$3,423
$3,354
$3,287
$3,221
$3,157
$3,094
$3,032
Manual to AMT
IC
$1,134
$1,128
$1,123
$1,117
$1,114
$830
$828
$825
$823
$821
Manual to AMT
TC
$4,884
$4,766
$4,651
$4,540
$4,468
$4,117
$4,049
$3,982
$3,917
$3,853
Manual to AMT
Alt
la
80%
80%
80%
80%
80%
80%
80%
80%
80%
80%
Manual to AMT
Alt 3
80%
80%
80%
85%
85%
85%
100%
100%
100%
100%
Manual to AMT
TCp
$0
$0
$0
$227
$223
$206
$810
$796
$783
$771
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-172 Costs for an Automated Transmission
Vocational Heavy HD Regional Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Manual to AMT
DMC
$3,750
$3,638
$3,528
$3,423
$3,354
$3,287
$3,221
$3,157
$3,094
$3,032
Manual to AMT
IC
$1,134
$1,128
$1,123
$1,117
$1,114
$830
$828
$825
$823
$821
Manual to AMT
TC
$4,884
$4,766
$4,651
$4,540
$4,468
$4,117
$4,049
$3,982
$3,917
$3,853
Manual to AMT
Alt
la
5%
5%
5%
5%
5%
5%
5%
5%
5%
5%
Manual to AMT
Alt 3
5%
5%
5%
35%
35%
35%
55%
55%
55%
85%
Manual to AMT
TCp
$0
$0
$0
$1,362
$1,340
$1,235
$2,024
$1,991
$1,958
$3,082
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-173 Costs for an AMT Transmission
Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Manual to AMT
DMC
$3,750
$3,638
$3,528
$3,423
$3,354
$3,287
$3,221
$3,157
$3,094
$3,032
Manual to AMT
IC
$1,134
$1,128
$1,123
$1,117
$1,114
$830
$828
$825
$823
$821
Manual to AMT
TC
$4,884
$4,766
$4,651
$4,540
$4,468
$4,117
$4,049
$3,982
$3,917
$3,853
Manual to AMT
Alt
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
la
Manual to AMT
Alt 3
0%
0%
0%
40%
40%
40%
50%
50%
50%
50%
Manual to AMT
TCp
$0
$0
$0
$1,816
$1,787
$1,647
$2,024
$1,991
$1,958
$1,926
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.3.3 Automatic Transmission Powershift
We have estimated the cost of a powershift automatic transmission, relative to a manual
transmission, based on an estimate by Tetra Tech of $15000 (retail, 2013$). Using that estimate,
we divided by an RPE of 1.36 to arrive at an estimated cost of $11883 (DMC, 2013$, in 2018).
We consider this technology to be on the flat portion of the learning curve (curve 12) and have
applied a medium complexity ICM with short term markups through 2022. The resultant
technology costs, penetration rates and total cost applied to the package are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-174 Costs for a Powershift Automatic Transmission
Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Manual to AT
powershift
DMC
$11,883
$11,527
$11,181
$10,846
$10,629
$10,416
$10,208
$10,004
$9,803
$9,607
Manual to AT
powershift
IC
$3,593
$3,575
$3,557
$3,540
$3,529
$2,630
$2,623
$2,616
$2,608
$2,602
Manual to AT
powershift
TC
$15,476
$15,101
$14,738
$14,386
$14,158
$13,046
$12,830
$12,619
$12,412
$12,209
Manual to AT
powershift
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Manual to AT
powershift
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
30%
Manual to AT
powershift
TCp
$0
$0
$0
$1,439
$1,416
$1,305
$2,566
$2,524
$2,482
$3,663
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.3.4 Dual-clutch Transmissions (DCT)
For tractors, we have based our estimated cost of a DCT relative to a manual transmission
on a Tetra Tech estimate of $17,500 (retail, 2013$). Using that estimate, we divided by an RPE
of 1.36 to arrive at an estimated cost of $12,868 (DMC, 2013$, in 2018). We consider this
technology to be on the flat portion of the learning curve (curve 12) and have applied a medium
complexity ICM with short term markups through 2022. The resultant technology costs,
penetration rates and total cost applied to the package are shown below.
Table 2-175 Costs for a Dual Clutch Transmission (DCT)
Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Manual to DCT
DMC
$12,868
$12,482
$12,107
$11,744
$11,509
$11,279
$11,053
$10,832
$10,616
$10,403
Manual to DCT
IC
$3,890
$3,871
$3,852
$3,833
$3,821
$2,848
$2,840
$2,832
$2,825
$2,817
Manual to DCT
TC
$16,758
$16,352
$15,959
$15,577
$15,331
$14,127
$13,893
$13,664
$13,440
$13,220
Manual to DCT
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Manual to DCT
Alt 3
0%
0%
0%
5%
5%
5%
10%
10%
10%
10%
Manual to DCT
TCp
$0
$0
$0
$779
$767
$706
$1,389
$1,366
$1,344
$1,322
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.3.5 High Efficiency Gearbox (HEG)
For this technology, we have relied on our light-duty technology referred to as high
efficiency gearbox (HEG). This technology was estimated at $200(DMC, in 2010$, in 2015).
For this analysis, we have used that estimate but have scaled upward the cost of HEG by 25
percent to account for differences between light-duty and HD. Converting to 2013$ results in
costs for this technology of $267 (DMC, 2013$, in 2021). We consider this technology to be on
the flat portion of the learning curve (curve 13) and have applied a low complexity ICM with

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
short term markups through 2022. The resultant technology costs, penetration rates and total cost
applied to the package are shown below.
Table 2-176 Costs of Improved Transmissions
Vocational Light/Medium/Heavy HD Urban/Multipurpose Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
HEG
DMC
$293
$284
$276
$267
$259
$252
$244
$239
$234
$230
HEG
IC
$48
$48
$48
$48
$48
$37
$37
$37
$37
$37
HEG
TC
$341
$332
$323
$315
$307
$289
$281
$276
$272
$267
HEG
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
HEG
Alt 3
0%
0%
0%
50%
50%
50%
60%
60%
60%
62%
HEG
TCp
$0
$0
$0
$158
$153
$144
$169
$166
$163
$165
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-177 Costs of Improved Transmissions
Vocational Light/Medium/Heavy HD Regional Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
HEG
DMC
$293
$284
$276
$267
$259
$252
$244
$239
$234
$230
HEG
IC
$48
$48
$48
$48
$48
$37
$37
$37
$37
$37
HEG
TC
$341
$332
$323
$315
$307
$289
$281
$276
$272
$267
HEG
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
HEG
Alt 3
0%
0%
0%
50%
50%
50%
60%
60%
60%
70%
HEG
TCp
$0
$0
$0
$158
$153
$144
$169
$166
$163
$187
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-178 Costs for High Efficiency Gearbox (HEG) on Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
HEG
DMC
$293
$284
$276
$267
$259
$252
$244
$239
$234
$230
HEG
IC
$48
$48
$48
$48
$48
$37
$37
$37
$37
$37
HEG
TC
$341
$332
$323
$315
$307
$289
$281
$276
$272
$267
HEG
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
HEG
Alt 3
0%
0%
0%
20%
20%
20%
40%
40%
40%
70%
HEG
TCp
$0
$0
$0
$63
$61
$58
$113
$111
$109
$187
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
2.11.3.6 Early Torque Converter Lockup (TORQ) - Vocational Vehicles
For this technology, we have relied on our light-duty technology of the same. This
technology was estimated at $25 (DMC, in 2010$, in 2015). For this analysis, we have used that
estimate converted to 2013$ resulting in a cost for this technology of $26 (DMC, 2013$, in
2021). We consider this technology to be on the flat portion of the learning curve (curve 8) and
have applied a low complexity ICM with short term markups through 2018. The resultant
technology costs, penetration rates and total cost applied to the package are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-179 Costs of Early Torque Converter Lockup (TORQ)
Vocational Light/Medium HD Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
TORQ
DMC
$28
$28
$27
$26
$25
$24
$24
$23
$23
$22
TORQ
IC
$5
$5
$5
$5
$5
$4
$4
$4
$4
$4
TORQ
TC
$33
$32
$31
$31
$30
$28
$27
$27
$26
$26
TORQ
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
TORQ
Alt 3
0%
0%
0%
30%
30%
30%
40%
40%
40%
50%
TORQ
TCp
$0
$0
$0
$9
$9
$8
$11
$11
$11
$13
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-180 Costs of Early Torque Converter Lockup (TORQ)
Vocational Heavy HD Urban/Multipurpose Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
TORQ
DMC
$28
$28
$27
$26
$25
$24
$24
$23
$23
$22
TORQ
IC
$5
$5
$5
$5
$5
$4
$4
$4
$4
$4
TORQ
TC
$33
$32
$31
$31
$30
$28
$27
$27
$26
$26
TORQ
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
TORQ
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
30%
TORQ
TCp
$0
$0
$0
$3
$3
$3
$5
$5
$5
$8
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
2.11.3.7 Driveline Integration - Vocational Vehicles
We have estimated the cost of driveline integration on comments regarding the cost of
neutral idle.207 While the comment was not speaking to driveline integration, we believe that the
rationale of the comment and the cost estimate made by the commenter are applicable to the
driveline integration technology in terms of sensors and calibration required. We have divided
this cost by 1.36 to arrive at a direct manufacturing cost of $74 (DMC, 2013$, in 2021). We
consider this technology to be on the flat portion of the learning curve (curve 13) and have
applied a low complexity ICM with short term markups through 2022. The resultant technology
costs, penetration rates and total cost applied to the package are shown below.
Table 2-181 Costs of Driveline Integration
Vocational Light/Medium/Heavy HD Urban/Multipurpose Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Improved trans
DMC
$81
$78
$76
$74
$71
$69
$67
$66
$64
$63
Improved trans
IC
$13
$13
$13
$13
$13
$10
$10
$10
$10
$10
Improved trans
TC
$94
$91
$89
$87
$84
$79
$77
$76
$75
$73
Improved trans
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Improved trans
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
24%
Improved trans
TCp
$0
$0
$0
$9
$8
$8
$15
$15
$15
$18
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-182 Costs of Driveline Integration
Vocational Light/Medium/Heavy HD Regional Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Improved trans
DMC
$81
$78
$76
$74
$71
$69
$67
$66
$64
$63
Improved trans
IC
$13
$13
$13
$13
$13
$10
$10
$10
$10
$10
Improved trans
TC
$94
$91
$89
$87
$84
$79
$77
$76
$75
$73
Improved trans
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Improved trans
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
30%
Improved trans
TCp
$0
$0
$0
$9
$8
$8
$15
$15
$15
$22
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
2.11.3.8 8 Speed Transmission Relative to a 6 Speed, HD Pickups & Vans
We have based the cost of this technology on several values used in the light-duty 2017-
2025 final rule. In that rule, we presented costs for 6 to 8 speed automatic transmission, high
efficiency gearbox (HEG) and aggressive shift logic (ASL1) as separate technologies. Here we
are treating these technologies as separate for costing (since some metrics differ for each) but
considering them as being applied together as a complete group. As such, the cost for moving to
an 8 speed transmission from the base 6 would always be the summation within any given year
of the total costs shown in the tables that follow. For adding 2 gears, we have estimated the cost
at $121 (DMC, 2012$, in 2012). We consider that technology to be on the flat portion of the
learning curve (curve 7) and have applied medium complexity markups with near term markups
through 2018. For HEG, we have estimated the cost at $263 (DMC, 2012$, in 2017). We
consider this technology to be on the flat portion of the learning curve (curve 6) and have applied
low complexity markups with near term markups through 2024. For shift logic, we have
estimated the cost at $28 (DMC, 2012$, in 2015). We consider this technology to be on the flat
portion of the learning curve (curve 8) and have applied low complexity markups with near term
markups through 2018. The resultant costs for these technologies are shown below.
Table 2-183 Costs to Add 2 Transmission Gears
HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Move from 6 to 8 gears
DMC
$97
$95
$93
$91
$89
$88
$88
Move from 6 to 8 gears
IC
$34
$34
$34
$34
$34
$34
$34
Move from 6 to 8 gears
TC
$131
$129
$127
$125
$123
$123
$122
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
Table 2-184 Costs for High Efficiency Gearbox (HEG)
HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
High efficiency gearbox
DMC
$232
$225
$221
$217
$212
$208
$204
High efficiency gearbox
IC
$63
$63
$63
$63
$50
$50
$50
High efficiency gearbox
TC
$296
$288
$284
$279
$262
$258
$254
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-185 Costs for Aggressive Shift Logic Level 1
HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Aggressive shift logic 1
DMC
$25
$24
$24
$23
$23
$22
$22
Aggressive shift logic 1
IC
$5
$5
$5
$5
$5
$5
$5
Aggressive shift logic 1
TC
$30
$30
$29
$29
$28
$28
$28
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
Table 2-186 Complete Cost of Moving from the Base 6 Speed to 8 Speed Transmission
2 Gears+HEG+ASLl
HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Move from 6speed to
8speed Transmission
TC
$457
$447
$440
$433
$414
$409
$403
Notes: TC=total cost.
2.11.4 Air Conditioning
2.11.4.1 Direct AC Controls - Vocational (all)
We have estimated the cost of this technology based on an estimate from TetraTech of
$30 (retail, 2013$). Using that estimate we divided by a 1.36 RPE to arrive at a cost of $22
(DMC, 2013$, in 2014). We consider this technology to be on the flat portion of the learning
curve (curve 2) and have applied a low complexity ICM with short term markups through 2022.
The resultant technology costs, penetration rates and total cost applied to the package are shown
below.
Table 2-187 Costs for Direct Air Conditioning Controls
All Vocational HD Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
A/C direct
DMC
$20
$19
$19
$18
$18
$18
$17
$17
$17
$17
A/C direct
IC
$4
$4
$4
$4
$4
$3
$3
$3
$3
$3
A/C direct
TC
$23
$23
$23
$22
$22
$21
$20
$20
$20
$20
A/C direct
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
A/C direct
Alt 3
0%
0%
0%
100%
100%
100%
100%
100%
100%
100%
A/C direct
TCp
$0
$0
$0
$22
$22
$21
$20
$20
$20
$20
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
2.11.4.2 Indirect AC Controls - Tractors (all)
We have estimated the cost of this technology based on an estimate from TetraTech of
$218 (retail, 2013$). Using that estimate we divided by a 1.36 RPE to arrive at a cost of $160
(DMC, 2013$, in 2018). We consider this technology to be on the flat portion of the learning
curve (curve 12) and have applied a low complexity ICM with short term markups through 2022.
The resultant technology costs, penetration rates and total cost applied to the package are shown
below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-188 Costs for Indirect AC Controls
Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
A/C indirect
DMC
$160
$155
$150
$146
$143
$140
$137
$135
$132
$129
A/C indirect
IC
$29
$28
$28
$28
$28
$22
$22
$22
$22
$22
A/C indirect
TC
$188
$184
$179
$174
$171
$162
$160
$157
$154
$152
A/C indirect
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
A/C indirect
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
30%
A/C indirect
TCp
$0
$0
$0
$17
$17
$16
$32
$31
$31
$45
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
2.11.5 Axles
2.11.5.1 6x2 Axle
We have estimated the cost of this technology based on an estimate from TetraTech of
$250 (retail, 2013$). Using that estimate we divided by a 1.36 RPE to arrive at a cost of $184
(DMC, 2013$, in 2018). We consider this technology to be on the flat portion of the learning
curve (curve 12) and have applied a low complexity ICM with short term markups through 2022.
The resultant technology costs, penetration rates and total cost applied to the package are shown
below.
Table 2-189 Costs for 6x2 Axles
Class 8 Day Cab and Sleeper Cab Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Axle 6x2
DMC
$184
$178
$173
$168
$164
$161
$158
$155
$152
$149
Axle 6x2
IC
$33
$33
$33
$33
$33
$26
$26
$26
$26
$26
Axle 6x2
TC
$217
$211
$206
$200
$197
$187
$183
$180
$177
$174
Axle 6x2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Axle 6x2
Alt 3
0%
0%
0%
15%
15%
15%
25%
25%
25%
30%
Axle 6x2
TCp
$0
$0
$0
$30
$30
$28
$46
$45
$44
$52
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
2.11.5.2 Axle Disconnect
We have estimated the cost of this technology based on an estimate from TetraTech of
$140 (retail, 2013$). Using that estimate we divided by a 1.36 RPE to arrive at a cost of $103
(DMC, 2013$, in all years). We consider this technology to be on the flat portion of the learning
curve with no additional learning to occur (curve 1) and have applied a low complexity ICM
with short term markups through 2022. The resultant technology costs, penetration rates and
total cost applied to the package are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-190 Costs for Axle Disconnect
Vocational Heavy HD Multipurpose Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Axle disconnect
DMC
$103
$103
$103
$103
$103
$103
$103
$103
$103
$103
Axle disconnect
IC
$18
$18
$18
$18
$18
$14
$14
$14
$14
$14
Axle disconnect
TC
$121
$121
$121
$121
$121
$117
$117
$117
$117
$117
Axle disconnect
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Axle disconnect
Alt 3
0%
0%
0%
5%
5%
5%
15%
15%
15%
25%
Axle disconnect
TCp
$0
$0
$0
$6
$6
$6
$18
$18
$18
$29
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-191 Costs for Axle Disconnect
Vocational Heavy HD Regional Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Axle disconnect
DMC
$103
$103
$103
$103
$103
$103
$103
$103
$103
$103
Axle disconnect
IC
$18
$18
$18
$18
$18
$14
$14
$14
$14
$14
Axle disconnect
TC
$121
$121
$121
$121
$121
$117
$117
$117
$117
$117
Axle disconnect
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Axle disconnect
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
30%
Axle disconnect
TCp
$0
$0
$0
$12
$12
$12
$23
$23
$23
$35
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
2.11.5.3 Axle Downspeed
We have estimated the cost of this technology based on engineering judgment at $50
(DMC, 2013$, in 2018). This DMC is expected to cover development and some testing and
integration work since there is no real hardware required for this technology. We consider this
technology to be on the flat portion of the learning curve (curve 12) and have applied a low
complexity ICM with short term markups through 2022. The resultant technology costs,
penetration rates and total cost applied to the package are shown below.
Table 2-192 Costs for Axle Downspeeding
Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Axle downspeed
DMC
$50
$49
$47
$46
$45
$44
$43
$42
$41
$40
Axle downspeed
IC
$9
$9
$9
$9
$9
$7
$7
$7
$7
$7
Axle downspeed
TC
$59
$57
$56
$54
$54
$51
$50
$49
$48
$47
Axle downspeed
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Axle downspeed
Alt 3
0%
0%
0%
20%
20%
20%
40%
40%
40%
60%
Axle downspeed
TCp
$0
$0
$0
$11
$11
$10
$20
$20
$19
$28
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.11.5.4 High Efficiency Axle (Axle HE)
We have estimated the cost of this technology based on an estimate from TetraTech of
$250 (retail, 2013$), an estimate applicable to tractors having 3 axles. Using that estimate we
divided by a 1.36 RPE to arrive at a cost of $184 (DMC, 2013$, in 2018). We consider this
estimate to be applicable also to vocational HH vehicles since these generally have 3 axles. For
vocational light/medium HD vehicles, which generally have 2 axles, we have estimated the
DMC at 2/3 the vocational heavy HD/tractor cost, or $123 (DMC, 2013$, in 2018). We consider
this technology to be on the flat portion of the learning curve (curve 12) and have applied a low
complexity ICM with short term markups through 2022. The resultant technology costs,
penetration rates and total cost applied to the package are shown below.
Table 2-193 Costs for High Efficiency Axles
Vocational Light/Medium HD Urban/Multipurpose/Regional Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Axle low friction lubes
DMC
$123
$119
$115
$112
$110
$107
$105
$103
$101
$99
Axle low friction lubes
IC
$22
$22
$22
$22
$22
$17
$17
$17
$17
$17
Axle low friction lubes
TC
$144
$141
$137
$134
$131
$124
$122
$120
$118
$116
Axle low friction lubes
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Axle low friction lubes
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
30%
Axle low friction lubes
TCp
$0
$0
$0
$13
$13
$12
$24
$24
$24
$35
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-194 Costs for High Efficiency Axles
Vocational Heavy HD Urban/Multipurpose/Regional Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Axle low friction lubes
DMC
$184
$178
$173
$168
$164
$161
$158
$155
$152
$149
Axle low friction lubes
IC
$33
$33
$33
$33
$33
$26
$26
$26
$26
$26
Axle low friction lubes
TC
$217
$211
$206
$200
$197
$187
$183
$180
$177
$174
Axle low friction lubes
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Axle low friction lubes
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
30%
Axle low friction lubes
TCp
$0
$0
$0
$20
$20
$19
$37
$36
$35
$52
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-195 Costs for High Efficiency Axles
Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Axle low friction lubes
DMC
$184
$178
$173
$168
$164
$161
$158
$155
$152
$149
Axle low friction lubes
IC
$33
$33
$33
$33
$33
$26
$26
$26
$26
$26
Axle low friction lubes
TC
$217
$211
$206
$200
$197
$187
$183
$180
$177
$174
Axle low friction lubes
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Axle low friction lubes
Alt 3
0%
0%
0%
30%
30%
30%
65%
65%
65%
80%
Axle low friction lubes
TCp
$0
$0
$0
$60
$59
$56
$119
$117
$115
$139
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.11.6 Idle Reduction
2.11.6.1 Auxiliary Power Units (APU)
We have estimated the cost of the APU technology at $8000 retail (2013$). We divided
that by 1.36 to arrive at a cost of $5882 (DMC, 2013$, in 2014). We consider this technology to
be on the flat portion of the learning curve (curve 2) and have applied a low complexity ICM
with short term markups through 2022. The resultant technology costs, penetration rates and
total cost applied to the package are shown below.
Table 2-196 Costs for Auxiliary Power Units (APU)
On Sleeper Cab Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
APU
DMC
$5,208
$5,103
$5,001
$4,901
$4,803
$4,707
$4,613
$4,521
$4,476
$4,431
APU
IC
$1,041
$1,039
$1,038
$1,037
$1,035
$817
$816
$816
$815
$815
APU
TC
$6,248
$6,143
$6,039
$5,938
$5,839
$5,524
$5,429
$5,336
$5,291
$5,246
APU
Alt
la
9%
9%
9%
9%
9%
9%
0%
0%
0%
0%
APU
Alt 3
9%
9%
9%
30%
30%
30%
0%
0%
0%
0%
APU
TCp
$0
$0
$0
$1,247
$1,226
$1,160
$0
$0
$0
$0
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.6.2 Auxiliary Power Units, Battery Powered (APUB)
We have estimated the cost of the battery powered APU technology at $6400 retail
(2013$). We divided that by 1.36 to arrive at a cost of $5070 (DMC, 2013$, in 2014). We
consider this technology to be on the flat portion of the learning curve (curve 2) and have applied
a low complexity ICM with short term markups through 2022. The resultant technology costs,
penetration rates and total cost applied to the package are shown below.
Table 2-197 Costs for Battery Powered Auxiliary Power Units (APU B) on Sleeper Cab Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
APU B
DMC
$4,489
$4,399
$4,311
$4,225
$4,140
$4,057
$3,976
$3,897
$3,858
$3,819
APU B
IC
$897
$896
$895
$894
$893
$704
$703
$703
$703
$702
APU B
TC
$5,386
$5,295
$5,206
$5,118
$5,033
$4,761
$4,680
$4,600
$4,560
$4,522
APUB
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
APU B
Alt 3
0%
0%
0%
10%
10%
10%
10%
10%
10%
15%
APU B
TCp
$0
$0
$0
$512
$503
$476
$468
$460
$456
$678
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.6.3 Auxiliary Power Units with Diesel Particulate Filters (APUwDPF)
We have estimated the cost of the DPF equipped APU technology at $10,000 retail
(2013$). See Preamble Section III.C for an explanation of the estimate for the cost of the APU.
We divided that by 1.36 to arrive at a cost of $7922 (DMC, 2013$, in 2014). We consider this
technology to be on the flat portion of the learning curve (curve 2) and have applied a low

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
complexity ICM with short term markups through 2022. The resultant technology costs,
penetration rates and total cost applied to the package are shown below.
Table 2-198 Costs for Auxiliary Power Units with Diesel Particulate Filters (APUwDPF) on Sleeper Cab
Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
APUwDPF
DMC
$7,013
$6,873
$6,736
$6,601
$6,469
$6,340
$6,213
$6,089
$6,028
$5,967
APUwDPF
IC
$1,402
$1,400
$1,398
$1,396
$1,395
$1,100
$1,099
$1,098
$1,098
$1,098
APUwDPF
TC
$8,415
$8,273
$8,134
$7,997
$7,864
$7,439
$7,312
$7,187
$7,126
$7,065
APUwDPF
Alt
la
0%
0%
0%
0%
0%
0%
9%
9%
9%
9%
APUwDPF
Alt 3
0%
0%
0%
0%
0%
0%
40%
40%
40%
40%
APUwDPF
TCp
$0
$0
$0
$0
$0
$0
$2,267
$2,228
$2,209
$2,190
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.6.4 Fuel Operated Heater (FOH)
We have estimated the cost of the FOH technology at $1200 retail (2013$). We divided
that by 1.36 to arrive at a cost of $882 (DMC, 2013$, in 2014). We consider this technology to
be on the flat portion of the learning curve (curve 2) and have applied a low complexity ICM
with short term markups through 2022. The resultant technology costs, penetration rates and
total cost applied to the package are shown below.
Table 2-199 Costs for Fuel Operated Heaters (FOH) on Sleeper Cab Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
FOH
DMC
$781
$766
$750
$735
$720
$706
$692
$678
$671
$665
FOH
IC
$156
$156
$156
$156
$155
$122
$122
$122
$122
$122
FOH
TC
$937
$921
$906
$891
$876
$829
$814
$800
$794
$787
FOH
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
FOH
Alt 3
0%
0%
0%
0%
10%
10%
10%
10%
10%
15%
FOH
TCp
$0
$0
$0
$0
$88
$83
$81
$80
$79
$118
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.6.5 Neutral Idle
We have estimated the cost of neutral idle on comments received.208 A commenter stated
that a cost of $100 would be more appropriate than the estimate used in the proposal. We have
considered the $100 estimate to be in 2013$ and applicable in all years meaning that we consider
this technology to be on the flat portion of the learning curve with no additional learning to occur
(curve 1) and have applied a low complexity ICM with short term markups through 2022. The
resultant technology costs, penetration rates and total cost applied to the package are below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-200 Costs for Neutral Idle Technology
Vocational Light/Medium/Heavy HD Urban/Multipurpose Vehicles
(2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Neutral idle
DMC
$100
$100
$100
$100
$100
$100
$100
$100
$100
$100
Neutral idle
IC
$18
$18
$18
$18
$18
$14
$14
$14
$14
$14
Neutral idle
TC
$118
$118
$118
$118
$118
$114
$114
$114
$114
$114
Neutral idle
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Neutral idle
Alt 3
0%
0%
0%
50%
50%
50%
70%
70%
70%
60%
Neutral idle
TCp
$0
$0
$0
$59
$59
$57
$80
$80
$80
$68
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
2.11.6.6 Stop-start with Enhancements (Stop-startenhanced)
We have estimated the cost of this technology based on several cost estimates. First, an
estimate from TetraTech of $700 (retail, 2013$) for gasoline HD pickups and vans and $1500
(retail, 2013$) for diesel HD pickups and vans. Using these values, we divided by a 1.36 RPE to
arrive at $515 (DMC, 2013$, in 2021) and $1103 (DMC, 2013$, in 2021) which were considered
appropriate for vocational MH and HH vehicles, respectively. To these estimates, we have
added the costs for improved accessories used for HD pickups and vans of $126 (DMC, 2013$,
in 2015) which is based on values from the 2017-2025 light-duty FRM. However, to account for
the heavier vocational vehicles relative to the HD pickup and vans, we have scaled upward the
improved accessory value by 50 percent to arrive at a cost of $189 (DMC, 2013$, in 2015). We
have then added these values to arrive at costs of $704 (DMC, 2013$, in 2021) and $1292
(DMC, 2013$, in 2021) and have applied the lower cost to vocational medium HD vehicles and
the higher cost to vocational heavy HD vehicles. For vocational light HD, we have used the
stop-start cost for the 2017-2025 rule for LD pickups ($377 DMC, 2012$, in 2015) but have
scaled upward that value by 25 percent to account for the weight difference between the LD and
vocational light HD vehicles. Doing this results in a cost of $479 (DMC, 2013$, in 2021).
Adding to that the $189 value for improved accessories mentioned earlier gives the resultant
vocational light HD cost of $669 (DMC, 2013$, in 2021). We consider all of these technologies
to be on the flat portion of the learning curve (curve 13) and have applied a medium complexity
ICM with short term markups through 2022. The resultant technology costs, penetration rates
and total cost applied to the package are shown below.
Table 2-201 Costs for Enhanced Stop-start with Enhancements
Vocational Light HD Urban/Multipurpose Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Stop-start enhanced
DMC
$733
$711
$689
$669
$648
$629
$610
$598
$586
$574
Stop-start enhanced
IC
$205
$204
$203
$202
$201
$149
$149
$148
$148
$148
Stop-start enhanced
TC
$938
$915
$892
$871
$850
$779
$759
$746
$734
$722
Stop-start enhanced
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Stop-start enhanced
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
30%
Stop-start enhanced
TCp
$0
$0
$0
$87
$85
$78
$152
$149
$147
$217
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-202 Costs for Enhanced Stop-start Vocational Medium HD Urban/Multipurpose Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Stop-start enhanced
DMC
$771
$748
$726
$704
$683
$662
$642
$630
$617
$605
Stop-start enhanced
IC
$216
$215
$214
$213
$212
$157
$157
$156
$156
$155
Stop-start enhanced
TC
$987
$963
$939
$917
$894
$820
$799
$786
$773
$760
Stop-start enhanced
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Stop-start enhanced
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
30%
Stop-startenhanced
TCp
$0
$0
$0
$92
$89
$82
$160
$157
$155
$228
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-203 Costs for Enhanced Stop-start Vocational Heavy HD Urban/Multipurpose Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Stop-start enhanced
DMC
$1,416
$1,373
$1,332
$1,292
$1,253
$1,216
$1,179
$1,156
$1,133
$1,110
Stop-start enhanced
IC
$397
$395
$393
$391
$389
$289
$288
$287
$286
$285
Stop-start enhanced
TC
$1,813
$1,768
$1,725
$1,683
$1,642
$1,505
$1,467
$1,442
$1,419
$1,395
Stop-start enhanced
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Stop-start enhanced
Alt 3
0%
0%
0%
0%
0%
0%
10%
10%
10%
20%
Stop-start enhanced
TCp
$0
$0
$0
$0
$0
$0
$147
$144
$142
$279
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
For HD pickups and vans, we have based our costs for stop-start systems on the values
used in the light-duty 2017-2025 final rule, but have scaled upward those costs by 25 percent to
account for the larger and harder starting HD engines. Using this approach and converting to
2012$ results in a cost of $471 (DMC, 2012$, in 2015). We consider this technology to be on
the steep portion of the learning curve (curve 9, note the different year of cost-applicability
relative to the vocational cost discussed above) and have applied medium complexity markups
with near term markups through 2018. The resultant costs for HD pickups and vans are shown
below.
Table 2-204 Costs of Stop-start
HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Stop-start
DMC
$404
$392
$380
$369
$358
$351
$344
Stop-start
IC
$134
$134
$134
$133
$133
$133
$132
Stop-start
TC
$539
$526
$514
$502
$491
$483
$476
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
2.11.6.7 Automatic Engine Shutdown System (AESS)
We have estimated the cost of an AESS at $50 retail (2013$). This system should be low
cost since the engine control software already features the necessary code. The cost here is
simply meant to cover the costs of setting the software correctly to take advantage of the already
existing feature. We have divided the $50 by 1.36 to arrive at a cost of $40 (DMC, 2013$, in
2014). We have placed this technology on the steep portion of the learning curve today but flat
by the 2019 timeframe (curve 4) and have applied a low complexity ICM with short term

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
markups through 2022. The resultant technology costs, penetration rates and total cost applied to
the package are shown below.
Table 2-205 Costs for Automatic Engine Shutdown System on Vocational Light/Medium/Heavy HD
Urban/Multipurpose Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
AESS
DMC
$25
$25
$24
$23
$22
$22
$21
$20
$20
$20
AESS
IC
$7
$7
$7
$7
$7
$5
$5
$5
$5
$5
AESS
TC
$32
$31
$31
$30
$29
$27
$27
$26
$26
$25
AESS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
AESS
Alt 3
0%
0%
0%
30%
30%
30%
60%
60%
60%
70%
AESS
TCp
$0
$0
$0
$9
$9
$8
$16
$16
$15
$18
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-206 Costs for Automatic Engine Shutdown System on Vocational Light/Medium/Heavy HD Regional
Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
AESS
DMC
$25
$25
$24
$23
$22
$22
$21
$20
$20
$20
AESS
IC
$7
$7
$7
$7
$7
$5
$5
$5
$5
$5
AESS
TC
$32
$31
$31
$30
$29
$27
$27
$26
$26
$25
AESS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
AESS
Alt 3
0%
0%
0%
40%
40%
40%
80%
80%
80%
90%
AESS
TCp
$0
$0
$0
$12
$12
$11
$21
$21
$20
$23
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-207 Costs for Automatic Engine Shutdown System on Sleeper Cab Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
AESS
DMC
$25
$25
$24
$23
$22
$22
$21
$20
$20
$20
AESS
IC
$7
$7
$7
$7
$7
$5
$5
$5
$5
$5
AESS
TC
$32
$31
$31
$30
$29
$27
$27
$26
$26
$25
AESS
Alt la
80%
80%
80%
80%
80%
80%
80%
80%
80%
80%
AESS
Alt 3
80%
80%
80%
40%
40%
40%
30%
30%
30%
15%
AESS
TCp
$0
$0
$0
-$12
-$12
-$11
-$13
-$13
-$13
-$16
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
2.11.6.8 Automatic Engine Shutdown System with Auto-Start
(AE S S_w Au to Sta rt)
We have estimated the cost of an AESS with auto-start at $2700 retail (2013$). We have
divided this value by 1.36 to arrive at a cost of $2139 (DMC, 2013$, in 2014). We have placed
this technology on the steep portion of the learning curve (curve 4) and have applied a low
complexity ICM with short term markups through 2022. The resultant technology costs,
penetration rates and total cost applied to the package are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-208 Costs for Automatic Engine Shutdown System with Auto-Start on Sleeper Cab Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
AESS wAutoStart
DMC
$1,369
$1,328
$1,288
$1,249
$1,212
$1,176
$1,140
$1,106
$1,084
$1,062
AESS wAutoStart
IC
$372
$371
$370
$370
$369
$294
$293
$293
$293
$293
AESS wAutoStart
TC
$1,740
$1,699
$1,659
$1,619
$1,581
$1,469
$1,434
$1,399
$1,377
$1,355
AESS wAutoStart
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
AESS wAutoStart
Alt 3
0%
0%
0%
10%
10%
10%
10%
10%
10%
15%
AESS wAutoStart
TCp
$0
$0
$0
$162
$158
$147
$143
$140
$138
$203
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
2.11.7 Electrification (strong/mild HEV, full EV)
2.11.7.1 Strong Hybrid Electric Vehicle (strong HEV)
We have estimated the cost of this technology using the costs estimated in the 2017-2025
light-duty rule for a light-duty pickup strong HEV. There we estimated the cost at $2729 (DMC,
2010$, in 2021) for a LD truck with a 5200 pound curb weight. We have then scaled upward
that value using the ratio of test weights for HD pickups in our MY2014 market file (8739
pounds) to the test weight of the 5200 pound LD truck (5500 pounds). The resultant strong
hybrid costs become $4335 (DMC, 2012$, in 2021) for HD pickups and vans. We consider this
technology to be on the steep portion of the learning curve today but on the flat portion by 2021
(curve 11) and have applied high complexity level 1 with short term markups through 2024. The
resultant technology costs are shown below for HD pickups and vans.
Table 2-209 Costs of Strong Hybrid
HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Strong HEV
DMC
$4,335
$4,205
$4,079
$3,957
$3,838
$3,723
$3,648
Strong HEV
IC
$2,443
$2,435
$2,427
$2,419
$1,482
$1,478
$1,476
Strong HEV
TC
$6,779
$6,640
$6,506
$6,376
$5,320
$5,201
$5,124
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
2.11.7.2 Mild hybrid Electric Vehicle (mild HEV)
We have estimated the cost of this technology using the costs estimated in the 2017-2025
light-duty rule for a light-duty pickup mild HEV. There we estimated the cost at $983 (DMC,
2010$, in 2021) for a LD truck with a 3500 pound curb weight. We have then scaled upward
that value using the ratio of curb weights for HD pickups of 6500 pounds to the 3500 pound curb
weight. The resultant mild hybrid costs become $1894 (DMC, 2012$, in 2017) for HD pickups
and vans. We consider this technology to be on the flat portion of the learning curve (curve 6)
and have applied high complexity level 1 with short term markups through 2024. The resultant
technology costs are shown below for HD pickups and vans.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-210 Costs of Mild Hybrid
HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Mild HEV
DMC
$1,677
$1,626
$1,594
$1,562
$1,531
$1,500
$1,470
Mild HEV
IC
$1,053
$1,050
$1,048
$1,046
$643
$642
$641
Mild HEV
TC
$2,730
$2,677
$2,642
$2,608
$2,173
$2,142
$2,111
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
For vocational vehicle mild hybrids, we have scaled upward from the HD pickup and van
values using best estimates of curb weights. For vocational vehicles, we have used curb weights
of 16,000 for light HD, 25,150 for medium HD and 42,000 for heavy HD relative to a 6500
pound value for HD pickups. Scaling based on curb weight here should provide an acceptable
scaling of costs with battery and motor sizes since those are generally directly correlated with the
weight of the vehicle itself. Using these scaling factors results in costs for complete mild hybrid
systems for light, medium and heavy HD, respectively, of $4747, $7462 and $12461 (DMC,
2012$, in 2018). We consider this technology to be on the flat portion of the learning curve
(curve 12) and have applied high complexity level 1 with short term markups through 2022. The
resultant technology costs, penetration rates and total cost applied to the package are shown are
shown below.
Table 2-211 Costs for Mild Hybrid
Vocational Light HD Urban/Multipurpose Vehicles (2013$)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Mild HEV
DMC
$4,747
$4,605
$4,467
$4,333
$4,246
$4,161
$4,078
$3,996
$3,916
$3,838
Mild HEV
IC
$2,018
$2,007
$1,997
$1,987
$1,981
$1,975
$1,969
$1,963
$1,247
$1,244
Mild HEV
TC
$6,765
$6,612
$6,464
$6,320
$6,227
$6,136
$6,046
$5,959
$5,164
$5,082
Mild HEV
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Mild HEV
Alt 3
0%
0%
0%
0%
0%
0%
3%
3%
3%
6%
Mild HEV
TCp
$0
$0
$0
$0
$0
$0
$181
$179
$155
$305
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-212 Costs for Mild Hybrid
Vocational Medium HD Urban/Multipurpose Vehicles (2013$)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Mild HEV
DMC
$7,462
$7,238
$7,021
$6,810
$6,674
$6,541
$6,410
$6,282
$6,156
$6,033
Mild HEV
IC
$3,171
$3,155
$3,139
$3,124
$3,114
$3,104
$3,094
$3,085
$1,961
$1,956
Mild HEV
TC
$10,633
$10,393
$10,160
$9,934
$9,788
$9,645
$9,504
$9,367
$8,116
$7,989
Mild HEV
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Mild HEV
Alt 3
0%
0%
0%
0%
0%
0%
3%
3%
3%
6%
Mild HEV
TCp
$0
$0
$0
$0
$0
$0
$285
$281
$243
$479
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-213 Costs for Mild Hybrid
Vocational Heavy HD Urban/Multipurpose Vehicles (2013$)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Mild HEV
DMC
$12,461
$12,087
$11,725
$11,373
$11,146
$10,923
$10,704
$10,490
$10,280
$10,075
Mild HEV
IC
$5,296
$5,269
$5,242
$5,217
$5,200
$5,184
$5,168
$5,152
$3,274
$3,267
Mild HEV
TC
$17,757
$17,356
$16,967
$16,590
$16,345
$16,106
$15,872
$15,642
$13,554
$13,341
Mild HEV
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Mild HEV
Alt 3
0%
0%
0%
0%
0%
0%
3%
3%
3%
6%
Mild HEV
TCp
$0
$0
$0
$0
$0
$0
$476
$469
$407
$800
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.7.3 Hybrid electric Vehicle without Stop-Start (HEVnoSS)
We have estimated the cost of a hybrid electric system without any stop-start technology
at $8500 retail (2013$). We have divided this value by 1.36 to arrive at a cost of $6250 (DMC,
2013$, in 2021). We have placed this technology on the steep portion of the learning curve
(curve 11) and have applied high complexity level 1 ICM with short term markups through 2022
The resultant technology costs, penetration rates and total cost applied to the package are shown
below.
Table 2-214 Costs for Hybrid Electric without Stop-start, Vocational Light/Medium/Heavy HD
Urban/Multipurpose Vehicles (2013$)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
HEVnoSS
DMC
$9,766
$7,813
$7,813
$6,250
$6,063
$5,881
$5,704
$5,533
$5,367
$5,260
HEVnoSS
IC
$2,914
$2,771
$2,771
$2,656
$2,643
$1,669
$1,662
$1,656
$1,650
$1,646
HEVnoSS
TC
$12,679
$10,583
$10,583
$8,906
$8,705
$7,549
$7,366
$7,189
$7,017
$6,906
HEVnoSS
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
HEVnoSS
Alt 3
0%
0%
0%
2%
2%
2%
5%
5%
5%
8%
HEVnoSS
TCp
$0
$0
$0
$178
$174
$151
$368
$359
$351
$552
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
2.11.8 Tires
2.11.8.1 Lower Rolling Resistance Tires ($/tire)
We have estimated the cost of lower rolling resistance tires based on an estimate from
TetraTech of $30 (retail, 2013$). Using that estimate we divided by a 1.36 RPE to arrive at a
cost of $22 (DMC, 2013$) but consider that cost valid in different years depending on the level
of rolling resistance. For LRR tires level 1 and 2, we consider that $22 value valid in 2014, level
3 in 2018, level 4 and level 5 (new for this FRM analysis) in 2021. We consider this technology
to be on the flat portion of the curve with LRR tires level 1 and 2 on curve 2, LRR tires level 3
on curve 12 and LRR tires level 4 and 5 on curve 13. We have applied a low complexity markup
to LRR tires levels 1 and 2 with short term markups through 2022. For LRR tires level 3, we
have applied a medium complexity markup with short term markups through 2025, for LRR tires
level 4, we have applied a medium complexity markup with short term markups through 2028,

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
and for LRR tires level 5, and we have applied a medium complexity markup with short term
markups through 2031. As a result, despite using the same DMC for each level of rolling
resistance, our tire costs can vary year-over-year for each of the 5 levels of rolling resistance
considered. The resultant costs on a per-tire basis are shown in Table 2-215. Table 2-216
through Table 2-239 show the costs per vocational vehicle, tractor or trailer depending on the
number of tires present.
Table 2-215 Costs for Lower Rolling Resistance Tires at each LRR Level (2013$/tire)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
DMC
$20
$19
$19
$18
$18
$18
$17
$17
$17
$17
LRR - level 2
DMC
$20
$19
$19
$18
$18
$18
$17
$17
$17
$17
LRR - level 3
DMC
$22
$21
$21
$20
$20
$19
$19
$19
$18
$18
LRR - level 4
DMC
$24
$23
$23
$22
$21
$21
$20
$20
$19
$19
LRR - level 5
DMC
$24
$23
$23
$22
$21
$21
$20
$20
$19
$19
LRR - level 1
IC
$4
$4
$4
$4
$4
$3
$3
$3
$3
$3
LRR - level 2
IC
$4
$4
$4
$4
$4
$3
$3
$3
$3
$3
LRR - level 3
IC
$7
$7
$7
$7
$7
$7
$7
$6
$5
$5
LRR - level 4
IC
$7
$7
$7
$7
$7
$7
$7
$7
$7
$7
LRR - level 5
IC
$7
$7
$7
$7
$7
$7
$7
$7
$7
$7
LRR - level 1
TC
$23
$23
$23
$22
$22
$21
$20
$20
$20
$20
LRR - level 2
TC
$23
$23
$23
$22
$22
$21
$20
$20
$20
$20
LRR - level 3
TC
$29
$28
$27
$27
$26
$26
$25
$25
$23
$23
LRR - level 4
TC
$31
$30
$29
$29
$28
$27
$27
$26
$26
$25
LRR - level 5
TC
$31
$30
$29
$29
$28
$27
$27
$26
$26
$25
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.11.8.2 Lower RR Steer Tires, Vocational Vehicles
Table 2-216 Costs for Lower Rolling Resistance Steer Tires
Vocational Light/Medium HD Urban Vehicles
(2013$/vehicle @ 2 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$47
$46
$45
$45
$44
$41
$41
$40
$40
$39
LRR - level 2
TC
$47
$46
$45
$45
$44
$41
$41
$40
$40
$39
LRR - level 3
TC
$57
$56
$55
$53
$53
$52
$51
$50
$46
$45
LRR - level 4
TC
$62
$60
$59
$57
$56
$55
$53
$53
$52
$51
LRR - level 5
TC
$62
$60
$59
$57
$56
$55
$53
$53
$52
$51
LRR - level 1
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
100%
100%
100%
0%
0%
0%
0%
0%
0%
0%
LRR - level 2
Alt 3
0%
0%
0%
100%
100%
100%
100%
100%
100%
0%
LRR - level 3
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
100%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
-$45
-$44
-$41
-$41
-$40
-$40
-$39
LRR - level 2
TCp
$0
$0
$0
$45
$44
$41
$41
$40
$40
$0
LRR - level 3
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$45
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-217 Costs for Lower Rolling Resistance Steer Tires
Vocational Light/Medium/Heavy HD Multipurpose/Regional and Heavy HD Urban Vehicles
(2013$/vehicle @ 2 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$47
$46
$45
$45
$44
$41
$41
$40
$40
$39
LRR - level 2
TC
$47
$46
$45
$45
$44
$41
$41
$40
$40
$39
LRR - level 3
TC
$57
$56
$55
$53
$53
$52
$51
$50
$46
$45
LRR - level 4
TC
$62
$60
$59
$57
$56
$55
$53
$53
$52
$51
LRR - level 5
TC
$62
$60
$59
$57
$56
$55
$53
$53
$52
$51
LRR - level 1
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
100%
100%
100%
0%
0%
0%
0%
0%
0%
0%
LRR - level 2
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt 3
0%
0%
0%
100%
100%
100%
0%
0%
0%
0%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt 3
0%
0%
0%
0%
0%
0%
100%
100%
100%
100%
LRR - level 1
TCp
$0
$0
$0
-$45
-$44
-$41
-$41
-$40
-$40
-$39
LRR - level 2
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 3
TCp
$0
$0
$0
$53
$53
$52
$0
$0
$0
$0
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 5
TCp
$0
$0
$0
$0
$0
$0
$53
$53
$52
$51
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.11.8.3 Lower RR Drive Tires, Vocational Vehicles
Table 2-218 Costs for Lower Rolling Resistance Drive Tires, Vocational Light HD Urban Vehicles
(2013$ @ 4 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 2
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 3
TC
$115
$112
$109
$107
$105
$103
$102
$100
$92
$91
LRR - level 4
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 5
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 1
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
100%
100%
100%
0%
0%
0%
0%
0%
0%
0%
LRR - level 2
Alt 3
0%
0%
0%
100%
100%
100%
100%
100%
100%
50%
LRR - level 3
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
50%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
-$89
-$88
-$83
-$81
-$80
-$79
-$79
LRR - level 2
TCp
$0
$0
$0
$89
$88
$83
$81
$80
$79
$39
LRR - level 3
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$45
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative
Table 2-219 Costs for Lower Rolling Resistance Drive Tires, Vocational Light HD Multipurpose Vehicles
(2013$ @ 4 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 2
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 3
TC
$115
$112
$109
$107
$105
$103
$102
$100
$92
$91
LRR - level 4
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 5
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 1
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
100%
100%
100%
0%
0%
0%
0%
0%
0%
0%
LRR - level 2
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt 3
0%
0%
0%
100%
100%
100%
100%
100%
100%
100%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
-$89
-$88
-$83
-$81
-$80
-$79
-$79
LRR - level 2
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 3
TCp
$0
$0
$0
$107
$105
$103
$102
$100
$92
$91
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-220 Costs for Lower Rolling Resistance Drive Tires, Vocational Light HD Regional Vehicles
(2013$ @ 4 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 2
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 3
TC
$115
$112
$109
$107
$105
$103
$102
$100
$92
$91
LRR - level 4
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 5
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 1
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
100%
100%
100%
0%
0%
0%
0%
0%
0%
0%
LRR - level 2
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt 3
0%
0%
0%
100%
100%
100%
100%
100%
100%
100%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
-$89
-$88
-$83
-$81
-$80
-$79
-$79
LRR - level 2
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 3
TCp
$0
$0
$0
$107
$105
$103
$102
$100
$92
$91
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative
Table 2-221 Costs for Lower Rolling Resistance Drive Tires, Vocational Medium HD Urban Vehicles
(2013$ @ 4 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 2
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 3
TC
$115
$112
$109
$107
$105
$103
$102
$100
$92
$91
LRR - level 4
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 5
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 1
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
100%
100%
100%
100%
100%
100%
100%
100%
100%
50%
LRR - level 2
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
50%
LRR - level 3
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
-$39
LRR - level 2
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$39
LRR - level 3
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-222 Costs for Lower Rolling Resistance Drive Tires, Vocational Medium HD Multipurpose Vehicles
(2013$ @ 4 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 2
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 3
TC
$115
$112
$109
$107
$105
$103
$102
$100
$92
$91
LRR - level 4
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 5
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 1
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
100%
100%
100%
100%
100%
100%
50%
50%
50%
0%
LRR - level 2
Alt 3
0%
0%
0%
0%
0%
0%
50%
50%
50%
0%
LRR - level 3
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
100%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
$0
$0
$0
-$41
-$40
-$40
-$79
LRR - level 2
TCp
$0
$0
$0
$0
$0
$0
$41
$40
$40
$0
LRR - level 3
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$91
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative
Table 2-223 Costs for Lower Rolling Resistance Drive Tires, Vocational Medium HD Regional Vehicles
	(2013$ @ 4 tires/vehicle)	
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 2
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 3
TC
$115
$112
$109
$107
$105
$103
$102
$100
$92
$91
LRR - level 4
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 5
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 1
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
100%
100%
100%
100%
100%
100%
0%
0%
0%
0%
LRR - level 2
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt 3
0%
0%
0%
0%
0%
0%
100%
100%
100%
100%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
$0
$0
$0
-$81
-$80
-$79
-$79
LRR - level 2
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 3
TCp
$0
$0
$0
$0
$0
$0
$102
$100
$92
$91
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-224 Costs for Lower Rolling Resistance Drive Tires, Vocational Heavy HD Urban Vehicles
(2013$ @ 8 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR - level 2
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR - level 3
TC
$230
$224
$219
$214
$210
$207
$204
$200
$184
$181
LRR - level 4
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR - level 5
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR - level 1
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
100%
100%
100%
100%
100%
100%
100%
100%
100%
0%
LRR - level 2
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
100%
LRR - level 3
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
-$157
LRR - level 2
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$157
LRR - level 3
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative
Table 2-225 Costs for Lower Rolling Resistance Drive Tires, Vocational Heavy HD Multipurpose Vehicles
(2013$ @ 8 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR - level 2
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR - level 3
TC
$230
$224
$219
$214
$210
$207
$204
$200
$184
$181
LRR - level 4
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR - level 5
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR - level 1
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
100%
100%
100%
0%
0%
0%
0%
0%
0%
0%
LRR - level 2
Alt 3
0%
0%
0%
100%
100%
100%
100%
100%
100%
0%
LRR - level 3
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
100%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
-$178
-$175
-$166
-$163
-$160
-$159
-$157
LRR - level 2
TCp
$0
$0
$0
$178
$175
$166
$163
$160
$159
$0
LRR - level 3
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$181
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-226 Costs for Lower Rolling Resistance Drive Tires, Vocational Heavy HD Regional Vehicles
(2013$ @ 8 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR - level 2
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR - level 3
TC
$230
$224
$219
$214
$210
$207
$204
$200
$184
$181
LRR - level 4
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR - level 5
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR - level 1
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR - level 2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
100%
100%
100%
0%
0%
0%
0%
0%
0%
0%
LRR - level 2
Alt 3
0%
0%
0%
100%
100%
100%
0%
0%
0%
0%
LRR - level 3
Alt 3
0%
0%
0%
0%
0%
0%
100%
100%
100%
100%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
-$178
-$175
-$166
-$163
-$160
-$159
-$157
LRR - level 2
TCp
$0
$0
$0
$178
$175
$166
$0
$0
$0
$0
LRR - level 3
TCp
$0
$0
$0
$0
$0
$0
$204
$200
$184
$181
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative
2.11.8.4 Lower RR Steer Tires, Tractors
Table 2-227 Costs for Lower Rolling Resistance Steer Tires
Day Cab Low Roof & Sleeper Cab Low/Medium Roof Tractors
(2013$/vehicle @ 2 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$47
$46
$45
$45
$44
$41
$41
$40
$40
$39
LRR - level 2
TC
$47
$46
$45
$45
$44
$41
$41
$40
$40
$39
LRR - level 3
TC
$57
$56
$55
$53
$53
$52
$51
$50
$46
$45
LRR - level 4
TC
$62
$60
$59
$57
$56
$55
$53
$53
$52
$51
LRR - level 1
Alt la
50%
50%
50%
50%
50%
50%
50%
50%
50%
50%
LRR - level 2
Alt la
10%
10%
10%
10%
10%
10%
10%
10%
10%
10%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
50%
50%
50%
35%
35%
35%
25%
25%
25%
20%
LRR - level 2
Alt 3
10%
10%
10%
50%
50%
50%
55%
55%
55%
50%
LRR - level 3
Alt 3
0%
0%
0%
10%
10%
10%
15%
15%
15%
25%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
-$7
-$7
-$6
-$10
-$10
-$10
-$12
LRR - level 2
TCp
$0
$0
$0
$18
$18
$17
$18
$18
$18
$16
LRR - level 3
TCp
$0
$0
$0
$5
$5
$5
$8
$8
$7
$11
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-228 Costs for Lower Rolling Resistance Steer Tires
Day & Sleeper Cab High Roof Tractors
(2013$/vehicle @ 2 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$47
$46
$45
$45
$44
$41
$41
$40
$40
$39
LRR - level 2
TC
$47
$46
$45
$45
$44
$41
$41
$40
$40
$39
LRR - level 3
TC
$57
$56
$55
$53
$53
$52
$51
$50
$46
$45
LRR - level 4
TC
$62
$60
$59
$57
$56
$55
$53
$53
$52
$51
LRR - level 1
Alt la
70%
70%
70%
70%
70%
70%
70%
70%
70%
70%
LRR - level 2
Alt la
20%
20%
20%
20%
20%
20%
20%
20%
20%
20%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
70%
70%
70%
35%
35%
35%
15%
15%
15%
10%
LRR - level 2
Alt 3
20%
20%
20%
50%
50%
50%
60%
60%
60%
50%
LRR - level 3
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
35%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
-$16
-$15
-$15
-$22
-$22
-$22
-$24
LRR - level 2
TCp
$0
$0
$0
$13
$13
$12
$16
$16
$16
$12
LRR - level 3
TCp
$0
$0
$0
$5
$5
$5
$10
$10
$9
$16
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative
2.11.8.5 Lower RR Drive Tires, Tractors
Table 2-229 Costs for Lower Rolling Resistance Drive Tires
Class 7 Day Cab Low Roof Tractors
(2013$/vehicle @ 4 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 2
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 3
TC
$115
$112
$109
$107
$105
$103
$102
$100
$92
$91
LRR - level 4
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 1
Alt la
50%
50%
50%
50%
50%
50%
50%
50%
50%
50%
LRR - level 2
Alt la
10%
10%
10%
10%
10%
10%
10%
10%
10%
10%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
50%
50%
50%
35%
35%
35%
25%
25%
25%
10%
LRR - level 2
Alt 3
10%
10%
10%
50%
50%
50%
65%
65%
65%
85%
LRR - level 3
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
-$13
-$13
-$12
-$20
-$20
-$20
-$31
LRR - level 2
TCp
$0
$0
$0
$36
$35
$33
$45
$44
$44
$59
LRR - level 3
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-230 Costs for Lower Rolling Resistance Drive Tires
Class 7 Day Cab High Roof Tractors
(2013$/vehicle @ 4 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 2
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR - level 3
TC
$115
$112
$109
$107
$105
$103
$102
$100
$92
$91
LRR - level 4
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR - level 1
Alt la
70%
70%
70%
70%
70%
70%
70%
70%
70%
70%
LRR - level 2
Alt la
20%
20%
20%
20%
20%
20%
20%
20%
20%
20%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
70%
70%
70%
35%
35%
35%
15%
15%
15%
10%
LRR - level 2
Alt 3
20%
20%
20%
50%
50%
50%
60%
60%
60%
50%
LRR - level 3
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
35%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
-$31
-$31
-$29
-$45
-$44
-$44
-$47
LRR - level 2
TCp
$0
$0
$0
$27
$26
$25
$33
$32
$32
$24
LRR - level 3
TCp
$0
$0
$0
$11
$11
$10
$20
$20
$18
$32
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative
Table 2-231 Costs for Lower Rolling Resistance Drive Tires
Class 8 Day Cab Low & Sleeper Cab Low/Medium Roof Tractors
(2013$/vehicle @ 8 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR - level 2
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR - level 3
TC
$230
$224
$219
$214
$210
$207
$204
$200
$184
$181
LRR - level 4
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR - level 1
Alt la
50%
50%
50%
50%
50%
50%
50%
50%
50%
50%
LRR - level 2
Alt la
10%
10%
10%
10%
10%
10%
10%
10%
10%
10%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
50%
50%
50%
35%
35%
35%
25%
25%
25%
10%
LRR - level 2
Alt 3
10%
10%
10%
50%
50%
50%
65%
65%
65%
85%
LRR - level 3
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
-$27
-$26
-$25
-$41
-$40
-$40
-$63
LRR - level 2
TCp
$0
$0
$0
$71
$70
$66
$90
$88
$87
$118
LRR - level 3
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-232 Costs for Lower Rolling Resistance Drive Tires
Class 8 Day & Sleeper Cab High Roof Tractors
(2013$/vehicle @ 8 tires/vehicle)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR - level 1
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR - level 2
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR - level 3
TC
$230
$224
$219
$214
$210
$207
$204
$200
$184
$181
LRR - level 4
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR - level 1
Alt la
70%
70%
70%
70%
70%
70%
70%
70%
70%
70%
LRR - level 2
Alt la
20%
20%
20%
20%
20%
20%
20%
20%
20%
20%
LRR - level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
Alt 3
70%
70%
70%
35%
35%
35%
15%
15%
15%
10%
LRR - level 2
Alt 3
20%
20%
20%
50%
50%
50%
60%
60%
60%
50%
LRR - level 3
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
35%
LRR - level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR - level 1
TCp
$0
$0
$0
-$62
-$61
-$58
-$90
-$88
-$87
-$94
LRR - level 2
TCp
$0
$0
$0
$53
$53
$50
$65
$64
$63
$47
LRR - level 3
TCp
$0
$0
$0
$21
$21
$21
$41
$40
$37
$63
LRR - level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative
2.11.8.6 Lower RR Tires, Trailers
Table 2-233 Costs for Lower Rolling Resistance Tires
Long Van, Full Aero Highway Trailers
(2013$/trailer @ 8 tires/trailer)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR-level 1
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR-level 2
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR-level 3
TC
$230
$224
$219
$214
$210
$207
$204
$200
$184
$181
LRR-level 4
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR-level 5
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR-level 1
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 2
Alt la
10%
10%
10%
10%
10%
10%
10%
10%
10%
10%
LRR-level 3
Alt la
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
LRR-level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 1
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 2
Alt 3
5%
5%
5%
5%
5%
5%
5%
5%
5%
5%
LRR-level 3
Alt 3
95%
95%
95%
0%
0%
0%
0%
0%
0%
0%
LRR-level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 5
Alt 3
0%
0%
0%
95%
95%
95%
95%
95%
95%
95%
LRR-level 1
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR-level 2
TCp
-$9
-$9
-$9
-$9
-$9
-$8
-$8
-$8
-$8
-$8
LRR-level 3
TCp
$11
$11
$11
-$192
-$189
-$186
-$183
-$180
-$166
-$163
LRR-level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR-level 5
TCp
$0
$0
$0
$218
$213
$208
$203
$200
$197
$193
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-234 Costs for Lower Rolling Resistance Tires
Long Van, Partial Aero Highway Trailers
(2013$/trailer @ 8 tires/trailer)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR-level 1
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR-level 2
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR-level 3
TC
$230
$224
$219
$214
$210
$207
$204
$200
$184
$181
LRR-level 4
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR-level 5
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR-level 1
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 2
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR-level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 1
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 2
Alt 3
5%
5%
5%
5%
5%
5%
5%
5%
5%
5%
LRR-level 3
Alt 3
95%
95%
95%
0%
0%
0%
0%
0%
0%
0%
LRR-level 4
Alt 3
0%
0%
0%
95%
95%
95%
95%
95%
95%
95%
LRR-level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 1
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR-level 2
TCp
-$178
-$175
-$172
-$169
-$166
-$157
-$155
-$152
-$151
-$150
LRR-level 3
TCp
$218
$213
$208
$0
$0
$0
$0
$0
$0
$0
LRR-level 4
TCp
$0
$0
$0
$218
$213
$208
$203
$200
$197
$193
LRR-level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative
Table 2-235 Costs for Lower Rolling Resistance Tires
Short Van, Full Aero Highway Trailers
(2013$/trailer @ 4 tires/trailer)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR-level 1
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR-level 2
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR-level 3
TC
$115
$112
$109
$107
$105
$103
$102
$100
$92
$91
LRR-level 4
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR-level 5
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR-level 1
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 2
Alt la
10%
10%
10%
10%
10%
10%
10%
10%
10%
10%
LRR-level 3
Alt la
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
LRR-level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 1
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 2
Alt 3
5%
5%
5%
5%
5%
5%
5%
5%
5%
5%
LRR-level 3
Alt 3
95%
95%
95%
0%
0%
0%
0%
0%
0%
0%
LRR-level 4
Alt 3
0%
0%
0%
95%
95%
95%
0%
0%
0%
0%
LRR-level 5
Alt 3
0%
0%
0%
0%
0%
0%
95%
95%
95%
95%
LRR-level 1
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR-level 2
TCp
-$5
-$5
-$5
-$4
-$4
-$4
-$4
-$4
-$4
-$4
LRR-level 3
TCp
$6
$6
$5
-$96
-$95
-$93
-$92
-$90
-$83
-$82
LRR-level 4
TCp
$0
$0
$0
$109
$107
$104
$0
$0
$0
$0
LRR-level 5
TCp
$0
$0
$0
$0
$0
$0
$101
$100
$98
$97
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-236 Costs for Lower Rolling Resistance Tires
Short Van, Partial Aero Highway Trailers
(2013$/trailer @ 4 tires/trailer)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR-level 1
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR-level 2
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR-level 3
TC
$115
$112
$109
$107
$105
$103
$102
$100
$92
$91
LRR-level 4
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR-level 5
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR-level 1
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 2
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR-level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 1
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 2
Alt 3
5%
5%
5%
5%
5%
5%
5%
5%
5%
5%
LRR-level 3
Alt 3
95%
95%
95%
0%
0%
0%
0%
0%
0%
0%
LRR-level 4
Alt 3
0%
0%
0%
95%
95%
95%
95%
95%
95%
95%
LRR-level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 1
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR-level 2
TCp
-$89
-$88
-$86
-$85
-$83
-$79
-$77
-$76
-$75
-$75
LRR-level 3
TCp
$109
$107
$104
$0
$0
$0
$0
$0
$0
$0
LRR-level 4
TCp
$0
$0
$0
$109
$107
$104
$101
$100
$98
$97
LRR-level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative
Table 2-237 Costs for Lower Rolling Resistance Tires
Long Van, No Aero Highway Trailers
(2013$/trailer @ 8 tires/trailer)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR-level 1
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR-level 2
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR-level 3
TC
$230
$224
$219
$214
$210
$207
$204
$200
$184
$181
LRR-level 4
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR-level 5
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR-level 1
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 2
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR-level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 1
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 2
Alt 3
5%
5%
5%
5%
5%
5%
5%
5%
5%
5%
LRR-level 3
Alt 3
95%
95%
95%
0%
0%
0%
0%
0%
0%
0%
LRR-level 4
Alt 3
0%
0%
0%
95%
95%
95%
95%
95%
95%
95%
LRR-level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 1
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR-level 2
TCp
-$178
-$175
-$172
-$169
-$166
-$157
-$155
-$152
-$151
-$150
LRR-level 3
TCp
$218
$213
$208
$0
$0
$0
$0
$0
$0
$0
LRR-level 4
TCp
$0
$0
$0
$218
$213
$208
$203
$200
$197
$193
LRR-level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-238 Costs for Lower Rolling Resistance Tires
Short Van, No Aero Highway Trailers
(2013$/trailer @ 4 tires/trailer)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR-level 1
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR-level 2
TC
$94
$92
$91
$89
$88
$83
$81
$80
$79
$79
LRR-level 3
TC
$115
$112
$109
$107
$105
$103
$102
$100
$92
$91
LRR-level 4
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR-level 5
TC
$124
$121
$118
$115
$112
$109
$107
$105
$103
$102
LRR-level 1
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 2
Alt la
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
LRR-level 3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 1
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 2
Alt 3
5%
5%
5%
5%
5%
5%
5%
5%
5%
5%
LRR-level 3
Alt 3
95%
95%
95%
0%
0%
0%
0%
0%
0%
0%
LRR-level 4
Alt 3
0%
0%
0%
95%
95%
95%
95%
95%
95%
95%
LRR-level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 1
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR-level 2
TCp
-$89
-$88
-$86
-$85
-$83
-$79
-$77
-$76
-$75
-$75
LRR-level 3
TCp
$109
$107
$104
$0
$0
$0
$0
$0
$0
$0
LRR-level 4
TCp
$0
$0
$0
$109
$107
$104
$101
$100
$98
$97
LRR-level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative
Table 2-239 Costs for Lower Rolling Resistance Tires
Non-Box Highway Trailers
(2013$/trailer @ 8 tires/trailer)
ITEM

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
LRR-level 1
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR-level 2
TC
$187
$184
$181
$178
$175
$166
$163
$160
$159
$157
LRR-level 3
TC
$230
$224
$219
$214
$210
$207
$204
$200
$184
$181
LRR-level 4
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR-level 5
TC
$248
$241
$236
$230
$224
$219
$214
$210
$207
$204
LRR-level 1
Alt la
40%
40%
40%
40%
40%
40%
40%
40%
40%
40%
LRR-level 2
Alt la
30%
30%
30%
30%
30%
30%
30%
30%
30%
30%
LRR-level 3
Alt la
30%
30%
30%
30%
30%
30%
30%
30%
30%
30%
LRR-level 4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 1
Alt 3
5%
5%
5%
5%
5%
5%
5%
5%
5%
5%
LRR-level 2
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 3
Alt 3
95%
95%
95%
95%
95%
95%
95%
95%
95%
95%
LRR-level 4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
LRR-level 1
TCp
-$66
-$65
-$63
-$62
-$61
-$58
-$57
-$56
-$56
-$55
LRR-level 2
TCp
-$56
-$55
-$54
-$53
-$53
-$50
-$49
-$48
-$48
-$47
LRR-level 3
TCp
$149
$146
$142
$139
$137
$135
$132
$130
$120
$118
LRR-level 4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
LRR-level 5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: TC=total cost; TCp=total cost applied to the package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.11.8.7 Lower RR Tires, HD Pickup & Van ($/tire)
We have estimated the costs of lower rolling resistance tires for HD pickups and vans
using the costs used in the 2017-2025 light-duty FRM. In that rule, we estimated the costs of
lower rolling resistance tires level 1 at $5/vehicle including a spare (DMC, 2010$, in all years)
and level 2 at $40/vehicle assuming no spare (DMC, 2010$, in 2021). For HD pickups and vans,
we have scaled upward both of those costs by 50 percent to account for the heavier and larger
HD tires. We consider the level 1 tires to be learned out (curve 1) and the level 2 tires to be on
the steep portion of the curve until 2021 after which it is on the flatter portion of the curve (curve
11). We have applied a low complexity markup to both with short term markups through 2018
for level 1 and through 2024 for level 2. With the exception of the 50 percent scaling factor, all
LRR tire costs for HD pickups and vans are identical to the 2017-2025 light-duty FRM. The
resultant costs are presented below.
Table 2-240 Costs for Lower Rolling Resistance Tires
HD Pickups & Vans
(2012$ @ 4 tires/vehicle)
ITEM

2021
2022
2023
2024
2025
2026
2027
LRR - level 1
DMC
$8
$8
$8
$8
$8
$8
$8
LRR - level 2
DMC
$63
$61
$59
$58
$56
$54
$53
LRR - level 1
IC
$2
$2
$2
$2
$2
$2
$2
LRR - level 2
IC
$15
$15
$15
$15
$12
$12
$12
LRR - level 1
TC
$10
$10
$10
$10
$10
$10
$10
LRR - level 2
TC
$78
$76
$74
$73
$68
$66
$65
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
2.11.8.8 Automatic Tire Inflation Systems (ATIS)
For tractors, we have estimated the cost of ATIS technology based on an estimate from
TetraTech of $1143 (retail, 2013$). Using that estimate we divided by a 1.36 RPE to arrive at a
cost of $840 (DMC, 2013$, in 2018). We consider this technology to be on the flat portion of
the learning curve (curve 12) and have applied a low complexity ICM with short term markups
through 2022. The resultant technology costs, penetration rates and total cost applied to the
package are shown below for tractors.
Table 2-241 Costs for Automatic Tire Inflation Systems
Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
ATIS
DMC
$840
$815
$790
$767
$751
$736
$722
$707
$693
$679
ATIS
IC
$150
$150
$149
$149
$149
$117
$117
$117
$117
$117
ATIS
TC
$990
$964
$940
$916
$900
$853
$839
$824
$810
$796
ATIS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
ATIS
Alt 3
0%
0%
0%
20%
20%
20%
25%
25%
25%
30%
ATIS
TCp
$0
$0
$0
$183
$180
$171
$210
$206
$202
$239
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
For trailers, we have estimated the cost of this technology based on an estimate from
TetraTech of $800 (retail, 2013$). We consider this estimate to be valid for all trailers except
short vans. For short vans, we have used an estimate of $600 (retail, 2013$) since they have just
one axle. Using these estimates we divided by a 1.36 RPE to arrive at a cost of $588 (DMC,
2013$, in 2018) for all but short vans and $441 (DMC, 2013$, in 2018) for short vans. We
consider this technology to be on the flat portion of the learning curve (curve 12) and have
applied a low complexity ICM with short term markups through 2022. The resultant technology
costs, penetration rates and total cost applied to the package are shown below for trailers.
Table 2-242 Costs for Automatic Tire Inflation Systems
Long Van, Full Aero Trailers (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
ATIS
DMC
$588
$571
$553
$537
$526
$516
$505
$495
$485
$476
ATIS
IC
$105
$105
$105
$104
$104
$82
$82
$82
$82
$82
ATIS
TC
$693
$675
$658
$641
$630
$598
$587
$577
$567
$557
ATIS
Alt la
45%
45%
45%
45%
45%
45%
45%
45%
45%
45%
ATIS
Alt 3
95%
95%
95%
95%
95%
95%
95%
95%
95%
95%
ATIS
TCp
$347
$338
$329
$321
$315
$299
$294
$289
$284
$279
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-243 Costs for Automatic Tire Inflation Systems
Long Van, Partial Aero Trailers (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
ATIS
DMC
$588
$571
$553
$537
$526
$516
$505
$495
$485
$476
ATIS
IC
$105
$105
$105
$104
$104
$82
$82
$82
$82
$82
ATIS
TC
$693
$675
$658
$641
$630
$598
$587
$577
$567
$557
ATIS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
ATIS
Alt 3
95%
95%
95%
95%
95%
95%
95%
95%
95%
95%
ATIS
TCp
$659
$642
$625
$609
$599
$568
$558
$548
$539
$529
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-244 Costs for Automatic Tire Inflation Systems
Short Van, Full Aero Trailers (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
ATIS
DMC
$441
$428
$415
$403
$395
$387
$379
$371
$364
$357
ATIS
IC
$79
$79
$78
$78
$78
$61
$61
$61
$61
$61
ATIS
TC
$520
$506
$493
$481
$473
$448
$440
$433
$425
$418
ATIS
Alt la
30%
30%
30%
30%
30%
30%
30%
30%
30%
30%
ATIS
Alt 3
95%
95%
95%
95%
95%
95%
95%
95%
95%
95%
ATIS
TCp
$338
$329
$321
$313
$307
$291
$286
$281
$276
$272
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-245 Costs for Automatic Tire Inflation Systems
Short Van, Partial Aero Trailers (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
ATIS
DMC
$441
$428
$415
$403
$395
$387
$379
$371
$364
$357
ATIS
IC
$79
$79
$78
$78
$78
$61
$61
$61
$61
$61
ATIS
TC
$520
$506
$493
$481
$473
$448
$440
$433
$425
$418
ATIS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
ATIS
Alt 3
95%
95%
95%
95%
95%
95%
95%
95%
95%
95%
ATIS
TCp
$494
$481
$469
$457
$449
$426
$418
$411
$404
$397
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
2.11.8.9 Tire Pressure Monitoring System (TPMS)
We have estimated the cost of TPMS technology based on price data from Ryder.209
These price data showed a price of $94/pair of tire pressure monitoring sensors along with a
price of $65 for a repeater. Using these values as DMCs in 2013$ and applicable in 2018, we
have costed 10 sensors per class 8 tractor, 6 per class 7 tractor, 10 sensors per heavy HD
vocational vehicle, 6 per light and medium HD vocational vehicle, 8 per long van and non-box
trailer, and 4 per short van trailer. We have also included a $65 repeater for all tractors. We
consider this technology to be on the flat portion of the learning curve (curve 12) and have
applied a low complexity ICM with short term markups through 2022. The resultant technology
costs, penetration rates and total cost applied to the package are shown in the tables below.
Table 2-246 Costs for Tire Pressure Monitoring Systems (TPMS)
Vocational Light/Medium HD Urban Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
TPMS
DMC
$282
$274
$265
$257
$252
$247
$242
$237
$233
$228
TPMS
IC
$50
$50
$50
$50
$50
$39
$39
$39
$39
$39
TPMS
TC
$332
$324
$315
$307
$302
$286
$281
$277
$272
$267
TPMS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
TPMS
Alt 3
0%
0%
0%
40%
40%
40%
55%
55%
55%
70%
TPMS
TCp
$0
$0
$0
$123
$121
$115
$155
$152
$150
$187
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-247 Costs for Tire Pressure Monitoring Systems (TPMS)
Vocational Light/Medium HD Multipurpose Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
TPMS
DMC
$282
$274
$265
$257
$252
$247
$242
$237
$233
$228
TPMS
IC
$50
$50
$50
$50
$50
$39
$39
$39
$39
$39
TPMS
TC
$332
$324
$315
$307
$302
$286
$281
$277
$272
$267
TPMS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
TPMS
Alt 3
0%
0%
0%
50%
50%
50%
65%
65%
65%
80%
TPMS
TCp
$0
$0
$0
$154
$151
$143
$183
$180
$177
$214
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-248 Costs for Tire Pressure Monitoring Systems (TPMS)
Vocational Light/Medium HD Regional Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
TPMS
DMC
$282
$274
$265
$257
$252
$247
$242
$237
$233
$228
TPMS
IC
$50
$50
$50
$50
$50
$39
$39
$39
$39
$39
TPMS
TC
$332
$324
$315
$307
$302
$286
$281
$277
$272
$267
TPMS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
TPMS
Alt 3
0%
0%
0%
60%
60%
60%
75%
75%
75%
90%
TPMS
TCp
$0
$0
$0
$184
$181
$172
$211
$207
$204
$240
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-249 Costs for Tire Pressure Monitoring Systems (TPMS)
Vocational Heavy HD Urban Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
TPMS
DMC
$535
$519
$503
$488
$479
$469
$460
$450
$441
$433
TPMS
IC
$95
$95
$95
$95
$95
$75
$74
$74
$74
$74
TPMS
TC
$630
$614
$598
$583
$573
$543
$534
$525
$516
$507
TPMS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
TPMS
Alt 3
0%
0%
0%
40%
40%
40%
55%
55%
55%
70%
TPMS
TCp
$0
$0
$0
$233
$229
$217
$294
$289
$284
$355
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-250 Costs for Tire Pressure Monitoring Systems (TPMS)
Vocational Heavy HD Multipurpose Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
TPMS
DMC
$535
$519
$503
$488
$479
$469
$460
$450
$441
$433
TPMS
IC
$95
$95
$95
$95
$95
$75
$74
$74
$74
$74
TPMS
TC
$630
$614
$598
$583
$573
$543
$534
$525
$516
$507
TPMS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
TPMS
Alt 3
0%
0%
0%
50%
50%
50%
65%
65%
65%
80%
TPMS
TCp
$0
$0
$0
$292
$287
$272
$347
$341
$335
$405
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-251 Costs for Tire Pressure Monitoring Systems (TPMS)
Vocational Heavy HD Regional Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
TPMS
DMC
$535
$519
$503
$488
$479
$469
$460
$450
$441
$433
TPMS
IC
$95
$95
$95
$95
$95
$75
$74
$74
$74
$74
TPMS
TC
$630
$614
$598
$583
$573
$543
$534
$525
$516
$507
TPMS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
TPMS
Alt 3
0%
0%
0%
60%
60%
60%
75%
75%
75%
90%
TPMS
TCp
$0
$0
$0
$350
$344
$326
$401
$394
$387
$456
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-252 Costs for Tire Pressure Monitoring Systems (TPMS)
Class 7 Day Cab Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
TPMS
DMC
$347
$337
$326
$317
$310
$304
$298
$292
$286
$281
TPMS
IC
$62
$62
$62
$62
$61
$48
$48
$48
$48
$48
TPMS
TC
$409
$398
$388
$378
$372
$352
$346
$340
$334
$329
TPMS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
TPMS
Alt 3
0%
0%
0%
20%
20%
20%
50%
50%
50%
70%
TPMS
TCp
$0
$0
$0
$76
$74
$70
$173
$170
$167
$230
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-253 Costs for Tire Pressure Monitoring Systems (TPMS)
Class 8 Day & Sleeper Cab Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
TPMS
DMC
$535
$519
$503
$488
$479
$469
$460
$450
$441
$433
TPMS
IC
$95
$95
$95
$95
$95
$75
$74
$74
$74
$74
TPMS
TC
$630
$614
$598
$583
$573
$543
$534
$525
$516
$507
TPMS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
TPMS
Alt 3
0%
0%
0%
20%
20%
20%
50%
50%
50%
70%
TPMS
TCp
$0
$0
$0
$117
$115
$109
$267
$262
$258
$355
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-254 Costs for Tire Pressure Monitoring Systems (TPMS)
Long Van, No Aero and Non-Box Trailers (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
TPMS
DMC
$376
$365
$354
$343
$336
$330
$323
$317
$310
$304
TPMS
IC
$67
$67
$67
$67
$67
$52
$52
$52
$52
$52
TPMS
TC
$443
$432
$421
$410
$403
$382
$375
$369
$362
$356
TPMS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
TPMS
Alt 3
95%
95%
95%
95%
95%
95%
95%
95%
95%
95%
TPMS
TCp
$421
$410
$400
$389
$383
$363
$357
$350
$344
$338
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative
Table 2-255 Costs for Tire Pressure Monitoring Systems (TPMS)
Short Van, No Aero Trailers (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
TPMS
DMC
$188
$182
$177
$172
$168
$165
$161
$158
$155
$152
TPMS
IC
$34
$33
$33
$33
$33
$26
$26
$26
$26
$26
TPMS
TC
$222
$216
$210
$205
$201
$191
$188
$184
$181
$178
TPMS
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
TPMS
Alt 3
95%
95%
95%
95%
95%
95%
95%
95%
95%
95%
TPMS
TCp
$210
$205
$200
$195
$191
$181
$178
$175
$172
$169
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the
package; alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.11.9 Aerodynamic Improvements (Aero)
The agencies' estimates for cost of tractor aero features are based the work done by ICF
in support of the Phase 1 HD rules. For trailers, we have based our estimates on the work
presented in the ICCT trailer technology report.210
2.11.9.1 Aero Improvements, Day Cab Low Roof Tractors
For low roof day cab tractors, Aero Bin 2 costs are estimated at $1020, Bin 3 at $2059
and Bin 4 at $2625 (all are DMC, in 2013$, and applicable in 2014). We consider Bin 2
technologies to be beyond the effects of learning (curve 1), Bin 3 technologies to be on the flat
portion of the curve (curve 2) and Bin 4 technologies to be on the steep portion of the curve
(curve 4). We have applied a low complexity ICMs to each with short term markups through
2022. The resultant technology costs, penetration rates and total cost applied to the package are
shown below.
Table 2-256 Costs of Aero Technologies
Day Cab Low Roof Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Aero Bin2
DMC
$1,020
$1,020
$1,020
$1,020
$1,020
$1,020
$1,020
$1,020
$1,020
$1,020
Aero Bin3
DMC
$1,823
$1,787
$1,751
$1,716
$1,681
$1,648
$1,615
$1,583
$1,567
$1,551
Aero Bint
DMC
$1,680
$1,630
$1,581
$1,534
$1,488
$1,443
$1,400
$1,358
$1,331
$1,304
Aero Bin2
IC
$182
$182
$182
$182
$182
$143
$143
$143
$143
$143
Aero Bin3
IC
$364
$364
$363
$363
$362
$286
$286
$285
$285
$285
Aero Bin4
IC
$456
$455
$455
$454
$454
$360
$360
$360
$360
$360
Aero Bin2
TC
$1,201
$1,201
$1,201
$1,201
$1,201
$1,162
$1,162
$1,162
$1,162
$1,162
Aero Bin3
TC
$2,187
$2,150
$2,114
$2,079
$2,044
$1,934
$1,901
$1,868
$1,852
$1,836
Aero Bint
TC
$2,136
$2,085
$2,036
$1,988
$1,941
$1,803
$1,760
$1,718
$1,690
$1,663
Aero Bin2
Alt la
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
Aero Bin3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bint
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin2
Alt 3
90%
90%
90%
95%
95%
95%
80%
80%
80%
50%
Aero Bin3
Alt 3
0%
0%
0%
5%
5%
5%
20%
20%
20%
50%
Aero Bint
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin2
TCp
$0
$0
$0
$60
$60
$58
-$116
-$116
-$116
-$465
Aero Bin3
TCp
$0
$0
$0
$104
$102
$97
$380
$374
$370
$918
Aero Bin4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.9.2 Aero Improvements, Day Cab High Roof Tractors
For high roof day cab tractors, Aero Bin 3 costs are estimated at $1046, Bin 4 at $2086,
Bin 5 at $2660, Bin 6 at $3234 and Bin 7 at $3807 (all are DMC, in 2013$, and applicable in
2014; note that the table below makes clear that we do not project use of aero improvements
above Bin 5). We consider Bin 3 technologies to be on the flat portion of the curve (curve 2) and
Bin 4 through 7 technologies to be on the steep portion of the curve (curve 4). We have applied
a low complexity ICMs to Bins 3 and 4 with short term markups through 2022. We have applied
medium complexity ICMs to Bins 5 through 7 with short term markups through 2025. The

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
resultant technology costs, penetration rates and total cost applied to the package are shown
below.
Table 2-257 Costs of Aero Technologies
Day Cab High Roof Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Aero Bin3
DMC
$926
$908
$890
$872
$854
$837
$821
$804
$796
$788
Aero Bin4
DMC
$1,335
$1,295
$1,256
$1,219
$1,182
$1,147
$1,112
$1,079
$1,057
$1,036
Aero Bin5
DMC
$1,702
$1,651
$1,602
$1,554
$1,507
$1,462
$1,418
$1,375
$1,348
$1,321
Aero Bin6
DMC
$2,069
$2,007
$1,947
$1,889
$1,832
$1,777
$1,724
$1,672
$1,639
$1,606
Aero Bin7
DMC
$2,437
$2,364
$2,293
$2,224
$2,157
$2,092
$2,030
$1,969
$1,929
$1,891
Aero Bin3
IC
$185
$185
$185
$184
$184
$145
$145
$145
$145
$145
Aero Bin4
IC
$362
$362
$361
$361
$360
$286
$286
$286
$286
$286
Aero Bin5
IC
$756
$753
$750
$748
$746
$743
$741
$739
$554
$553
Aero Bin6
IC
$919
$915
$912
$909
$907
$904
$901
$898
$673
$672
Aero Bin7
IC
$1,082
$1,078
$1,074
$1,071
$1,067
$1,064
$1,061
$1,058
$793
$792
Aero Bin3
TC
$1,112
$1,093
$1,074
$1,056
$1,039
$983
$966
$949
$941
$933
Aero Bin4
TC
$1,697
$1,657
$1,618
$1,579
$1,542
$1,433
$1,398
$1,365
$1,343
$1,322
Aero Bin5
TC
$2,458
$2,404
$2,352
$2,302
$2,253
$2,205
$2,159
$2,114
$1,902
$1,874
Aero Bin6
TC
$2,988
$2,923
$2,860
$2,798
$2,739
$2,681
$2,625
$2,571
$2,312
$2,278
Aero Bin7
TC
$3,518
$3,441
$3,367
$3,295
$3,224
$3,156
$3,091
$3,027
$2,722
$2,682
Aero Bin3
Alt la
80%
80%
80%
80%
80%
80%
80%
80%
80%
80%
Aero Bin4
Alt la
10%
10%
10%
10%
10%
10%
10%
10%
10%
10%
Aero Bin5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin6
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin7
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin3
Alt 3
80%
80%
80%
60%
60%
60%
40%
40%
40%
30%
Aero Bin4
Alt 3
10%
10%
10%
35%
35%
35%
40%
40%
40%
30%
Aero Bin5
Alt 3
0%
0%
0%
5%
5%
5%
20%
20%
20%
40%
Aero Bin6
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin7
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin3
TCp
$0
$0
$0
-$211
-$208
-$197
-$386
-$380
-$376
-$467
Aero Bin4
TCp
$0
$0
$0
$395
$386
$358
$419
$409
$403
$264
Aero Bin5
TCp
$0
$0
$0
$115
$113
$110
$432
$423
$380
$750
Aero Bin6
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin7
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.9.3 Aero Improvements, Sleeper Cab Low/Mid Roof Tractors
For low and mid roof sleeper cab tractors, Aero Bin 2 costs are estimated at $1244, Bin 3
at $2356 and Bin 4 at $3003 (all are DMC, in 2013$, and applicable in 2014). We consider Bin
2 technologies to be beyond the effects of learning (curve 1), Bin 3 technologies to be on the flat
portion of the curve (curve 2) and Bin 4 technologies to be on the steep portion of the curve
(curve 4). We have applied a low complexity ICMs to each with short term markups through
2022. The resultant technology costs, penetration rates and total cost applied to the package are
shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-258 Costs of Aero Technologies
Sleeper Cab Low/Mid Roof Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Aero Bin2
DMC
$1,244
$1,244
$1,244
$1,244
$1,244
$1,244
$1,244
$1,244
$1,244
$1,244
Aero Bin3
DMC
$2,085
$2,044
$2,003
$1,963
$1,923
$1,885
$1,847
$1,810
$1,792
$1,774
Aero Bin4
DMC
$1,922
$1,864
$1,808
$1,754
$1,702
$1,651
$1,601
$1,553
$1,522
$1,492
Aero Bin2
IC
$222
$222
$222
$222
$222
$174
$174
$174
$174
$174
Aero Bin3
IC
$417
$416
$416
$415
$415
$327
$327
$327
$326
$326
Aero Bin4
IC
$522
$521
$520
$519
$519
$412
$412
$412
$412
$411
Aero Bin2
TC
$1,466
$1,466
$1,466
$1,466
$1,466
$1,419
$1,419
$1,419
$1,419
$1,419
Aero Bin3
TC
$2,502
$2,460
$2,418
$2,378
$2,338
$2,212
$2,174
$2,137
$2,119
$2,101
Aero Bin4
TC
$2,444
$2,385
$2,329
$2,274
$2,220
$2,063
$2,013
$1,965
$1,933
$1,903
Aero Bin2
Alt la
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
Aero Bin3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin2
Alt 3
90%
90%
90%
90%
90%
90%
90%
90%
90%
60%
Aero Bin3
Alt 3
0%
0%
0%
5%
5%
5%
10%
10%
10%
40%
Aero Bin4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin2
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
-$426
Aero Bin3
TCp
$0
$0
$0
$119
$117
$111
$217
$214
$212
$840
Aero Bin4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.9.4 Aero Improvements, Sleeper Cab High Roof Tractors
For high roof sleeper cab tractors, Aero Bin 3 costs are estimated at $1413, Bin 4 at
$2423, Bin 5 at $3089, Bin 6 at $3755 and Bin 7 at $4422 (all are DMC, in 2013$, and
applicable in 2014; note that the table below makes clear that we do not project use of aero
improvements above Bin 5). We consider Bin 3 technologies to be on the flat portion of the
curve (curve 2) and Bin 4 through 7 technologies to be on the steep portion of the curve (curve
4). We have applied a low complexity ICMs to Bins 3 and 4 with short term markups through
2022. We have applied medium complexity ICMs to Bins 5 through 7 with short term markups
through 2025. The resultant technology costs, penetration rates and total cost applied to the
package are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-259 Costs of Aero Technologies
Sleeper Cab High Roof Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Aero Bin3
DMC
$1,251
$1,226
$1,201
$1,177
$1,154
$1,131
$1,108
$1,086
$1,075
$1,064
Aero Bin4
DMC
$1,551
$1,504
$1,459
$1,415
$1,373
$1,332
$1,292
$1,253
$1,228
$1,203
Aero Bin5
DMC
$1,977
$1,918
$1,860
$1,804
$1,750
$1,698
$1,647
$1,597
$1,565
$1,534
Aero Bin6
DMC
$2,403
$2,331
$2,261
$2,194
$2,128
$2,064
$2,002
$1,942
$1,903
$1,865
Aero Bin7
DMC
$2,830
$2,745
$2,663
$2,583
$2,505
$2,430
$2,357
$2,286
$2,241
$2,196
Aero Bin3
IC
$250
$250
$249
$249
$249
$196
$196
$196
$196
$196
Aero Bin4
IC
$421
$420
$420
$419
$418
$333
$332
$332
$332
$332
Aero Bin5
IC
$878
$875
$872
$869
$866
$863
$861
$858
$643
$642
Aero Bin6
IC
$1,067
$1,063
$1,060
$1,056
$1,053
$1,050
$1,046
$1,043
$782
$781
Aero Bin7
IC
$1,256
$1,252
$1,248
$1,244
$1,240
$1,236
$1,232
$1,229
$921
$919
Aero Bin3
TC
$1,501
$1,475
$1,450
$1,426
$1,402
$1,327
$1,304
$1,282
$1,271
$1,260
Aero Bin4
TC
$1,971
$1,924
$1,879
$1,834
$1,791
$1,664
$1,624
$1,585
$1,560
$1,535
Aero Bin5
TC
$2,855
$2,792
$2,732
$2,673
$2,616
$2,561
$2,508
$2,456
$2,209
$2,176
Aero Bin6
TC
$3,470
$3,395
$3,321
$3,250
$3,181
$3,114
$3,048
$2,985
$2,685
$2,646
Aero Bin7
TC
$4,086
$3,997
$3,910
$3,826
$3,745
$3,666
$3,589
$3,515
$3,162
$3,115
Aero Bin3
Alt la
80%
80%
80%
80%
80%
80%
80%
80%
80%
80%
Aero Bin4
Alt la
10%
10%
10%
10%
10%
10%
10%
10%
10%
10%
Aero Bin5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin6
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin7
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin3
Alt 3
80%
80%
80%
60%
60%
60%
40%
40%
40%
20%
Aero Bin4
Alt 3
10%
10%
10%
30%
30%
30%
40%
40%
40%
30%
Aero Bin5
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
50%
Aero Bin6
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin7
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin3
TCp
$0
$0
$0
-$285
-$280
-$265
-$522
-$513
-$508
-$756
Aero Bin4
TCp
$0
$0
$0
$367
$358
$333
$487
$476
$468
$307
Aero Bin5
TCp
$0
$0
$0
$267
$262
$256
$502
$491
$442
$1,088
Aero Bin6
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin7
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.9.5 Aero Improvements, Trailers
For long van trailers, Aero Bin 3 costs are based on and ICCT estimate of $700 (retail,
2013$), Bin 4 costs are based on an ICCT estimate of $1000 (retail, 2013$), Bin 5 costs are
based on an ICCT estimate of $1600 (retail, 2013$), Bin 6 costs are based on an ICCT estimate
of $1900 (retail, 2013$), Bin 7 costs are based on an ICCT estimate of $2200 (retail, 2013$), and
Bin 8 costs are based on an ICCT estimate of $2900 (retail, 2013$). We have used these costs
and divided by a 1.36 RPE to arrive at direct manufacturing costs of $515, $735, $1176, $1397,
$1617 and $2132 for Bins 3 through 8, respectively (all are DMC, in 2013$, applicable in 2014).
We consider each of these technologies to be on the flat portion of the learning curve (curve 2)
and have applied low complexity ICMs with short term markups through 2018. The resultant
technology costs, penetration rates and total cost applied to the package are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-260 Costs of Aero Technologies
Long Van, Full Aero Trailers (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Aero Bin3
DMC
$456
$447
$438
$429
$420
$412
$404
$396
$392
$388
Aero Bin4
DMC
$651
$638
$625
$613
$600
$588
$577
$565
$559
$554
Aero Bin5
DMC
$1,042
$1,021
$1,000
$980
$961
$941
$923
$904
$895
$886
Aero Bin6
DMC
$1,237
$1,212
$1,188
$1,164
$1,141
$1,118
$1,096
$1,074
$1,063
$1,052
Aero Bin7
DMC
$1,432
$1,403
$1,375
$1,348
$1,321
$1,294
$1,269
$1,243
$1,231
$1,218
Aero Bin8
DMC
$1,888
$1,850
$1,813
$1,777
$1,741
$1,706
$1,672
$1,639
$1,622
$1,606
Aero Bin3
IC
$91
$72
$72
$72
$72
$71
$71
$71
$71
$71
Aero Bin4
IC
$130
$102
$102
$102
$102
$102
$102
$102
$102
$102
Aero Bin5
IC
$208
$164
$164
$164
$163
$163
$163
$163
$163
$163
Aero Bin6
IC
$247
$195
$194
$194
$194
$194
$194
$194
$194
$194
Aero Bin7
IC
$286
$225
$225
$225
$225
$225
$224
$224
$224
$224
Aero Bin8
IC
$377
$297
$297
$296
$296
$296
$296
$296
$296
$295
Aero Bin3
TC
$547
$518
$509
$500
$492
$483
$475
$467
$463
$459
Aero Bin4
TC
$781
$740
$727
$715
$703
$690
$679
$667
$661
$656
Aero Bin5
TC
$1,250
$1,185
$1,164
$1,144
$1,124
$1,105
$1,086
$1,067
$1,058
$1,049
Aero Bin6
TC
$1,484
$1,407
$1,382
$1,358
$1,335
$1,312
$1,289
$1,267
$1,257
$1,246
Aero Bin7
TC
$1,718
$1,629
$1,600
$1,573
$1,546
$1,519
$1,493
$1,467
$1,455
$1,443
Aero Bin8
TC
$2,265
$2,147
$2,110
$2,073
$2,037
$2,002
$1,968
$1,934
$1,918
$1,902
Aero Bin3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin4
Alt la
40%
40%
40%
40%
40%
40%
40%
40%
40%
40%
Aero Bin5
Alt la
5%
5%
5%
5%
5%
5%
5%
5%
5%
5%
Aero Bin6
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin7
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin8
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin3
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin4
Alt 3
95%
95%
95%
0%
0%
0%
0%
0%
0%
0%
Aero Bin5
Alt 3
0%
0%
0%
95%
95%
95%
0%
0%
0%
0%
Aero Bin6
Alt 3
0%
0%
0%
0%
0%
0%
95%
95%
95%
30%
Aero Bin7
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
70%
Aero Bin8
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin3
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin4
TCp
$430
$407
$400
-$286
-$281
-$276
-$271
-$267
-$265
-$262
Aero Bin5
TCp
-$62
-$59
-$58
$1,029
$1,012
$994
-$54
-$53
-$53
-$52
Aero Bin6
TCp
$0
$0
$0
$0
$0
$0
$1,225
$1,204
$1,194
$374
Aero Bin7
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$1,010
Aero Bin8
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-261 Costs of Aero Technologies
Long Van, Partial Aero Trailers (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Aero Bin3
DMC
$456
$447
$438
$429
$420
$412
$404
$396
$392
$388
Aero Bin4
DMC
$651
$638
$625
$613
$600
$588
$577
$565
$559
$554
Aero Bin5
DMC
$1,042
$1,021
$1,000
$980
$961
$941
$923
$904
$895
$886
Aero Bin6
DMC
$1,237
$1,212
$1,188
$1,164
$1,141
$1,118
$1,096
$1,074
$1,063
$1,052
Aero Bin7
DMC
$1,432
$1,403
$1,375
$1,348
$1,321
$1,294
$1,269
$1,243
$1,231
$1,218
Aero Bin8
DMC
$1,888
$1,850
$1,813
$1,777
$1,741
$1,706
$1,672
$1,639
$1,622
$1,606
Aero Bin3
IC
$91
$72
$72
$72
$72
$71
$71
$71
$71
$71
Aero Bin4
IC
$130
$102
$102
$102
$102
$102
$102
$102
$102
$102
Aero Bin5
IC
$208
$164
$164
$164
$163
$163
$163
$163
$163
$163
Aero Bin6
IC
$247
$195
$194
$194
$194
$194
$194
$194
$194
$194
Aero Bin7
IC
$286
$225
$225
$225
$225
$225
$224
$224
$224
$224
Aero Bin8
IC
$377
$297
$297
$296
$296
$296
$296
$296
$296
$295
Aero Bin3
TC
$547
$518
$509
$500
$492
$483
$475
$467
$463
$459
Aero Bin4
TC
$781
$740
$727
$715
$703
$690
$679
$667
$661
$656
Aero Bin5
TC
$1,250
$1,185
$1,164
$1,144
$1,124
$1,105
$1,086
$1,067
$1,058
$1,049
Aero Bin6
TC
$1,484
$1,407
$1,382
$1,358
$1,335
$1,312
$1,289
$1,267
$1,257
$1,246
Aero Bin7
TC
$1,718
$1,629
$1,600
$1,573
$1,546
$1,519
$1,493
$1,467
$1,455
$1,443
Aero Bin8
TC
$2,265
$2,147
$2,110
$2,073
$2,037
$2,002
$1,968
$1,934
$1,918
$1,902
Aero Bin3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin6
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin7
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin8
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin3
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin4
Alt 3
95%
95%
95%
95%
95%
95%
95%
95%
95%
95%
Aero Bin5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin6
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin7
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin8
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin3
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin4
TCp
$742
$703
$691
$679
$667
$656
$645
$634
$628
$623
Aero Bin5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin6
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin7
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin8
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-262 Costs of Aero Technologies
Short Van, Full Aero Trailers (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Aero Bin2
DMC
$456
$447
$438
$429
$420
$412
$404
$396
$392
$388
Aero Bin3
DMC
$911
$893
$875
$858
$841
$824
$807
$791
$783
$775
Aero Bin4
DMC
$1,107
$1,084
$1,063
$1,042
$1,021
$1,000
$980
$961
$951
$942
Aero Bin5
DMC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin6
DMC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin7
DMC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin8
DMC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin2
IC
$91
$72
$72
$72
$72
$71
$71
$71
$71
$71
Aero Bin3
IC
$182
$143
$143
$143
$143
$143
$143
$143
$143
$143
Aero Bin4
IC
$221
$174
$174
$174
$174
$174
$173
$173
$173
$173
Aero Bin5
IC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin6
IC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin7
IC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin8
IC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin2
TC
$547
$518
$509
$500
$492
$483
$475
$467
$463
$459
Aero Bin3
TC
$1,093
$1,036
$1,018
$1,001
$984
$967
$950
$934
$926
$918
Aero Bin4
TC
$1,328
$1,259
$1,237
$1,215
$1,194
$1,174
$1,154
$1,134
$1,124
$1,115
Aero Bin5
TC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin6
TC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin7
TC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin8
TC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin2
Alt la
5%
5%
5%
5%
5%
5%
5%
5%
5%
5%
Aero Bin3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin6
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin7
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin8
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin2
Alt 3
5%
5%
5%
95%
95%
95%
0%
0%
0%
0%
Aero Bin3
Alt 3
0%
0%
0%
0%
0%
0%
95%
95%
95%
30%
Aero Bin4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
60%
Aero Bin5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
10%
Aero Bin6
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin7
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin8
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin2
TCp
$0
$0
$0
$450
$443
$435
-$24
-$23
-$23
-$23
Aero Bin3
TCp
$0
$0
$0
$0
$0
$0
$903
$887
$880
$275
Aero Bin4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$669
Aero Bin5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin6
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin7
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin8
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-263 Costs of Aero Technologies
Short Van, Partial Aero Trailers (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Aero Bin2
DMC
$456
$447
$438
$429
$420
$412
$404
$396
$392
$388
Aero Bin3
DMC
$911
$893
$875
$858
$841
$824
$807
$791
$783
$775
Aero Bin4
DMC
$1,107
$1,084
$1,063
$1,042
$1,021
$1,000
$980
$961
$951
$942
Aero Bin5
DMC
$1,107
$1,084
$1,063
$1,042
$1,021
$1,000
$980
$961
$951
$942
Aero Bin6
DMC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin7
DMC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin8
DMC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin2
IC
$91
$72
$72
$72
$72
$71
$71
$71
$71
$71
Aero Bin3
IC
$182
$143
$143
$143
$143
$143
$143
$143
$143
$143
Aero Bin4
IC
$221
$174
$174
$174
$174
$174
$173
$173
$173
$173
Aero Bin5
IC
$221
$174
$174
$174
$174
$174
$173
$173
$173
$173
Aero Bin6
IC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin7
IC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin8
IC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin2
TC
$547
$518
$509
$500
$492
$483
$475
$467
$463
$459
Aero Bin3
TC
$1,093
$1,036
$1,018
$1,001
$984
$967
$950
$934
$926
$918
Aero Bin4
TC
$1,328
$1,259
$1,237
$1,215
$1,194
$1,174
$1,154
$1,134
$1,124
$1,115
Aero Bin5
TC
$1,328
$1,259
$1,237
$1,215
$1,194
$1,174
$1,154
$1,134
$1,124
$1,115
Aero Bin6
TC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin7
TC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin8
TC
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin2
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin3
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin4
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin5
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin6
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin7
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin8
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin2
Alt 3
0%
0%
0%
95%
95%
95%
95%
95%
95%
95%
Aero Bin3
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin4
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin5
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin6
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin7
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin8
Alt 3
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Aero Bin2
TCp
$0
$0
$0
$475
$467
$459
$451
$444
$440
$436
Aero Bin3
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin4
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin5
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin6
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin7
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Aero Bin8
TCp
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.11.9.6 Aero Improvements, HD Pickups and Vans
For HD pickups and vans, we have based our aero improvement costs on values used in
our light-duty 2017-2025 final rule. Using those values updated to 2012$ results in costs for
aero 1 (passive aero treatments) and active aero treatments of $47 and $142 (both are DMC, in
2012$, in 2015). Note that the aero 2 costs are the passive aero 1 plus the active aero costs. We
consider both of these technologies to be on the flat portion of the learning curve (curve 8) and,
to aero 1, have applied low complexity markups with near term markups through 2018 and, to
active aero, and have applied medium complexity markups with near term markups through
2024. The resultant costs for HD pickups and vans are shown below for aero 1 and active aero
and then for aero 2 (the two combined, passive+active aero).
Table 2-264 Costs for Passive Aero Treatments - Aero 1
Gasoline & Diesel HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Aero 1 - passive aero
DMC
$42
$41
$40
$39
$38
$38
$38
Aero 1 - passive aero
IC
$9
$9
$9
$9
$9
$9
$9
Aero 1 - passive aero
TC
$51
$50
$49
$48
$47
$47
$47
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
Table 2-265 Costs for Active Aero Treatments
Gasoline & Diesel HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Aero 2 - active aero
DMC
$125
$122
$120
$118
$115
$114
$113
Aero 2 - active aero
IC
$54
$54
$54
$54
$40
$40
$40
Aero 2 - active aero
TC
$179
$177
$174
$172
$156
$154
$153
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
Table 2-266 Costs for Aero 2 (passive plus active aero)
Gasoline & Diesel HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Aero 2 - active aero
DMC
$166
$163
$160
$157
$154
$152
$151
Aero 2 - active aero
IC
$63
$63
$63
$63
$50
$49
$49
Aero 2 - active aero
TC
$230
$227
$223
$220
$203
$201
$200
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
2.11.10 Other Technologies
2.11.10.1 Advanced Cruise Controls, Tractors
We have estimated the cost of this technology based on an estimate from TetraTech of
$1100 (retail, 2013$). Using that estimate we divided by a 1.36 RPE to arrive at a cost of $809
(DMC, 2013$, in 2018). We consider this technology to be on the flat portion of the learning
curve (curve 12) and have applied a low complexity ICM with short term markups through 2022.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The resultant technology costs, penetration rates and total cost applied to the package are shown
below.
Table 2-267 Costs for Advanced Cruise Controls
Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Advanced cruise control
DMC
$809
$785
$761
$738
$723
$709
$695
$681
$667
$654
Advanced cruise control
IC
$144
$144
$144
$143
$143
$113
$113
$112
$112
$112
Advanced cruise control
TC
$953
$929
$905
$882
$867
$822
$807
$793
$780
$766
Advanced cruise control
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Advanced cruise control
Alt 3
0%
0%
0%
20%
20%
20%
40%
40%
40%
40%
Advanced cruise control
TCp
$0
$0
$0
$176
$173
$164
$323
$317
$312
$307
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.10.2 Improved Accessories
For vocational vehicles, we have estimated the cost of this technology based on an
estimate from TIAX of $530 (retail) for light HD, $1000 for medium HD and $2000 for heavy
HD vocational vehicles. These estimates include costs of upgrading to a 42 Volt electrical
system, electric power steering and electric air conditioning. Using these estimates, we divided
by a 1.36 RPE to arrive at cost of $390, $735 and $1471, respectively (DMC, 2013$, in 2018).
We consider this technology to be on the flat portion of the learning curve (curve 12) and have
applied a low complexity ICM with short term markups through 2022. The resultant technology
costs, penetration rates and total cost applied to the package are shown below.
Table 2-268 Costs for Improved Accessories
Vocational Light HD Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Improved accessories
DMC
$390
$378
$367
$356
$349
$342
$335
$328
$322
$315
Improved accessories
IC
$70
$69
$69
$69
$69
$54
$54
$54
$54
$54
Improved accessories
TC
$459
$447
$436
$425
$418
$396
$389
$382
$376
$369
Improved accessories
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Improved accessories
Alt 3
0%
0%
0%
5%
5%
5%
10%
10%
10%
15%
Improved accessories
TCp
$0
$0
$0
$21
$21
$20
$39
$38
$38
$55
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
Table 2-269 Costs for Improved Accessories
Vocational Medium HD Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Improved accessories
DMC
$735
$713
$692
$671
$658
$645
$632
$619
$607
$594
Improved accessories
IC
$131
$131
$131
$130
$130
$102
$102
$102
$102
$102
Improved accessories
TC
$867
$844
$822
$801
$788
$747
$734
$721
$709
$697
Improved accessories
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Improved accessories
Alt 3
0%
0%
0%
5%
5%
5%
10%
10%
10%
15%
Improved accessories
TCp
$0
$0
$0
$40
$39
$37
$73
$72
$71
$104
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-270 Costs for Improved Accessories
Vocational Heavy HD Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Improved accessories
DMC
$1,471
$1,426
$1,384
$1,342
$1,315
$1,289
$1,263
$1,238
$1,213
$1,189
Improved accessories
IC
$262
$262
$261
$261
$260
$205
$205
$205
$204
$204
Improved accessories
TC
$1,733
$1,688
$1,645
$1,603
$1,576
$1,494
$1,468
$1,443
$1,418
$1,393
Improved accessories
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Improved accessories
Alt 3
0%
0%
0%
5%
5%
5%
10%
10%
10%
15%
Improved accessories
TCp
$0
$0
$0
$80
$79
$75
$147
$144
$142
$209
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
For tractors, we have estimated the cost of this technology based on an estimate from
TetraTech of $350 (retail, 2013$). Using that estimate we divided by a 1.36 RPE to arrive at a
cost of $257 (DMC, 2013$, in 2018). We consider this technology to be on the flat portion of
the learning curve (curve 12) and have applied a low complexity ICM with short term markups
through 2022. The resultant technology costs, penetration rates and total cost applied to the
package are shown below for tractors.
Table 2-271 Costs for Improved Accessories
Tractors (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Improved accessories
DMC
$257
$250
$242
$235
$230
$226
$221
$217
$212
$208
Improved accessories
IC
$46
$46
$46
$46
$46
$36
$36
$36
$36
$36
Improved accessories
TC
$303
$295
$288
$281
$276
$261
$257
$252
$248
$244
Improved accessories
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Improved accessories
Alt 3
0%
0%
0%
10%
10%
10%
20%
20%
20%
30%
Improved accessories
TCp
$0
$0
$0
$28
$28
$26
$51
$50
$50
$73
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
For HD pickups and vans, we have estimated the costs for two levels of improved
accessories based on estimates presented in the light-duty 2017-2025 final rule. In that rule, we
estimated the costs of IACC1 and IACC2 at $73 and $118, respectively (both are DMC, 2009$,
in 2015). With updates to 2012$, these costs become $77 and $124, respectively (both are DMC,
2012$, in 2015). Note that IACC2 includes IACC1. We consider these technologies to be on the
flat portion of the learning curve (curve 8) and have applied low complexity markups with near
term markups through 2018. The resultant cost for both are shown below.
Table 2-272 Costs for Improved Accessories
Gasoline & Diesel HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Improved accessories 1 (IACC1)
DMC
$67
$66
$64
$63
$62
$61
$61
Improved accessories 1 (IACC2)
DMC
$109
$106
$104
$102
$100
$99
$98
Improved accessories 1 (IACC1)
IC
$15
$15
$15
$15
$15
$15
$15
Improved accessories 1 (IACC2)
IC
$24
$24
$24
$24
$24
$24
$24
Improved accessories 1 (IACC1)
TC
$82
$80
$79
$78
$77
$76
$75
Improved accessories 1 (IACC2)
TC
$132
$130
$128
$126
$124
$123
$122
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.11.10.3 Weight Reduction, Vocational Vehicles
We have estimated the cost of a 200 pound weight reduction on vocational vehicles at
$4/pound (retail, 2013$). Using that cost we have divided by a 1.36 RPE to arrive at costs of
$588 (DMC, in 2013$, applicable in 2021). We consider this weight reduction level to be on the
flat portion of the learning curve (curve 13) and have applied low complexity ICMs with short
term markups through 2022. We have applied the 200 pound weight reduction level to light and
medium HD vocational vehicles. The resultant technology costs, penetration rates and total cost
applied to the package are shown below.
Table 2-273 Costs for a 200 Pound Weight Reduction
Vocational Light/Medium HD Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Weight reduction, 200 lbs
DMC
$645
$625
$606
$588
$571
$553
$537
$526
$516
$505
Weight reduction, 200 lbs
IC
$106
$105
$105
$105
$105
$82
$82
$82
$82
$82
Weight reduction, 200 lbs
TC
$750
$731
$712
$693
$675
$636
$619
$608
$598
$587
Weight reduction, 200 lbs
Alt la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Weight reduction, 200 lbs
Alt 3
0%
0%
0%
10%
10%
10%
30%
30%
30%
50%
Weight reduction, 200 lbs
TCp
$0
$0
$0
$69
$68
$64
$186
$182
$179
$294
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
We have estimated the cost of weight reduction from use of aluminum wheels based on
the aluminum steer wheel technology discussed in the Phase 1 rules. That technology was
estimated at $459 for two wheels (DMC, 2008$, in 2014). With updates to 2013$, we estimate
the costs at $494 (DMC, 2013$, in 2014). We consider this technology to be on the flat portion
of the learning curve (curve 2) and have applied a low complexity ICM with short term markups
through 2022. The resultant technology costs, penetration rates and total cost applied to the
package are shown below. We apply this technology to heavy HD vocational vehicles having 10
wheels per vehicle.
Table 2-274 Costs for Weight Reduction via use of Aluminum Wheels
Vocational Heavy HD Vehicles (2013$)
TECHNOLOGY

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Weight reduction, A1 wheels
DMC
$2,188
$2,144
$2,102
$2,060
$2,018
$1,978
$1,938
$1,900
$1,881
$1,862
Weight reduction, A1 wheels
IC
$437
$437
$436
$436
$435
$343
$343
$343
$343
$342
Weight reduction, A1 wheels
TC
$2,626
$2,581
$2,538
$2,495
$2,453
$2,321
$2,281
$2,242
$2,223
$2,204
Weight reduction, A1 wheels
Alt
la
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Weight reduction, A1 wheels
Alt 3
0%
0%
0%
10%
10%
10%
30%
30%
30%
50%
Weight reduction, A1 wheels
TCp
$0
$0
$0
$250
$245
$232
$684
$673
$667
$1,102
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost; TCp=total cost applied to the package;
alt=alternative
2.11.10.4 Weight Reduction in HD Pickups and Vans
For this rule, we are estimating weight reduction costs for HD pickups and vans using the
same cost curve used in support of the 2017-2025 light-duty GHG/CAFE FRM. That curve can
be expressed as:

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Mass Reduction Direct Manufacturing Cost (DMC) ($/lb) = 4.55 x Percentage of Mass
Reduction (2012$)
For example, this results in an estimated $80 (2012$) DMC increase for a 5 percent mass
reduction of a 7,000 pound vehicle and $318 (2012$) DMC increase for a 10 percent mass
reduction of a 7,000 pound vehicle, or $0,227 $/lb and $0.455/lb, respectively (both in 2012$).
Consistent with the 2017-2025 light-duty FRM, the agencies consider this DMC to be
applicable to MY2017 and consider mass reduction technology to be on the flat portion of the
learning curve in the 2017-2025MY timeframe. To estimate indirect costs for applied mass
reduction of up to 10 percent, the agencies have applied a low complexity ICM with near term
markups through 2018.
2.11.10.5 Electric Power Steering, HD Pickups and Vans
We have based the costs for electric power steering on the costs used in the light-duty
2017-2025 FRM but have scaled upward that cost by 50 percent to account for the larger HD
vehicles. Using that cost and converting to 2012$ results in a cost of $141 (DMC, 2012$, in
2015). We consider this technology to be on the flat portion of the learning curve (curve 8) and
have applied low complexity markups with near term markups through 2018. The resultant costs
for HD pickups and vans are shown in are shown below.
Table 2-275 Costs for Electric Power Steering
Gasoline & Diesel HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Electric power steering (EPS)
DMC
$124
$121
$119
$117
$114
$113
$112
Electric power steering (EPS)
IC
$27
$27
$27
$27
$27
$27
$27
Electric power steering (EPS)
TC
$151
$148
$146
$144
$141
$140
$139
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
2.11.10.6 Low Drag Brakes, HD Pickups and Vans
We have based the costs for low drag brakes on the costs used in the light-duty 2017-
2025 FRM but have scaled upward that cost by 50 percent to account for the larger HD vehicles.
Using that cost and converting to 2012$ results in a cost of $91 (DMC, 2012$, in any year). We
consider this technology to be beyond the learning curve (curve 1) and have applied low
complexity markups with near term markups through 2018. The resultant costs for HD pickups
and vans are shown in are shown below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-276 Costs for Low Drag Brakes
Gasoline & Diesel HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Low drag brakes
DMC
$91
$91
$91
$91
$91
$91
$91
Low drag brakes
IC
$18
$18
$18
$18
$18
$18
$18
Low drag brakes
TC
$109
$109
$109
$109
$109
$109
$109
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
2.11.10.7 Driveline Friction Reduction, Diesel HD Pickups & Vans
We have estimated the cost of driveline friction reduction based on the cost of secondary
axle disconnect in the light-duty 2017-2025 final rule. Using that cost of $80 (DMC, 2009$, in
2015), we have scaled upward by 50 percent to account for the larger HD componentry to arrive
at a cost of $126 (DMC, 2012$, in 2015). We consider this technology to be on the flat portion
of the learning curve (curve 3) and have applied low complexity markups with near term
markups through 2022. The resultant costs for driveline friction reduction (applied only to diesel
HD pickups & vans) are shown below.
Table 2-277 Costs for Driveline Friction Reduction
Diesel HD Pickups and Vans (2012$)
ITEM

2021
2022
2023
2024
2025
2026
2027
Driveline friction reduction
DMC
$108
$106
$104
$102
$100
$99
$98
Driveline friction reduction
IC
$30
$30
$24
$24
$24
$24
$24
Driveline friction reduction
TC
$139
$136
$128
$126
$124
$123
$122
Notes: DMC=direct manufacturing cost; IC=indirect cost; TC=total cost
2.12 Package Costs
Chapter 2.11 presents detailed technology costs along with penetration rates to illustrate
how each technology is accounted for in the package costs. Here we present package costs by
regulated sector (i.e., vocational heavy HD, urban vehicles) and package costs by MOVES
sourcetype (i.e., diesel refuse trucks). We determine package costs by MOVES sourcetype so
that we can calculate total program costs (i.e., package costs multiplied by vehicle sales) since
sourcetypes are the sales figures that we can glean from MOVES. As a result, the sourcetype
package costs presented here are the costs used in our program cost estimations.
2.12.1 Package Costs by Regulated Sector
2.12.1.1 Vocational Vehicles
We have estimated costs for 9 vocational segments and 2 fuels. We present package
costs in the tables below for these for alternative 3 relative to alternatives la and lb and
separately for diesel and gasoline vehicles.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-278 Package Costs for Regulated Vocational Segment
Alternative 3 Incremental to Alternative la & lb
Diesel (2013$)
WEIGHT CLASS
SPEED
2021
2022
2023
2024
2025
2026
2027
Light HD
Urban
$1,106
$1,083
$1,020
$1,959
$1,925
$1,873
$2,533
Light HD
Multipurpose
$1,164
$1,140
$1,079
$2,018
$1,983
$1,919
$2,571
Light HD
Regional
$873
$855
$825
$1,272
$1,251
$1,224
$1,486
Medium HD
Urban
$1,116
$1,092
$1,030
$2,082
$2,046
$1,977
$2,727
Medium HD
Multipurpose
$1,146
$1,123
$1,058
$2,110
$2,074
$2,004
$2,771
Medium HD
Regional
$851
$833
$800
$1,274
$1,252
$1,226
$1,500
Heavy HD
Urban
$1,334
$1,308
$1,236
$2,932
$2,882
$2,785
$4,151
Heavy HD
Multipurpose
$1,625
$1,595
$1,502
$3,813
$3,749
$3,638
$5,025
Heavy HD
Regional
$2,562
$2,517
$2,359
$4,009
$3,942
$3,869
$5,670
Table 2-279 Package Costs for Regulated Vocational Segment
Alternative 3 Incremental to Alternative la & lb
Gasoline (2013$)
WEIGHT CLASS
SPEED
2021
2022
2023
2024
2025
2026
2027
Light HD
Urban
$947
$930
$872
$1,649
$1,616
$1,569
$2,177
Light HD
Multipurpose
$1,004
$986
$931
$1,708
$1,673
$1,615
$2,215
Light HD
Regional
$714
$701
$677
$962
$941
$921
$1,130
Medium HD
Urban
$979
$961
$904
$1,805
$1,770
$1,705
$2,406
Medium HD
Multipurpose
$1,010
$991
$932
$1,833
$1,797
$1,732
$2,450
Medium HD
Regional
$715
$702
$674
$997
$975
$954
$1,179
Heavy HD
Urban
$1,198
$1,177
$1,110
$2,655
$2,606
$2,513
$3,830
Heavy HD
Multipurpose
$1,489
$1,464
$1,376
$3,536
$3,472
$3,366
$4,704
Heavy HD
Regional
$2,426
$2,386
$2,233
$3,732
$3,665
$3,598
$5,349
2.12.1.2 Tractors
We have estimated costs for 7 tractor segments and 1 fuel. We present package costs in
the tables below for these for alternative 3 relative to alternatives la and lb.
Table 2-280 Package Costs for Regulated Tractor Segment
Alternative 3 Incremental to Alternative la
Diesel (2013$)
CLASS
TYPE
2021
2022
2023
2024
2025
2026
2027
7
Day cab, low roof
$5,134
$5,052
$4,682
$8,037
$7,859
$7,728
$10,235
7
Day cab, high roof
$5,240
$5,151
$4,772
$8,210
$8,026
$7,852
$10,298
8
Day cab, low roof
$5,228
$5,143
$4,769
$8,201
$8,020
$7,887
$10,439
8
Day cab, high roof
$5,317
$5,227
$4,844
$8,358
$8,172
$7,993
$10,483
8
Sleeper cab, low roof
$7,181
$7,061
$6,580
$11,100
$10,871
$10,714
$13,535
8
Sleeper cab, mid roof
$7,175
$7,056
$6,574
$11,100
$10,871
$10,714
$13,574
8
Sleeper cab, high roof
$7,276
$7,239
$6,751
$11,306
$11,068
$10,857
$13,749

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-281 Package Costs for Regulated Tractor Segment
Alternative 3 Incremental to Alternative lb
Diesel (2013$)
CLASS
TYPE
2021
2022
2023
2024
2025
2026
2027
7
Day cab, low roof
$5,267
$5,112
$4,659
$7,944
$7,705
$7,536
$9,937
7
Day cab, high roof
$5,093
$4,977
$4,594
$8,016
$7,816
$7,621
$10,042
8
Day cab, low roof
$5,360
$5,203
$4,745
$8,108
$7,866
$7,695
$10,141
8
Day cab, high roof
$5,170
$5,053
$4,667
$8,164
$7,962
$7,763
$10,227
8
Sleeper cab, low roof
$7,195
$6,988
$6,438
$10,883
$10,614
$10,404
$13,140
8
Sleeper cab, mid roof
$7,102
$6,886
$6,337
$10,800
$10,514
$10,306
$13,043
8
Sleeper cab, high roof
$7,115
$7,057
$6,577
$11,122
$10,871
$10,656
$13,515
2.12.1.3 Trailers
We have estimated costs for seven trailer types (i.e. for each of the subcategories). The
dry and refrigerated vans have identical stringency and technology packages, so costs are
presented by length category only. The tire-based design standards for non-aero box vans are a
single category, but separate non-aero costs were considered for long vans and short vans,
because we assumed all short vans have a single axle, which results in fewer wheels and tires and
lower costs. We present package costs in the tables below for these for alternative 3 relative to
alternative la and lb.
Table 2-282 Costs for Trailers
Alternative 3 Incremental to Alternative la (2013$)
TYPE
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Long van, Full aero
$716
$688
$673
$1,081
$1,061
$1,030
$1,204
$1,184
$1,183
$1,370
Long van, Partial aero
$1,441
$1,383
$1,352
$1,337
$1,313
$1,274
$1,251
$1,229
$1,213
$1,196
Long van, No aero
$461
$448
$435
$438
$429
$413
$405
$398
$390
$382
Short van, Full aero
$339
$330
$322
$772
$757
$733
$1,171
$1,151
$1,144
$1,204
Short van, Partial aero
$514
$500
$487
$957
$940
$910
$894
$879
$867
$855
Short van, No aero
$231
$224
$218
$219
$215
$207
$202
$199
$195
$191
Non-box
$448
$436
$424
$412
$406
$390
$383
$377
$361
$354
Table 2-283 Costs for Trailers
Alternative 3 Incremental to Alternative lb (2013$)
TYPE
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Long van, Full aero
$716
$676
$650
$1,047
$1,016
$975
$1,139
$1,109
$1,098
$1,276
Long van, Partial aero
$1,441
$1,383
$1,352
$1,337
$1,313
$1,274
$1,251
$1,229
$1,213
$1,196
Long van, No aero
$461
$448
$435
$438
$429
$413
$405
$398
$390
$382
Short van, Full aero
$339
$330
$322
$772
$757
$733
$1,171
$1,151
$1,144
$1,204
Short van, Partial aero
$514
$500
$487
$957
$940
$910
$894
$879
$867
$855
Short van, No aero
$231
$224
$218
$219
$215
$207
$202
$199
$195
$191
Non-box
$448
$436
$424
$412
$406
$390
$383
$377
$361
$354

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
2.12.1.4 HD Pickups and Vans
The costs presented in the table below are CAFE model outputs used in analysis Method
B. We describe the CAFE model and how these costs were generated in Chapter 6 and 11 of this
RIA.
Table 2-284 Package Costs for HD Pickups and Vans (2013$)
ALTERNATIVE
BASELINE
CASE
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
3
la
$114
$105
$108
$524
$516
$804
$963
$1,180
$1,244
$1,364
3
lb
$113
$105
$102
$513
$505
$793
$952
$1,168
$1,233
$1,349
2.12.2 Package Costs by MOVES Sourcetype
The package costs by segment can then be used to calculate package costs by MOVES
sourcetype. To do this, we need the percentage of the MOVES sourcetype fleet comprised of
each regulated sector. Table 2-285 shows this breakout for the vocational sector and Table 2-286
shows it for tractors. Package costs for vocational vehicles make the conservative assumption of
full program compliance rather than compliance with the more flexible, less costly custom
chassis program.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-285 Fleet Mix by MOVES Sourcetype and Regulated Sector — Vocational"
ENGINE
FUEL
SPEED
INTERCITY
TRANSIT
SCHOOL
REFUSE
SINGLE
SINGLE
MOTOR



BUS
BUS
BUS
TRUCKS
UNIT
SHORT
HAUL
UNIT
LONG
HAUL
HOMES
Light HD
Gasoline
Urban
0%
27%
1%
0%
41%
0%
0%
Light HD
Gasoline
Multipurpose
0%
0%
0%
0%
33%
0%
0%
Light HD
Gasoline
Regional
0%
0%
0%
0%
7%
0%
54%
Medium HD
Gasoline
Urban
0%
10%
85%
0%
7%
0%
0%
Medium HD
Gasoline
Multipurpose
0%
0%
9%
0%
9%
0%
0%
Medium HD
Gasoline
Regional
0%
0%
0%
0%
3%
0%
41%
Heavy HD
Gasoline
Urban
0%
63%
4%
0%
0%
0%
0%
Heavy HD
Gasoline
Multipurpose
0%
0%
0%
0%
0%
0%
0%
Heavy HD
Gasoline
Regional
0%
0%
0%
0%
0%
0%
5%
Light HD
Diesel
Urban
0%
0%
1%
0%
21%
0%
0%
Light HD
Diesel
Multipurpose
0%
0%
0%
0%
17%
0%
0%
Light HD
Diesel
Regional
2%
0%
0%
0%
4%
25%
54%
Medium HD
Diesel
Urban
0%
0%
85%
2%
12%
0%
0%
Medium HD
Diesel
Multipurpose
0%
0%
9%
0%
17%
0%
0%
Medium HD
Diesel
Regional
15%
0%
0%
0%
5%
37%
41%
Heavy HD
Diesel
Urban
0%
100%
4%
88%
5%
0%
0%
Heavy HD
Diesel
Multipurpose
0%
0%
0%
10%
15%
0%
0%
Heavy HD
Diesel
Regional
83%
0%
0%
0%
5%
37%
5%
Heavy HD
CNG
Urban
0%
100%
0%
0%
0%
0%
0%
Heavy HD
CNG
Multipurpose
0%
0%
0%
0%
0%
0%
0%
Heavy HD
CNG
Regional
0%
0%
0%
0%
0%
0%
0%
Note:
a Columns add to 100% or 0% within each fuel type.
Table 2-286 Fleet Mix by MOVES Sourcetype and Regulated Sector - Tractors3
ENGINE
MOVES
CLASS
CLASS
CLASS
CLASS
CLASS 8
CLASS 8
CLASS 8

SOURCETYPE
7
7
8
8
SLEEPER
SLEEPER
SLEEPER


DAY
DAY
DAY
DAY
CAB
CAB
CAB


CAB
CAB
CAB
CAB
LOW
MID
HIGH


LOW
HIGH
LOW
HIGH
ROOF
ROOF
ROOF


ROOF
ROOF
ROOF
ROOF



Medium
HD
Combination
Short haul
11%
11%
0%
0%
0%
0%
0%
Heavy
HD
Combination
Short haul
0%
0%
39%
39%
0%
0%
0%
Heavy
HD
Combination
Long haul
0%
0%
0%
0%
5%
15%
80%
Note:
a Combination short haul adds to 100% and long haul to 100%.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Using the fleet mix information shown in Table 2-285 and Table 2-286, along with the
package costs shown in Chapter 2.12.1, we can generate the package costs by MOVES
sourcetype (note that package costs by MOVES sourcetype differ from package costs by
regulated sector only for vocational vehicles and tractors; trailer and HD pickup and van costs do
not change). These costs are shown below.
Table 2-287 Package Costs by MOVES Sourcetype
Alternative 3 Incremental to Alternative la (2013$)
SOURCETYPE
FUEL
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Intercity Bus
Diesel
$0
$0
$0
$2,266
$2,225
$2,089
$3,534
$3,475
$3,411
$4,946
Transit Bus
Diesel
$0
$0
$0
$1,334
$1,308
$1,236
$2,932
$2,882
$2,785
$4,151
School Bus
Diesel
$0
$0
$0
$1,130
$1,106
$1,043
$2,127
$2,090
$2,019
$2,799
Refuse Truck
Diesel
$0
$0
$0
$1,357
$1,330
$1,256
$2,996
$2,945
$2,847
$4,198
SingleUnit ShortHaul
Diesel
$0
$0
$0
$1,270
$1,244
$1,174
$2,392
$2,351
$2,281
$3,142
SingleUnit LongHaul
Diesel
$0
$0
$0
$1,497
$1,468
$1,389
$2,296
$2,258
$2,214
$3,056
MotorHome
Diesel
$0
$0
$0
$954
$934
$896
$1,418
$1,394
$1,365
$1,714
Intercity Bus
Gasoline










Transit Bus
Gasoline
$0
$0
$0
$1,109
$1,089
$1,026
$2,302
$2,258
$2,181
$3,247
School Bus
Gasoline
$0
$0
$0
$993
$975
$917
$1,850
$1,813
$1,747
$2,477
Refuse Truck
Gasoline










SingleUnit ShortHaul
Gasoline
$0
$0
$0
$951
$933
$880
$1,628
$1,595
$1,544
$2,126
SingleUnit LongHaul
Gasoline










MotorHome
Gasoline
$0
$0
$0
$805
$791
$758
$1,123
$1,100
$1,076
$1,374
Transit Bus
CNG
$0
$0
$0
$1,059
$1,039
$973
$2,519
$2,476
$2,384
$3,705
Comb ShortHaul
Tractor
Diesel
$0
$0
$0
$5,254
$5,167
$4,789
$8,245
$8,062
$7,907
$10,418
Comb LongHaul
Tractor
Diesel
$0
$0
$0
$7,256
$7,203
$6,716
$11,265
$11,029
$10,829
$13,712
Long Van, Full Aero

$716
$688
$673
$1,081
$1,061
$1,030
$1,204
$1,184
$1,183
$1,370
Long Van, Partial
Aero

$1,441
$1,383
$1,352
$1,337
$1,313
$1,274
$1,251
$1,229
$1,213
$1,196
Long Van, No Aero

$461
$448
$435
$438
$429
$413
$405
$398
$390
$382
Short Van, Full Aero

$339
$330
$322
$772
$757
$733
$1,171
$1,151
$1,144
$1,204
Short Van, Partial
Aero

$514
$500
$487
$957
$940
$910
$894
$879
$867
$855
Short Van, No Aero

$231
$224
$218
$219
$215
$207
$202
$199
$195
$191
Non-Box

$448
$436
$424
$412
$406
$390
$383
$377
$361
$354
Vocational
Weighted
Avg
$0
$0
$0
$1,110
$1,088
$1,027
$2,022
$1,986
$1,927
$2,662
Tractor/Trailer
Weighted
Avg
$568
$548
$535
$7,352
$7,269
$6,799
$11,134
$10,901
$10,712
$13,550
Note: Blank cells indicate no such vehicles of that sourcetype/fuel combination.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 2-288 Package Costs by MOVES Sourcetype
Alternative 3 Incremental to Alternative lb (2013$)
SOURCETYPE
FUEL
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
Intercity Bus
Diesel
$0
$0
$0
$2,266
$2,225
$2,089
$3,534
$3,475
$3,411
$4,946
Transit Bus
Diesel
$0
$0
$0
$1,334
$1,308
$1,236
$2,932
$2,882
$2,785
$4,151
School Bus
Diesel
$0
$0
$0
$1,130
$1,106
$1,043
$2,127
$2,090
$2,019
$2,799
Refuse Truck
Diesel
$0
$0
$0
$1,357
$1,330
$1,256
$2,996
$2,945
$2,847
$4,198
SingleUnit
ShortHaul
Diesel
$0
$0
$0
$1,270
$1,244
$1,174
$2,392
$2,351
$2,281
$3,142
SingleUnit
LongHaul
Diesel
$0
$0
$0
$1,497
$1,468
$1,389
$2,296
$2,258
$2,214
$3,056
MotorHome
Diesel
$0
$0
$0
$954
$934
$896
$1,418
$1,394
$1,365
$1,714
Intercity Bus
Gasoline










Transit Bus
Gasoline
$0
$0
$0
$1,109
$1,089
$1,026
$2,302
$2,258
$2,181
$3,247
School Bus
Gasoline
$0
$0
$0
$993
$975
$917
$1,850
$1,813
$1,747
$2,477
Refuse Truck
Gasoline










SingleUnit
ShortHaul
Gasoline
$0
$0
$0
$951
$933
$880
$1,628
$1,595
$1,544
$2,126
SingleUnit
LongHaul
Gasoline










MotorHome
Gasoline
$0
$0
$0
$805
$791
$758
$1,123
$1,100
$1,076
$1,374
Transit Bus
CNG
$0
$0
$0
$1,059
$1,039
$973
$2,519
$2,476
$2,384
$3,705
Comb
ShortHaul
Diesel
$0
$0
$0
$5,246
$5,110
$4,689
$8,101
$7,880
$7,696
$10,141
Comb
LongHaul
Diesel
$0
$0
$0
$7,117
$7,028
$6,534
$11,061
$10,804
$10,591
$13,426
Long Van, Full
Aero

$716
$676
$650
$1,047
$1,016
$975
$1,139
$1,109
$1,098
$1,276
Long Van,
Partial Aero

$1,441
$1,383
$1,352
$1,337
$1,313
$1,274
$1,251
$1,229
$1,213
$1,196
Long Van, No
Aero

$461
$448
$435
$438
$429
$413
$405
$398
$390
$382
Short Van, Full
Aero

$339
$330
$322
$772
$757
$733
$1,171
$1,151
$1,144
$1,204
Short Van,
Partial Aero

$514
$500
$487
$957
$940
$910
$894
$879
$867
$855
Short Van, No
Aero

$231
$224
$218
$219
$215
$207
$202
$199
$195
$191
Non-Box

$448
$436
$424
$412
$406
$390
$383
$377
$361
$354
Vocational
Weighted
Avg
$0
$0
$0
$1,110
$1,088
$1,027
$2,022
$1,986
$1,927
$2,662
Tractor/Trailer
Weighted
Avg
$639
$548
$482
$7,248
$7,120
$6,624
$10,925
$10,660
$10,447
$13,226

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
References
1	Committee to Assess Fuel Economy Technologies for Medium- and Heavy-Duty Vehicles; National Research
Council; Transportation Research Board (2010). "Technologies and Approaches to Reducing the Fuel Consumption
of Medium- and Heavy-Duty Vehicles," (hereafter, "NAS Report"). Washington, D.C., The National Academies
Press. Available electronically from the National Academies Press Website at
http://www.nap.edu/catalog.php?record_id=12845 (last accessed September 10, 2010).
2	TIAX, LLC. "Assessment of Fuel Economy Technologies for Medium- and Heavy-Duty Vehicles," Final Report
to the National Academy of Sciences, November 19, 2009.
3	U.S. EPA. EPA Lumped Parameter Model HD Version 1.0.0.5, 2010. Docket #EPA-HQ-OAR-2010-0162.
4	NESCCAF, ICCT, Southwest Research Institute, and TIAX. Reducing Heavy-Duty Long Haul Combination
Truck Fuel Consumption and CO2 Emissions. October 2009.
5ICF International. Investigation of Costs for Strategies to Reduce Greenhouse Gas Emissions for Heavy-Duty On-
Road Vehicles. July 2010. Docket Identification Number EPA-HQ-OAR-2010-0162-0044.
6	Reinhart, T.E. (2015, June). Commercial Medium- and Heavy-Duty Truck Fuel Efficiency Technology Study -
Report #1. (Report No. DOT HS 812 146). Washington, DC: National Highway Traffic Safety Administration.
7	Reinhart, T. (February 2016). Commercial Medium- and Heavy-Duty (MD/HD) Truck Fuel Efficiency
Technology Study - Report #2. Washington, DC: National Highway Traffic Safety Administration. EPA-HQ-OAR-
2014-0827-1623.
8	Schubert, R., Chan, M., Law, K. (2015). Commercial Medium- and Heavy-Duty Truck Fuel Efficiency
Technology Cost Study. Washington, DC: National Highway Traffic Safety Administration.
9	Northeast States Center for a Clean Air Future. "Reducing Greenhouse Gas Emissions from Light-Duty Motor
Vehicles." September 2004.
10	Energy and Environmental Analysis, Inc. "Technology to Improve the Fuel Economy of Light Duty Trucks to
2015." May 2006.
11	"Light-Duty Automotive Technology, Carbon Dioxide Emissions, and Fuel Economy Trends: 1975 - 2014," EPA-
420-R-14-023, October 2014. Available at http://www3.epa.gov/otaq/fetrends.htm (last accessed October 31, 2014).
12	"Development and Optimization of the Ford 3.5L V6 EcoBoost Combustion System," Yi,J., Wooldridge, S.,
Coulson, G., Hilditch, J. Iyer, C.O., Moilanen, P., Papaioannou, G., Reiche, D. Shelby, M., VanDerWege, B.,
Weaver, C. Xu, Z., Davis, G., Hinds, B. Schamel, A. SAE Technical Paper No. 2009-01-1494, 2009, Docket EPA-
HQ-OAR-2009-0472-2860.
13	David Woldring and Tilo Landenfeld of Bosch, and Mark J. Christie of Ricardo, "DI Boost: Application of a
High Performance Gasoline Direct Injection Concept," SAE 2007-01-1410. Available at
http://www.sae.org/technical/papers/2007-01-1410 (last accessed Nov. 9, 2008).
14	Yves Boccadoro, Loi'c Kermanac'h, Laurent Siauve, and Jean-Michel Vincent, Renault Powertrain Division, "The
New Renault TCE 1.2L Turbocharged Gasoline Engine," 28th Vienna Motor Symposium, April 2007.
15	Tobias Heiter, Matthias Philipp, Robert Bosch, "Gasoline Direct Injection: Is There a Simplified, Cost-Optimal
System Approach for an Attractive Future of Gasoline Engines?" AVL Engine & Environment Conference,
September 2005.
16	Stanton, Donald. "Enabling High Efficiency Clean Combustion." 2009 Semi-Mega Merit Review of the
Department of Energy. May 21, 2009. Last accessed on August 25, 2010 at
http://wwwl.eere.energy.gOv/vehiclesandfuels/pdfs/merit_review_2009/advanced_combustion/ace_40_stanton.pdf.
17	Zhang, H. Heavy Truck Engine Development & HECC. 2009 DOE Semi-Mega Merit Review, May 21, 2009.
Last accessed on August 25, 2010 at
http://wwwl.eere.energy.gov/vehiclesandfuels/pdfs/merit_review_2009/advanced_combustion/ace_42_zhang.pdf.
18	de Ojeda, W., SuperTruck - Development and Demonstration of a Fuel-Efficient Class 8 Tractor & Trailer Engine
Systems, http://energy.gov/eere/vehicles/downloads/supertruck-development-and-demonstration-fuel-efficient-class-
8-tractor-.\
19	Zhang, H. Detroit Diesel. "High Efficiency Clean Combustion for Heavy-Duty Engine." August 6, 2008
presentation to DEER Conference. Last accessed on August 25, 2010 at
http://wwwl.eere.energy.gov/vehiclesandfuels/pdfs/deer_2008/session5/deer08_zhang.pdf

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
20	Singh, S., SuperTruck Program: Engine Project Review Recovery Act - Class 8 Truck Freight Efficiency
Improvement Project, DOE SuperTruck Merit Review, Department of Energy, 2014.
21	See the NAS Report, page 53 cites Detroit Diesel Corporation, DD15 Brochure, DDC-EMC-BRO-0003-0408,
April 2008
22	Greszler, A., "View from the Bridge - Commercial Vehicle Perspective," DEER Conference, Department of
Energy, 2012. http://wwwl.eere.energy.gov/vehiclesandfuels/resources/proceedings/2012_deer_presentations.html.
23	de Ojeda, W., "Development and Demonstration of a Fuel-Efficient HD Engine (Department of Energy
SuperTruck Program)," DEER Conference, Department of Energy, 2012.
http://wwwl.eere.energy.gov/vehiclesandfuels/resources/proceedings/2012_deer_presentations.html,
24	Jadin, D., "SuperTruck - Development and Demonstration of a Fuel-Efficient Class 8 Tractor & Trailer," DOE
Supertruck Merit Review, Department of Energy, 2012.
http://wwwl.eere.energy.gov/vehiclesandfuels/resources/proceedings/2012_deer_presentations.html,
25	Koeberlein, D., Technology and System Level Demonstration of Highly Efficient and Clean, Diesel Powered
Class 8 Trucks, 2012, http://energy.gov/sites/prod/files/2014/03/fl0/ace057_koeberlein_2012_o.pdf
26	Sisken, K., Rotz, D., "Recovery Act - Class 8 Truck Freight Efficiency Improvement Project," DOE Supertruck
Merit Review, Department of Energy, 2012.
http://wwwl.eere.energy.gov/vehiclesandfuels/resources/proceedings/2012_deer_presentations.html.
27	de Ojeda, W., "Development and Demonstration of a Fuel-Efficient HD Truck," DEER Conference, Department
of Energy, 2011,
http://wwwl.eere.energy.gov/vehiclesandfuels/resources/proceedings/201 l_deer_presentations.html,
28	SAE Off-Highway Magazine. Federal-Mogul's Latest Piston Ring Developments help reduce friction in heavy-
duty engines. October 2, 2014. Last accessed on October 3, 2014 at
http://www.oemoffhighway.com/press_release/12007334/federal-moguls-latest-piston-ring-developments-help-
rcducc-friction-in-hcavY-duty-cngincs.
29	Sisken, K., SuperTruck Program: Engine Project Review Recovery Act - Class 8 Truck Freight Efficiency
Improvement Project, DOE Supertruck Merit Review, Department of Energy, 2013.
30	Amar, P., SuperTruck Development and Demonstration of a Fuel-Efficient Class 8 Highway Vehicle Systems,
DOE Supertruck Merit Review, Department of Energy, 2013.
31	Koeberlein, D., Technology and System Level Demonstration of Highly Efficient and Clean, Diesel Powered
Class 8 Trucks, DOE Supertruck Merit Review, Department of Energy, 2013.
32	Stanton, D. W., Systematic development of highly efficient and clean engines to meet future commercial vehicle
greenhouse gas regulations. Presented at SAE Commercial Vehicle Engineering Congress. Chicago, IL, October
2013.
33	Amar, P., Volvo SuperTruck Powertrain Technologies for Efficiency Improvement, DOE Supertruck Merit
Review, Department of Energy, 2014.
34	Rotz, D., Super Truck Program: Vehicle Project Review, DOE Supertruck Merit Review, Department of Energy,
2014.
35	Damon, K., DOE SuperTruck Program - Technology and System Level Demonstration of Highly Efficient b and
Clean, Diesel Powered Class 8 Trucks, DOE Supertruck Merit Review, Department of Energy, 2014.
36	2015 Annual Fleet Fuel Study, report of a study conducted by the North American Council for Freight Efficiency
(NACFE), May 6, 2015.
37	Assumes travel on level road at 65 MPH. (21st Century Truck Partnership Roadmap and Technical White Papers,
December 2006. U.S. Department of Energy, Energy Efficiency and Renewable Energy Program. 21CTP-003. p.
36.
38	Reducing Heavy-Duty Long Haul Combination Truck Fuel Consumption and CO2 Emissions, ICCT, October
2009.
39	Overdrive magazine September 2014, "Western Star introduces re-designed on-highway tractor,"
http://www.overdriveonline.com/photo-video-western-star-unveils-re-designed-on-highway-
tractor/?utm_medium=overdrive&utm_campaign=site_click&utm_source=top_stories
40	Cornelius Pfeifer (Rochling Automotive), Society of Automotive Engineer (SAE) Paper #2014-01-0633:
"Evolution of Active Grille Shutters." April 1, 2014.
41	Jason Leuschen and Kevin R. Cooper (National Research Council of Canada), Society of Automotive Engineer
(SAE) Paper #2006-01-3456: "Full-Scale Wind Tunnel Tests of Production and Prototype, Second-Generation
Aerodynamic Drag-Reducing Devices for Tractor-Trailers.," November 2, 2006.
42	http://www.todaystrucking.com/supertruck-flies-to-future-of-fuel-efficiency-at-107-mpg.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
43	"Tires & Truck Fuel Economy," A New Perspective. Bridgestone Firestone, North American Tire, LLC, Special
Edition Four, 2008.
44	"Michelin's Green Meters," Press Kit, October, 30, 2007. http://www.michelin-green-
meter.com/main.php?cLang=en (Complete Press File, Viewed March 6, 2010).
45	Argonne National Laboratory. "Evaluation of Fuel Consumption Potential of Medium and Heavy Duty Vehicles
through Modeling and Simulation." October 2009. Page 91.
46	21st Century Truck Partnership, "Roadmap and technical White Papers," U.S. Department of Energy, Technical
paper: 21CTP-0003, December 2006.
47	NACFE December 2010 "Executive Report - Wide Base Tires," available at http://nacfe.org/wp-
content/uploads/2010/12/NACFE-ER-1002-Wide-Base-Tires-Dec-2010.pdf.
48	"Energy Efficiency Strategies for Freight Trucking: Potential Impact on Fuel Use and Greenhouse Gas
Emissions," J. Ang-Olson, W. Schroer, Transportation Research Record: Journal of the Transportation Research
Board, 2002(1815): 11-18.
49	"Effect of Single Wide Tires and Trailer Aerodynamics on Fuel efficiency and NOx Emissions of Class 8 Line-
Haul Tractor-Trailer," J. Bachman, A. Erb, C. Bynum, U.S. Environmental Protection Agency, SAE International,
Paper Number 05CV-45, 2005.
50	"Class 8 Heavy Truck Duty Cycle Project Final Report," U.S. Department of Energy, Oak Ridge National
Laboratory, ORNL/TM-2008/122, p. 21, December 2008. Accessed January 19, 2010 at
http ://cta. ornl. gov/cta/Publications/Reports/ORNL_TM_2008-122.pdf.
51	"Are Ultra-Wide, Ultra-Low Aspect Ratio Tires the Next Big Thing?" K. Rohlwing, Today's Tire Industry, Vol.
I,	Issue 1, July/August, July 2003.
52	"New Generation Wide Base Single Tires," American Trucking Association, White paper presented at the
International Workshop on the use of wide tires sponsored by Federal Highway Administration, Turner-Fairbank
Highway Research Center, October 25-26, 2007, Revision 9, December 21, 2007, Accessed on February 3, 2010 at
http://www.arc.unr.edu/Workshops/Wide_Tires/Wide_Base_Summary-v9-ATA-whitepaper.pdf
53	HDT Trucking info, March 2014,"Switching to Wide-Base Singles: Keys to Success," Available online:
http://www.truckinginfo.eom/channel/equipment/article/story/2014/03/switching-to-wide-base-singles-keys-to-
success.asp.v
54	"Recommended Practice: Guidelines for Outset Wide Base Wheels for Drive, Trailer and Auxiliary Axle
Applications (Draft)," Technology and Maintenance Council, Council of American Trucking Associations,
circulated September 28, 2009.
55	"Tire Pressure Systems - Confidence Report." North American Council for Freight Efficiency. 2013. Available
online: http://nacfe.org/wp-content/uploads/2014/01/TPS-Detailed-Confidence-Reportl.pdf.
56	Technology and Maintenance Council of the American Trucking Associations, Tire Air Pressure Study, Tire
Debris Prevention Task Force S.2 Tire & Wheel Study Group; May 2002.
57	"Commercial Vehicle Tire Condition Sensors," Federal Motor Carrier Safety Administration. Report No.
FMCSA-PSV-04-002, November 2003.
58	U.S. Environmental Protection Agency Office of Transportation and Air Quality SmartWay Transport
Partnership, A Glance at Clean Freight Strategies: Automatic Tire Inflation Systems EPA 420-F-04-0010; February
2004.
59	"A Day in the Life of a Tire," Pressure Systems International, Presented to EPA on August 20, 2014.
60	"Buses & Retread Tires," The Tire Retread & Repair Information Bureau, Pacific Grove, Ca., Accessed on
January 27, 2010 at http://www.retread.org/packet/index.cfm/ID/284.htm.
61	"What are Retreaders Doing to Improve Fuel Efficiency?" H. Inman, Tire Review, December 11, 2006, Accessed
on February 18, 2010 at
http://www.tirereview.com/Article/59777/what_are_retreaders_doing_to_improve_fuel_efficiency.aspx.
62	Modern Tire Dealer, Top Retreaders, ibid.
63	Tire Industry Association and the Tire Retread and Repair Information Bureau, Understanding Retreading,
accessed August 2, 2010, at http://www.retread.org/pdf/UR/UnderstandingRetreading_web.pdf.
64	Todaystrucking.com, Retread Tires FAQ (07/17/2006), accessed August 2, 2010.
65	"Better Fuel Efficiency? Start with a Strong Tire Program," H. Inman, Fleet & Tire 2006, Tire Review, December
II,	2006, Accessed on February 18, 2010 at http://www.tirereview.com/better-fuel-economy-start-with-a-strong-
tire-program/.
66	See 2010 NAS Report, Note 1, Finding 4-6, page 87.
67	See NHTSA Technology Report #1 (2015), Note 6.
68	NACFE. Executive Report - 6x2 (Dead Axle) Tractors. November 2010. See Docket.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
69	Reinhart, T.E. (2015, June). Commercial Medium- and Heavy-Duty Truck Fuel Efficiency Technology Study -
Report #1. (Report No. DOT HS 812 146). Washington, DC: National Highway Traffic Safety Administration.
70	Fleet Owner, "Meritor Expects to offer new tandem axle in 2013," http://fleetowner.com/equipment/meritor-
expects-offer-new-tandem-axle-2013, December 2012.
71	DanaHolding Corporation Patents (8,523,738, 8,795,125, and 8,911,321).
72	See 4WD Axle Actuator Housing at 1A Auto, accessed November 2014 from http://www.laauto.com/4wd-axle-
actuator-housing/c/264.
73	DriveAluminum. The Aluminum Advantage in Commercial Vehicle Applications Webinar. 2009. Last accessed
on May 21, 2015 at
http://www.drivealuminum.org/research-resources/PDF/Webinars/2009/The-Aluminum-Advantage-Commercial-
Vehicle-Applications-Webinar-2009-Dec-Commercial-Vehicle-Webinar.pdf/view.
74	Committee on the Assessment of Technologies for Improving Light-Duty Vehicle Fuel Economy; National
Research Council, "Assessment of Fuel Economy Technologies for Light-Duty Vehicles," 2011. Available at
http://www.nap.edu/catalog.php?record_id=12924 (last accessed Jun 27, 2012).
75	Energy Savings through Increased Fuel Economy for Heavy Duty Trucks," Therese Langer,
American Council for an Energy-Efficient Economy prepared for the National Commission on Energy
Policy, February 11, 2004.
76	Best practices Guidebook for Greenhouse Gas Reductions in Freight Transportation," H.
Christopher Fey, Po-Yao Kuo, North Carolina State University prepared for the U.S. Department of
Transportation, October 4, 2007.
77U.S. EPA. http://www3.epa.gov/smartway/documents/weightreduction.pdf.
78	SAE World Congress, "Focus B-pillar 'tailor rolled' to 8 different thicknesses," Feb. 24, 2010. Available at
http://www.sae.org/mags/AEI/7695 (last accessed Jun. 10, 2012).
79	"Preliminary Vehicle Mass Estimation Using Empirical Subsystem Influence Coefficients," Malen, D.E., Reddy,
K. Auto-Steel Partnership Report, May 2007, Docket EPA-HQ-OAR-2009-0472-0169. Accessed on the Internet on
May 30, 2009 at: http://www.a-sp.org/database/custom/Mass%20Compounding%20-%20Final%20Report.pdf.
80	"Benefit Analysis: Use of Aluminum Structures in Conjunction with Alternative Powertrain Technologies in
Automobiles," Bull, M. Chavali, R., Mascarin, A., Aluminum Association Research Report, May 2008, Docket
EPA-HQ-OAR-2009-0472-0168. Accessed on the Internet on April 30, 2009 at:
http://www.autoaluminum.org/downloads/IBIS-Powertrain-Study.pdf.
81	American Trucking Association. Last viewed on January 29, 2010 at
http://www.trucksdeliver.org/recommendations/speed-limits.html.
82	U.S. EPA SmartWay Transport Partnership. Last viewed on January 28, 2010 at
http://www3.epa.gov/smartway/transport/documents/tech/reducedspeed.pdf.
83	Department for Transport, Vehicle and Operator Services Agency. Last viewed on January 6, 2010 at
http://www.dft.gov.uk/vosa/newsandevents/pressreleases/2006pressreleases/28-12-06speedlimiterlegislation. htnx
84	Transport Canada. Summary Report - Assessment of a Heavy Truck Speed Limiter Requirement in Canada. Last
viewed on January 6, 2010 at http://www.tc.gc.ca/eng/roadsafety/tp-tpl4808-menu-370.htnx
85	See TIAX 2009, Note 2, at page 4-98.
86	Gaines, L. and D. Santini. Argonne National Laboratory, Economic Analysis of Commercial Idling Reduction
Technologies.
87	Gaines, L. and D. Santini. Argonne National Laboratory, Economic Analysis of Commercial Idling Reduction
Technologies.
88ICF International. Investigation of Costs for Strategies to Reduce Greenhouse Gas Emissions for Heavy-Duty On-
Road Vehicles. July 2010. Docket Identification Number EPA-HQ-OAR-2010-0162-0044.
89	National Academy of Science. Technologies and Approaches to Reducing the Fuel Consumption of Medium- and
Heavy-Duty Vehicles. March 2010. Page 124.
90	Gaines, L., A. Vyas, J. Anderson. Estimation of Fuel Use by Idling Commercial Trucks. 2006. Page 6.
91	Brodrick, C., T.Lipman, M. Farshchi, H. Dwyer, S. Gouse III, D.B. Harris, and F.King, Jr. Potential Benefits of
Utilizing Fuel Cell Auxiliary Power Units in Lieu of Heavy-Duty Truck Engine Idling. 2001. Page 3.
92	Lim, Han. Study of Exhaust Emissions from Idling Heavy-Duty Diesel Trucks and Commercially Available Idle-
Reducing Devices. EPA420-R-02-052. 2002. Page 2.
93	Kahn, ABM, N. Clark, G. Thompson, W.S. Wayne, M. Gautam, and D. Lyons. Idle Emissions from Heavy-Duty
Diesel Vehicles: Review and Recent Data. 2006. Journal of Air and Waste Management Association. Page 1405.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
94NACFE, June 2014. Confidence Report: Idle-Reduction Solutions. Available at
http://www.carbonwarroom.com/sites/default/files/reports/Idle-Reduction_Confidence_Report.pdf (accessed
November 2014).
95	See Vanner battery-inverter Systems at http://www.vanner.com/.
96	See eNow solar systems, http://www.enowenergy.com/.
97	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.
98	Vincent, R., Cleary, K., Ayala, A., Corey, R. 2004. "Emissions of HFC-134a from Light-Duty Vehicles in
California." SAE 2004-01-2256.
99	Society of Automotive Engineers, "IMAC Team 1 - Refrigerant Leakage Reduction, Final Report to Sponsors,"
2006.
100	California Air Resources Board. Letter from Michael Carter to Matthew Spears dated December 3, 2014. Solar
Control: Heavy-Duty Vehicles White Paper. Docket EPA-HA-OAR-2014-0827.
101	National Renewal Energy Laboratory. Reducing Long-Haul Truck Idle Loads and Resulting Fuel Use.
Presented by Jason A. Lustbader to EPA on July 29, 2014. See Docket EPA-HA-OAR-2014-0827.
102	See NHTSA Technology Report #1 (2015), Note 6.
103	U.S. EPA and NHTSA, "Final Rulemaking to Establish Light-Duty Vehicle Greenhouse Gas Emission Standards
and Corporate Average Fuel Economy Standards - Joint Technical Support Document," 2010. Last viewed on June
3, 2010 at http://www3.epa.gov/otaq/climate/regulations/420rl0901.pdf.
104	"Light-Duty Automotive Technology, Carbon Dioxide Emissions, and Fuel Economy Trends: 1975 - 2014,"
EPA-420-R-14-023, October 2014. Available at http://www3.epa.gov/otaq/fetrends.htm (last accessed October 31,
2014).
105	"Development and Optimization of the Ford 3.5L V6 EcoBoost Combustion System," Yi, J., Wooldridge, S.,
Coulson, G., Hilditch, J. Iyer, C.O., Moilanen, P., Papaioannou, G., Reiche, D. Shelby, M., VanDerWege, B.,
Weaver, C. Xu, Z., Davis, G., Hinds, B. Schamel, A. SAE Technical Paper No. 2009-01-1494, 2009, Docket EPA-
HQ-OAR-2009-0472-2860.
106	David Woldring and Tilo Landenfeld of Bosch, and Mark J. Christie of Ricardo, "DI Boost: Application of a
High Performance Gasoline Direct Injection Concept," SAE 2007-01-1410. Available at
http://www.sae.org/technical/papers/2007-01-1410 (last accessed Nov. 9, 2008).
107	Yves Boccadoro, Loi'c Kermanac'h, Laurent Siauve, and Jean-Michel Vincent, Renault Powertrain Division,
"The New Renault TCE 1.2L Turbocharged Gasoline Engine," 28th Vienna Motor Symposium, April 2007.
108	Tobias Heiter, Matthias Philipp, Robert Bosch, "Gasoline Direct Injection: Is There a Simplified, Cost-Optimal
System Approach for an Attractive Future of Gasoline Engines?" AVL Engine & Environment Conference,
September 2005.
109	U.S. EPA and NHTSA, "Final Rulemaking to Establish Light-Duty Vehicle Greenhouse Gas Emission Standards
and Corporate Average Fuel Economy Standards - Joint Technical Support Document," 2010. Last viewed on June
3, 2010 at http://www3.epa.gOv/otaq/climate/regulations/420rl0901.pdf.
no "All-new-Ford-engineered, Ford-tested, Ford-built diesel maximizes 2011 Super Duty productivity," Ford press
release, August 3, 2010. Available at: http://www.media.ford.com (last accessed June 27, 2011).
111	Stanton, Donald. "Enabling High Efficiency Clean Combustion." 2009 Semi-Mega Merit Review of the
Department of Energy. May 21, 2009. Last accessed on August 25, 2010 at
http://wwwl.eere.energy.gOv/vehiclesandfuels/pdfs/merit_review_2009/advanced_combustion/ace_40_stanton.pdf.
112	Committee on the Assessment of Technologies for Improving Light-Duty Vehicle Fuel Economy; National
Research Council, "Assessment of Fuel Economy Technologies for Light-Duty Vehicles," 2011. Available at
http://www.nap.edu/catalog.php?record_id=12924 (last accessed Jun 27, 2012).
113	SAE World Congress, "Focus B-pillar 'tailor rolled' to 8 different thicknesses," Feb. 24, 2010. Available at
http://www.sae.org/mags/AEI/7695 (last accessed Jun. 10, 2012).
114	http://www.techtimes.com/articles/87961/20150925/ford-s-2017-f-25Q-super-dutv-with-an-aluminum-bodv-is-
the-toughest-smartest-and-most-capable-super-dutv-ever.htm. September 25, 2015.
115	https ://www.ford.com/trucks/superduty/2017/.
116	"2008/9 Blueprint for Sustainability," Ford Motor Company. Available at: http://
www.ford.com/go/sustainability (last accessed February 8, 2010).
117	"2015 North American Light Vehicle Aluminum Content Study - Executive Summary," June 2014,
http://www.drivealuminum.org/research-resources/PDF/Research/2014/2014-ducker-report (last accessed February
26, 2015).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
118	http://www.foxnews.eom/leisure/2014/09/30/ford-confirms-increased-aluminum-use-on-next-gen-super-duty-
pickups/.
119	See EPA's heavy-duty engine certification database at http://www3.epa.gOv/otaq/certdata.htm#largeng.
120	See Phase 1 Federal Register at 76 FR 57231.
121	H. Zhang, J, Sanchez, M, Spears, "Alternative Heavy-duty Engine Test Procedure for Full Vehicle Certification,".
SAE Int. J. Commer. Veh. 8(2): 2015, doi: 10.4271/2015-01-2768,
122	EPA Docket Memo, Fleet Average Fuel Maps Projected for HD Phase 2 Vehicles, July 2016.
123	Cummins, 'Cummins visit on Phase 2 Engine standard," Greenhouse Gas Emissions Standards and Fuel
Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles - Phase 2 - Docket EPA-HQ-OAR-2014-
0827.
124	NACFE. Confidence Report: Programmable Engine Parameters. February 2015. Page 23.
125	Ostrander, Robert, et.al. (Meritor). Understanding the Effects of Engine Downspeeding on Drivetrain
Components. 2014. Page 2.
126	NACFE. Confidence Report for Idle Reduction Technologies.
127	Southwest Research Institute. Aerodynamic Test Report. July 2016. Docket EPA-HQ-OAR-2014-0827.
128	"Aerodynamic data from EPA's wind tunnel tests performed at Auto Research Center," Supplemental
Aerodynamic Data from EPA Testing. Available in the docket to this rulemaking: EPA-HQ-OAR-2014- 0827-
1624.
129	Delgado, Oscar. N. Lutsey. Advanced Tractor-Trailer Efficiency Technology Potential in the 2020-2030
Timeframe. April 2015. Docket EPA-HQ-OAR-2014-0827.
130	Delgado, Oscar. N. Lutsey. Advanced Tractor-Trailer Efficiency Technology Potential in the 2020-2030
Timeframe. April 2015. Docket EPA-HQ-OAR-2014-0827.
131	U.S. EPA. Memo to Docket. Coefficient of Rolling Resistance and Coefficient of Drag Certification Data for
Tractors. See Docket EPA-HQ-OAR-2014-0827.
132	Delgado, Oscar. N. Lutsey. Advanced Tractor-Trailer Efficiency Technology Potential in the 2020-2030
Timeframe. April 2015. Docket EPA-HQ-OAR-2014-0827.
133	North American Council for Freight Efficiency. Confidence Report: Electronically Controlled Transmissions.
December 2014.
134	Oak Ridge National Laboratory. "Powertrain Test Procedure Development for EPA GHG Certification of
Medium- and Heavy-Duty Engines and Vehicles." July 2016. Docket # EPA-HQ-OAR-2014-0827.
135	Stoltz, T and Dorobantu, M. Transmission Potential to Contribute to CO2 Reduction: 2020 and Beyond Line Haul
Perspective. ACEEE/ICCT Workshop on Emerging Technologies for Heavy-Duty Fuel Efficiency. July 2014.
136	U.S. EPA. Memorandum to the Docket. "Effectiveness of Technology to Increase Transmission Efficiency."
July 2016. EPA-HQ-OAR-2014-0827.
137	Ibid.
138	See the 2010 NAS Report, Note 1, page 67.
139	U.S. EPA. Memorandum to the Docket. "Effectiveness of Technology to Increase Axle Efficiency." July 2016.
EPA Docket # EPA-HQ-OAR-2014-0827.
140	Delgado, Oscar. N. Lutsey. "Advanced Tractor-Trailer Efficiency Technology Potential in the 2020-2030
Timeframe." April 2015. EPA Docket EPA-HQ-OAR-2014.0827.
141	U.S. Department of Energy. Transportation Energy Data Book, Edition 28-2009. Table 5.7.
142	U.S. EPA. Memo to Docket. Coefficient of Rolling Resistance and Coefficient of Drag Certification Data for
Tractors. See Docket EPA-HQ-OAR-2014-0827.
143	Memo to Docket. Coefficient of Rolling Resistance and Coefficient of Drag Certification Data for Tractors.
Docket EPA-HQ-OAR-2014-0827.
144	North American Council for Freight Efficiency. Confidence Report: Idle- Reduction Solutions. 2014. Page 13.
145	EPA Docket Memo, Fleet Average Fuel Maps Projected for HD Phase 2 Vehicles, July 2016.
146	Michael Ross, Validation Testing for Phase 2 Greenhouse Gas Test Procedures and the Greenhouse Gas
Emission Model (GEM) for Medium and Heavy-Duty Engines and Powertrains, Final Report to EPA, Southwest
Research Institute, found in docket of this rulemaking, EPA-HQ-OAR-2014-0827, June, 2016.
147	Joint Technical Support Document: Final Rulemaking for 2017-2025 Light-Duty Vehicle Greenhouse Gas
Emission Standards and Corporate Average Fuel Economy Standards, August 2012, available at
http://www3 .epa.gov/otaq/climate/documents/420rl2901 .pdf.
148	National Renewable Energy Laboratory July 2016, "The Development of Vocational Vehicle Drive Cycles and
Segmentation," NREL/TP-5400-65921.
149	See memorandum dated July 2016 titled, "Summary of Comments on Vocational Vehicle Baselines."

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
150	See Cummins-Eaton partnership at http://smartadvantagepowertrain.com/.
151	See TIAX 2009, Note 2, Table 4-48.
152	See http://www.truckinginfo.eom/channel/equipment/article/story/2014/10/2015-medium-duty-trucks-the-
vehicles-and-trends-to-look-for/page/3.aspx (downloaded November 2014).
153	For example, see XL Hybrids at http://www.xlhybrids.com/content/assets/Uploads/XL-BoxTruck-US-FLY-
8.5x11-0519-LR.pdf, and Crosspoint Kinetics at http://crosspointkinetics.com/members/kinetics-hybrid-partners
154	See spreadsheet file titled, "HD GHG Simple Hybrid Model v7.xlsx".
155	Green Fleet Magazine, The Latest Developments in EV Battery Technology, November 2013, available at
http://www.greenfleetmagazine.eom/article/story/2013/12/the-latest-developments-in-ev-battery-technology-
grn/page/l.aspx.
156	See memorandum on axle efficiency improvements, Note 139.
157	See Argonne National Laboratory 2009 report, page 91.
158	See memorandum dated May 2016 on Vocational Vehicle Tire Rolling Resistance Certification Data.
159	See NREL data at http://www.nrel.gov/vehiclesandfuels/fleettest/research_fleet_dna.htmL
160	See spreadsheet file titled, "Vocational-Standards GEMpostprocess.xlsx
161	See Cummins maintenance schedule, available at
http://www.cumminsbridgeway.com/pdf/parts/Recommended_Maintenance_Schedule.pdf (accessed March 2016).
162	NTEA, 2015 Work Truck Electrification and Idle Management Study.
163	See TIAX 2009, Note 2.
164	Morton, C. and Spargo, C.M., IET Electrical Systems in Transportation, Electrified hydraulic power steering
system in hybrid electric heavy trucks, Sept 2014, accessed June 2016 from
https://www.researchgate.net/publication/264983979_Electrified_hydraulic_power_steering_system_in_hybrid_elec
tricheavytrucks.
165	TIAX 2009, Note 2 pp. 3-5.
166	The Minnesota refrigerant leakage data can be found at
http://www.pca. state.mn.us/climatechange/mobileair.html#leakdata.
167	See Phase 1 RIA, Chapter 2.7.
168	Society of Automotive Engineers, "IMAC Team 1 - Refrigerant Leakage Reduction, Final Report to Sponsors,"
2006.
169	Society of Automotive Engineers Surface Vehicle Standard J2727, issued August 2008,
http://www.sae.org.
170	See Attachment 1 of public comments from CARB, "Aerodynamic Drag Reduction Technologies Testing for
Heavy- Duty Vocational Vehicles— Preliminary Results, July 2015, NREL/TP-5400-64610.
171	National Renewable Energy Laboratory July 2016, "Characterization of PTO and Idle Behavior for Utility
Vehicles" NREL/TP-5400-66747.
172	See http://westcoastcollaborative.Org/files/sector-fleets/WCC-LA-BEVBusinessCase2011-08-15 .pdf
173	Silver, Fred, and Brotherton, Tom. (CalHEAT) Research and Market Transformation Roadmap to 2020 for
Medium- and Heavy-Duty Trucks. California Energy Commission, June 2013.
174	Gallo, Jean-Baptiste, and Jasna Tomic (CalHEAT). 2013. Battery Electric Parcel Delivery Truck Testing and
Demonstration. California Energy Commission.
175	See memorandum titled, Vocational Vehicle Tire Rolling Resistance Certification Data.
176	American Public Transportation Association, "An Analysis of Transit Bus Axle Weight Issues", November 2014.
177	See spreadsheet file dated July 2016 titled, "FRM_Vocational-Standards_GEMpostprocess.xls.
178	Final Rulemaking to Establish Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium-
and Heavy-Duty Engines and Vehicles: Regulatory Impact Analysis, Environmental Protection Agency, Page 3-42.
Available at: https://www3 .epa.gov/otaq/climate/documents/420rl 1901 .pdf.
179	"Improving the Aerodynamic Efficiency of Heavy-Duty Vehicles - Wind Tunnel Test Results of Trailer-Based
Drag Reduction Technologies," McAuliffe, Brian R., National Research Council Canada, Report #: LTR-AL-2015-
0272. Available online: https://www.tc.gc.ca/eng/programs/environment-etv-menu-eng-2980.html.
180	"Aerodynamic data from EPA's wind tunnel tests performed at Auto Research Center," Supplemental
Aerodynamic Data from EPA Testing. Available in the docket to this rulemaking: EPA-HQ-OAR-2014- 0827-
1624.
181	"Aerodynamic data from EPA's wind tunnel tests performed at Auto Research Center," Supplemental
Aerodynamic Data from EPA Testing. Available in the docket to this rulemaking: EPA-HQ-OAR-2014- 0827-
1624.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
182	"Tires & Truck Fuel Economy: A New Perspective," The Tire Topic Magazine, Special Edition Four, 2008,
Bridgestone Firestone, North American Tire, LLC. Available online:
http://www.trucktires.com/bridgestone/us_eng/brochures/pdf/08-Tires_and_Truck_Fuel_Economy.pdf.
183	Truck Trailer Manufacturers Association letter to EPA. October 16, 2014. Docket EPA-HQ-OAR-2014-0827.
184	"T^e pressure Systems - Confidence Report." North American Council for Freight Efficiency. 2013. Available
online: http://nacfe.org/wp-content/uploads/2014/01/TPS-Detailed-Confidence-Reportl.pdf.
185	"A Day in the Life of a Tire," Pressure Systems International, Presented to EPA on August 20, 2014.
186	Federal Motor Carriers Safety Administration, "Commercial Vehicle Tire Condition Sensors," conducted by
Booz-Allen-Hamilton, Inc. November, 2003.
187	TMC Technology & Maintenance Council, "TMC Tire Air Pressure Study," May 2002.
188	FMCSA "Advanced Sensors and Applications: Commercial Motor Vehicle Tire Pressure Monitoring and
Maintenance," February 2014.
189	"Underinflated Commercial Low Rolling Resistance Truck Tires & It's Impact on Fuel Economy," October
2010. Pressure Systems International, Presentation Material shared with EPA on May 21, 2014.
190	Scarcelli, Jamie. "Fuel Efficiency for Trailers" Presented at ACEEE/ICCT Workshop: Emerging Technologies
for Heavy-Duty Vehicle Fuel Efficiency, Wabash National Corporation. July 22, 2014.
191	"Weight Reduction: A Glance at Clean Freight Strategies," EPA SmartWay. EPA420F09-043. Available at:
http://permanent.access.gpo.gov/gpo38937/EPA420F09-043.pdf.
192	Memo to docket regarding confidential weight reduction information obtained during SBREFA Panel, June 4,
2015.
193	Randall Scheps, Aluminum Association, "The Aluminum Advantage: Exploring Commercial Vehicles
Applications," presented in Ann Arbor, Michigan, June 18, 2009.
194ICF International. Investigation of Costs for Strategies to Reduce Greenhouse Gas Emissions for Heavy-Duty
On-Road Vehicles. July 2010. See also, TIAX, LLC. "Assessment of Fuel Economy Technologies forMedium-
and Heavy-Duty Vehicles," Final Report to the National Academy of Sciences, November 19, 2009.
195	Schubert, R., Chan, M., Law, K. (2015). Commercial Medium- and Heavy-Duty (MD/HD) Truck Fuel Efficiency
Cost Study. Washington, DC: National Highway Traffic Safety Administration, Tetra Tech Technology Cost Report
(2015).
196	Schubert, R., Chan, M., Law, K. (2015). Commercial Medium- and Heavy-Duty Truck Fuel Efficiency
Technology Cost Study. Washington, DC: National Highway Traffic Safety Administration.
197	A. Rogozhin et al., Int. J. Production Economics 124 (2010) 360-368.
198	RTI International. Heavy Duty Truck Retail Price Equivalent and Indirect Cost Multipliers. July 2010.
199	Rogozhin, A., et al., "Using indirect cost multipliers to estimate the total cost of adding new technology in the
automobile industry," International Journal of Production Economics (2009), doi:10.1016/j.ijpe.2009.11.031.
200	Helfand, Gloria, and Todd Sherwood, "Documentation of the Development of Indirect Cost Multipliers for Three
Automotive Technologies," August 2009.
201	RTI International. Heavy Duty Truck Retail Price Equivalent and Indirect Cost Multipliers. July 2010.
202	See "Learning Curves in Manufacturing," L. Argote and D. Epple, Science, Volume 247; "Toward Cost Buy
down Via Learning-by-Doing for Environmental Energy Technologies, R. Williams, Princeton University,
Workshop on Learning-by-Doing in Energy Technologies, June 2003; "Industry Learning Environmental and the
Heterogeneity of Firm Performance, N. Balasubramanian and M. Lieberman, UCLA Anderson School of
Management, December 2006, Discussion Papers, Center for Economic Studies, Washington DC.
203	See "Learning Curves in Manufacturing," L. Argote and D. Epple, Science, Volume 247; "Toward Cost Buy
down Via Learning-by-Doing for Environmental Energy Technologies, R. Williams, Princeton University,
Workshop on Learning-by-Doing in Energy Technologies, June 2003; "Industry Learning Environments and the
Heterogeneity of Firm Performance, N. Balasubramanian and M. Lieberman, UCLA Anderson School of
Management, December 2006, Discussion Papers, Center for Economic Studies, Washington DC.
204	U.S. Energy Information Administration, Annual Energy Outlook 2014, Early Release; Report Number
DOE/EIA-0383ER (2014), December 16, 2013.
205	Bureau of Economic Analysis, Table 1.1.9 Implicit Price Deflators for Gross Domestic Product; as revised on
March 27, 2014.
206	Reducing Heavy-Duty Long Haul Combination Truck Fuel Consumption and CO2 Emissions; prepared by
Northeast States Center for a Clean Air Future (NESCCAF), International Council on Clean Transportation (ICCT),
Southwest Research Institute, TIAX, LLC; Final Report October 2009.
207	Comments to the docket from Allison Transmission, at I.D.2.a, page 6.
208	Comments to the docket from Allison Transmission, at I.D.2.a, page 6.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
209	Ryder website at http://www.ryderfleetproducts.com/tpms-accessories-c-l 1434; a PDF version of the site has
been placed in the docket.
210	Ben Sharpe (ICCT) and Mike Roeth (North American Council for Freight Efficiency), "Costs and Adoption
Rates of Fuel-Saving Technologies for Trailer in the North American On-Road Freight Sector," Feb 2014.

-------
Chapter 3: Test Procedures
Test procedures are a crucial aspect of the heavy-duty vehicle GHG and fuel
consumption program. This rulemaking establishes several new test procedures to be used as
part of compliance process for both engine and vehicle compliance. Specifically, these test
procedures are used to generate inputs to GEM. This chapter will describe the development
process for the test procedures, including the assessment of engines, aerodynamics, rolling
resistance, chassis dynamometer testing, powertrain testing, and duty cycles. The final
subsection of this chapter (3.10) describes the chassis test procedure used to verify compliance
with the standards for heavy duty pickups and vans.
This section focuses on the actual measurements procedures and generally does not
address how manufacturers will use this data to certify their engines and vehicles. For example,
Chapter 3.2 below discusses how to measure aerodynamic drag, but does not detail how
manufacturers will use the data to develop GEM aerodynamic inputs for certification.
3.1 Heavy-Duty Engine Test Procedure
The agencies are controlling heavy-duty engine fuel consumption and greenhouse gas
emissions through the use of engine certification. The program will mirror existing engine
regulations for the control of both GHG and non-GHG pollutants in many aspects. The
following sections provide an overview of the test procedures.
3.1.1 Existing Regulation Reference
Heavy-duty engines currently are certified for GHG and non-GHG pollutants using test
procedures developed by EPA. The Heavy-Duty Federal Test Procedure (FTP) is a transient test
consisting of second-by-second sequences of engine speed and torque pairs with values given in
normalized percent of maximum form. The cycle was computer generated from a dataset of 88
heavy-duty trucks in urban operation in New York and Los Angeles. These procedures are well-
defined, mirror in-use operating parameters, and thus we believe appropriate also for the
assessment of GHG emissions from heavy duty engines. Further, EPA is concerned that we
maintain a regulatory relationship between the non-GHG emissions and GHG emissions,
especially for control of CO2 and NOx. Therefore, the agencies will continue using the same
criteria pollutant test procedures for both the CO2 and fuel consumption standards.
For 2007 and later Heavy-Duty engines, 40 CFR parts 86 - "Control of Emissions from
New and In-Use Highway Vehicles and Engines" and 1065 - "Engine Testing Procedures" detail
the certification process. 40 CFR 86.007-11 defines the standard settings of Oxides of Nitrogen,
Non-Methane Hydrocarbons, Carbon Monoxide, and Particulate Matter. The duty cycles are
defined in 40 CFR part 86. The Federal Test Procedure engine test cycle is defined in 40 CFR
part 86 Appendix I. The Supplemental Emissions Test engine cycle is defined in 40 CFR
86.1360(b). All emission measurements and calculations are defined in 40 CFR part 1065, with
exceptions as noted in 40 CFR 86.007-11. The data requirements are defined in 40 CFR 86.001-
23 and 40 CFR 1065.695.

-------
The measurement method for CO2 is described in 40 CFR 1065.250. For measurement
of CH4 refer to 40 CFR 1065.260. For measurement of N2O refer to 40 CFR 1065.275. We
recommend that you use an analyzer that meets performance specifications shown in Table 1 of
40 CFR 1065.205. Note that your system must meet the linearity verification of 40 CFR
1065.307. To calculate the brake specific mass emissions for CO2, CH4 and N2O refer to 40
CFR 1065.650.
3.1.2 Engine Dynamometer Test Procedure Modifications
3.1.2.1	Fuel Consumption Calculation
EPA and NHTSA will calculate fuel consumption, as defined as gallons per brake
horsepower-hour, from the CO2 measurement, just as in the Phase 1 rule. The agencies are
continuing to use 8,887 grams of CO2 per gallon of gasoline and 10,180 g CO2 per gallon of
diesel fuel.
3.1.2.2	Regeneration Impact on Fuel Consumption and CO2 Emissions
The current engine test procedures also require the development of regeneration emission
rate and frequency factors to account for the emission changes during a regeneration event.1 In
Phase 1, the agencies adopted provisions to exclude CO2 emissions and fuel consumption due to
regeneration. However, for Phase 2, we will include CO2 emissions and fuel consumption due to
regeneration over the FTP and RMC cycles as determined using the infrequently regenerating
aftertreatment devices (IRAF) provisions in 40 CFR 1065.680. However, we are not finalizing
the inclusion of fuel consumption due to regeneration in the creation of the steady-state and cycle
average fuel maps used in GEM for vehicle compliance. Our assessment of the current non-
GHG regulatory program indicates that engine manufacturers have significantly reduced the
frequency of regeneration events. In addition, market forces already exist which create
incentives to reduce fuel consumption during regeneration.
3.1.2.3	Fuel Heating Value Correction
In the Phase 1 rule, the agencies collected baseline CO2 performance of diesel engines
from testing which used fuels with similar properties. The agencies will continue using a fuel-
specific correction factor for the fuel's energy content. This maintains consistency between test
labs, as well as prevents potential fuel changes that could occur in the future from changing the
effective stringency of the Phase 2 standards. The agencies found the average energy content of
the diesel fuel used at EPA's National Vehicle Fuel and Emissions Laboratory was 21,200 BTU
per pound of carbon. This value was determined by dividing the Net Heating Value (BTU per
pound) by the carbon weight fraction of the fuel used in testing. We will continue using the
Phase 1 corrections for diesel fuel, gasoline, natural gas, and liquid petroleum gas in 40 CFR
1036.530. We will also expand the table by adding dimethyl ether.
In addition to the fuel heating value correction, we are finalizing the addition of reference
carbon mass fraction values for these fuels to the Table 1 of 40 CFR 1036.530. These reference
values are used in the powertrain calculations 40 CFR 1037.550, steady-state engine fuel
mapping and fuel consumption at idle in 40 CFR 1036.535, and cycle average engine fuel

-------
mapping in 40 CFR 1036.540 to account for the difference in carbon mass fraction between the
test fuel and the reference fuel prior to correcting for the test fuel's mass-specific net energy
content.
The agencies are not finalizing fuel corrections for alcohols because the fuel chemistry is
homogeneous.
3.1.2.4	Urea Derived CO2 Correction
The agencies will allow manufacturers to correct compression ignition engine and
powertrain CO2 emission results (for engines utilizing urea SCR for NOx control) to account for
the contribution of urea derived CO2 emissions to the total engine CO2 emissions.
Urea derived CO2 can account for up to 1 percent of the total CO2 emissions. Urea is
produced from gaseous NH3 and gaseous CO2 that is captured from the atmosphere, thus CO2
derived from urea decomposition in diesel SCR emission control systems results in a net
emission of zero CO2 to the environment. In our test procedures for Phase 2, we allow
manufacturers to determine CO2 emissions either by measuring the CO2 emitted from the engine
or to determine it by measuring fuel flow rate during the test. If we do not allow for correction
of the urea derived CO2 emissions, this will result in a positive CO2 bias for CO2 emissions
determined by measuring the CO2 emitted from the engine. To perform this correction, we are
allowing you to determine the mass rate of urea injected over the duty cycle from the engine's
J1939 CAN signal or you may measure urea flow rate independently using good engineering
judgment. This value is used as an input to an equation that allows you to determine the mass
rate of CO2 from urea during the duty cycle. This resulting CO2 mass emission rate value is then
used as an input to the steady-state engine fuel map and engine fuel consumption at idle fuel
mass flow rate calculation in 40 CFR 1036.535, the cycle average engine fuel map calculation in
40 CFR 1036.540, and the total mass of CO2 emissions over the duty cycle calculations in 40
CFR 1037.550. Note that this correction is only allowed for CO2 measured from the engine and
not CO2 derived from fuel flow measurement.
The calculation for determination of the mass rate of CO2 from urea requires the user to
input the urea solution urea percent by mass. This calculation uses prescribed molecular weights
for CO2 and urea as given in 40 CFR 1065.1005 of 44.0095 and 60.05526 respectively. A 1:1
molar ratio of urea reactant to CO2 product is assumed.
To facilitate the ability of the agencies to make this correction, we are requiring that the
urea mass flow rate be broadcasted on the non-proprietary J1939 PGN (Parameter Group
Number) 61475 (and 61478 if applicable).
3.1.2.5	Multiple Fuel Maps
Engine manufacturers are being required to certify fuel maps to enable vehicle
manufacturers to run GEM for each vehicle configuration. However, modern heavy-duty
engines often have multiple fuel maps, commonly meant to improve performance or fuel
efficiency under certain operating conditions. CO2 emissions can also be different depending on
which map is tested, so it is important to specify a procedure to properly deal with engines with

-------
multiple fuel maps. Consistent with criteria-pollutant emissions certification, engine
manufacturers will be required to address this during certification, either by declaring worst case
maps that cover more than one in-use map, or by submitting multiple fuel maps. The agencies
may require the manufacturer to include other fuel map information, such as when the conditions
under which a given fuel map is used {i.e. transmission gear, vehicle speed, etc.).
3.1.2.6 Measuring GEM Engine Inputs
To recognize the contribution of the engine in GEM, the engine fuel map, full load torque
curve and motoring torque curve have to be input into GEM. To insure the robustness of each of
those inputs, a standard procedure has to be followed. Both the full load and motoring torque
curve procedures are already defined in 40 CFR part 1065 subpart F for engine testing.
However, the fuel mapping procedures we are finalizing are new. The agencies have compared
the new procedures to other accepted engine mapping procedures with a number of engines at
various labs including EPA's NVFEL, Southwest Research Institute, and Environment Canada's
laboratory. The procedure was selected because it proved to be accurate and repeatable, while
limiting the test burden to create the fuel map. This provision is consistent with NAS's
recommendation (3.8).
The agencies are requiring that engine manufacturers must certify fuel maps as part of
their certification to the engine standards, and that they provide those maps to vehicle
manufacturers. These maps consist of steady-state and cycle average fuel maps. The one
exception to this requirement would be for cases in which the engine manufacturer certifies
based on powertrain testing, as described in Chapter 3.6. In such cases, engine manufacturers
would not be required to also certify the otherwise applicable fuel maps. We are not allowing
vehicle manufacturers to develop their own fuel maps for engines they do not manufacture.
In addition to the steady-state engine fuel map procedure for cruise cycles the agencies
are also requiring use of the cycle-average engine map test procedure for the transient duty-cycle
as defined in 40 CFR 1036.540. The cycle-average approach can optionally be used in place of
the steady-state fuel maps by performing cycle-average testing over the cruise cycles. The
NPRM to this rule, along with the two journal publications, one from the US EPA and one
authored by an industry group, discussed in length the benefits of this test procedure.2'3 The
benefits ranged from capturing transient fueling to protecting intellectual property. We have
tested four different engines with two different engine ratings for each engine since the proposal.
The results of these tests confirmed our earlier findings that the cycle average engine test
procedure is much more accurate than the steady-state mapping procedure with respect to
representing the engine over transient engine operation. The results also showed that the cycle
average engine map can be applied to the cruise cycles but required that the agencies update the
test points to ensure that overlap doesn't occur. Overlap happens when the lower axle ratio
causes the vehicle to operate in the next lowest gear at increased engine speed. The agencies
updated the test points in 40 CFR 1037.540 to address the overlap issue. The agencies are
finalizing the requirement to use the steady-state engine procedure over the cruise cycles and the
cycle average engine map procedure for the transient cycle (optional for cruise).
Along with testing additional engines, the agencies have done significant work to define
the mathematical form the cycle average engine map data should take in GEM. The first

-------
approach the agencies evaluated was an interpolation and extrapolation scheme.4 Since then we
have looked at many different least square fits of the data using different dependent (fuel mass
and BSFC) and independent (average engine speed, average engine torque, average engine speed
divided by average vehicle speed (N/V) and positive cycle work) variables. The results of this
work showed that the cycle average map is most accurately described with fuel mass as the
dependent variable and N/V and positive work as the independent variables. The form of the
equation is fuel mass ~ 1 + N/V + W.
3.1.3 Engine Family Definition and Test Engine Selection
3.1.3.1	Criteria for Engine Families
The current regulations outline the criteria for grouping engine models into engine
families sharing similar emission characteristics. A few of these defining criteria include bore-
center dimensions, cylinder block configuration, valve configuration, and combustion cycle; a
comprehensive list can be found in 40 CFR 86.096-24(a)(2). While this set of criteria was
developed with criteria pollutant emissions in mind, similar effects on CO2 emissions can be
expected. For this reason, this methodology should continue to be followed when considering
CO2 emissions, just as it was in the Phase 1 rules.
3.1.3.2	Emissions Test Engine
Manufacturers must select at least one engine per engine family for emission testing. The
methodology for selecting the test engine(s) should be consistent with 40 CFR 86.096-24(b)(2)
(for heavy-duty Otto cycle engines) and 40 CFR 86.096-24(b)(3) (for heavy-duty diesel engines).
An inherent characteristic of these methodologies is selecting the engine with the highest fuel
feed per stroke (primarily at the speed of maximum rated torque and secondarily at rated speed)
as the test engine, as this is expected to produce the worst-case criteria pollutant emissions. To
be consistent, it is recommended that the same methodology continue to be used for selecting test
engines.
3.2 Aerodynamic Assessment
For the Phase 1 rule, the agencies promulgated requirements whereby the coefficient of
drag assessment was a product of test data and modeling using good engineering judgment. A
group of aerodynamic bins for tractors corresponding to certain known aerodynamic design
features (e.g., Classic, Conventional, SmartWay, etc.) were established based on the results of an
agency sponsored aerodynamic assessment test program. The rules require tractor manufacturers
to take the aerodynamic test result from a tractor and determine the tractor's appropriate bin. To
ensure the consistency of the drag assessment results, certain aspects of the truck were defined,
including the trailer, location of payload, and tractor-trailer gap. In addition, the agencies
specified test procedures for aerodynamic assessment: coastdown testing (also used as the
reference method), wind tunnel testing (reduced and full scale), and computational fluid
dynamics (CFD). Constant speed testing was also permitted as an alternative test procedure, but
the agencies did not develop a specific procedure.

-------
For the Phase 2 rule, we are retaining many of the aspects of the aerodynamic assessment
protocols from Phase 1 with the following revisions and additions: enhancement of the analysis
methodology for the coastdown test procedure, which we will keep as the reference method for
the tractor program; inclusion of trailers in the aerodynamic assessment test protocols;
modifications to the standard trailer used for tractor aerodynamic assessment and establishing a
reference tractor for trailer aerodynamic assessment; and use of wind-average drag area (CdA wa)
as the required aerodynamic Greenhouse Gas Emissions Model (GEM) input for tractors.
Another modification to the aerodynamic assessment for Phase 2 is the use of drag area
(coefficient of drag multiplied by the frontal area, or CdA), rather than the coefficient of drag
(Cd), for tractor aerodynamic bin standards. Although this modification will not alter the
aerodynamic assessment protocols, it is important to note this since all Phase 2 aerodynamic
assessment results will be presented in this format, rather than the Cd format used for Phase 1.
The Phase 2 trailer program will also be in the wind-averaged drag area domain, instead of drag
coefficient. However, the trailer program will be based on a drag area reduction from a baseline
configuration.
3.2.1 Aerodynamics Baselines for Tractors
To establish GHG standards, the aerodynamic assessment methods and baselines needed
to be evaluated. A combination of coastdown, wind tunnel, CFD, and constant speed tests were
used to determine the wind-averaged drag performance of several sleeper cab and day cab
tractors. The coastdown was used as the reference method, due to its familiarity within the
industry and the ability to test a real full-scale truck instead of relying on scale models or
simulations, which would require simplifications to the vehicle geometry and other factors.
The agencies used a multistep process for determining baseline performance. First, we
evaluated which Phase 1 aerodynamic bin our test tractors were in by doing a Phase 1-style
analysis from coastdown tests in the Phase 1 trailer configuration. Then, we tested the same
tractors with trailer skirts (the Phase 2 trailer configuration) and analyzed the data using the
Phase 2 analysis procedure that is being finalized in this rulemaking. Finally, we translated this
Phase 2 coastdown result to a wind-averaged drag area value.
For this final step, the agencies conducted or obtained test and simulation data using a
variety of alternate aerodynamic methods: scale wind tunnels at Auto Research Center (ARC)
and National Research Council Canada (NRC), CFD using Navier-Stokes and Lattice-Boltzmann
codes, and constant speed on-road tests at Southwest Research Institute (SwRI) conducted with
the same vehicles used in the coastdown tests.5'6'7'8 Given that tunnels, simulations, and road
load tests are all approximations of aerodynamic performance, the agencies made an effort to
have multiple methods for a given tractor to the extent possible. The aerodynamic drag as a
function of yaw angle determined from these alternate methods was used to adjust the coastdown
results to a wind-averaged drag area value.
This analysis provided a basis to derive aerodynamic bins for Phase 2. By evaluating the
aerodynamic performance of tractors in both the Phase 1 and Phase 2 domains, we were able to
create numerical values for the Phase 2 aerodynamic bin boundaries by aligning the relative
aerodynamic performance from both test procedures.

-------
3.2.1.1 Coastdown Testing
During development and since the beginning of Phase 1 implementation, we received
persuasive suggestions for improving the coastdown test procedure analysis methodology to
reduce data post processing and improve data resolution. Accordingly, for Phase 2 aerodynamic
assessment methods, we modified the coastdown test procedure analysis methodology, made
changes to the specifications and protocols for conducting and analyzing the results of the
constant speed test procedure, and updated the conditions for performing CFD analysis.
Based on feedback from the heavy-duty vehicle manufacturing industry and other
entities, the agencies finalized a Modified SAE J1263 coastdown procedure in the Phase 1
rulemaking. During and since the finalization of those 1 regulations, stakeholders suggested
increasing accuracy and precision by analyzing portions of the data generated during coastdown
testing rather than the full data set. One OEM suggested the use of the high speed portion of the
coastdown test procedure speed range to solely or predominantly isolate the aerodynamic forces.
Another OEM suggested using the high speed and low speed portions of the coastdown test
procedure speed range in an iterative fashion to isolate the mechanical/frictional losses and
rolling resistance predominantly present at lower speeds and removing these forces from the
higher speed forces to capture predominantly aerodynamic forces. Comments to the Phase 2
NPRM indicated a preference to include a tire rolling resistance dependence on speed, which
was assumed to be zero in the proposal.
To develop baseline aerodynamic performance and refine the aerodynamic test
procedures, the agencies (via contractors ICF Corporation and SwRI) coasted down combination
tractors on Farm-to-Market Highway 70, a rural highway between Bishop, Texas and Chapman
Ranch, Texas. Testing was performed by SwRI. Filtered USGS elevation data were obtained for
the same stretch of roadway.9 The grade information was incorporated into our analysis. The
testing was conducted overnight, usually between 12 am and 4 am, to minimize traffic and wind.
To get a comprehensive data set to conduct various analysis techniques, the vehicles were
coasted down from 70 mph to 0 mph, where 14 to 20 runs were conducted for each test. Some
tests were conducted with only high-speed and low-speed coastdowns, where up to 32 runs were
conducted. An ultrasonic anemometer was mounted 0.85 m above the leading edge of the trailer
at the midpoint of the trailer width. This anemometer recorded air speed and direction onboard
the vehicle at 10 Hz. A weather station, which measured wind speed, wind direction,
temperature, and air pressure at 1 Hz, was placed alongside the road at the approximate midpoint
of the stretch of road being used for the tests. Details of the test setup and vehicle information
can be found in the on-road testing summary report from SwRI. The tractors that were used in
this analysis are represented in this chapter using the following numbers: Sleeper Cab tractors 1
through 5 and Day Cab tractors 20, 30, and 31.
The average and maximum wind speeds were calculated for each run to determine
validity of the run with respect the wind restrictions. Some tests were performed outside of these
specifications to assess the impact of wind variation. Table 3-1 below shows the ambient
conditions desired for each coastdown run within a coastdown test, which resembles the SAE
J1263 recommended practice.

-------
Table 3-1 Desired Ambient Conditions for Coastdown Tests
PARAMETER
LIMIT
Maximum average wind speed
lOmph
Maximum wind speed
12.3 mph
Maximum average cross wind component
5 mph
3.2.1.1.1 Phase 1 analysis
To first understand how our test vehicles performed, we conducted coastdown tests using
the Phase 1 trailer configuration and Phase 1 test and analysis procedure. Force was calculated
for every 10-Hz measurement. Grade force was calculated at every 10-Hz measurement and
incorporated into the force value. The data was not filtered. The regression was applied between
force and vehicle speed for the entire test (not run by run). The results are plotted against the
Phase 1 aerodynamic bin structure below. An additional tractor not included in the SwRI report,
Sleeper Cab 11, was included for reference. It was an identical model ("sister tractor") to
Sleeper Cab 1 and produced a similar drag area result. The Phase 1 C&A bins are superimposed
on the plots to show the aerodynamic levels of the various tractors. Every tractor tested in this
program (both sleeper cabs and day cabs) were in Bin III or Bin IV.
Sleeper cabs
7.0

Bin II
6.5
4

6.0-
5' •
•' °
Bin III
5.5

Bin IV
5.0-

Bin V


Figure 3-1 Drag Area Values by Truck Number from Sleeper Cab Tractors Using Phase 1 Analysis; With
Phase 1 Bin Boundaries

-------
7.0
6.5
CM
<
— 6.0
<
T3
O
5.5
5.0
Figure 3-2 Drag Area Values by Truck Number from Day Cab Tractors Using Phase 1 Analysis; With Phase
1 Bin Boundaries.
3.2.1.1.2 Phase 2 Analysis
3.2.1.1.2.1 Data filtering
In the analysis for the NPRM, air speed and vehicle speed data, collected at 10 Hz, were
filtered using a 1-second weighted centered moving average. Given that the coastdown analysis
procedure already involves averaging over certain vehicle speed intervals, instead of the moving
average filter, a different filtering scheme was used in this analysis only to remove outliers.
Based on feedback from the heavy-duty vehicle manufacturers, a moving median filter was used
to remove the outliers, which were defined as points differing by more than three standard
deviations from the three-second centered moving median. The standard deviation was
calculated as the 1.4826 times the median absolute deviation of the three-second window. The
outlier was then replaced by the median of the three-second window. This technique is
equivalent to the Hampel filter in Matlab.
This filter was not applied to the weather station measurements (wind speed and wind
direction), as these measurements were collected only at 1 Hz. However, we finalizing that the
wind speed and wind direction must be collected at 10 Hz to be consistent with the air speed and
vehicle speed measurement frequencies. We are finalizing that the Hampel filter described
above must be applied to vehicle speed, air speed, yaw angle, wind speed, and wind direction
measurements.
Day cabs

31



O

Bin III
\



Bin IV



Bin V





-------
3.2.1.1.2.2 Air Speed Measurements
The position of the onboard anemometer is such that the air speed readings need to be
corrected. Located above the trailer, the anemometer's air velocity measurements will typically
be greater than the free stream air speed. The roadside weather station was used to correct the
onboard air speed measurements, using the trigonometric calculations below. Yaw angles
counter-clockwise to the direction of travel were considered as positive. Yaw angles clockwise
to the direction of travel were considered as negative.
Figure 3-3 Diagram Of Vehicle Speed and Air Speed Vectors during Coastdowns in Opposite Directions for
a Given Vehicle Speed, Wind Speed and Wind Direction
Basic trigonometric relationships were used to calculate the theoretical air speed vr.th
from the vehicle speed and weather station measurements, as described in the equation below.
The vehicle travel direction affects the resultant vector, and is included as a ^veh value of either
The resulting theoretical air speed values were regressed against the measured air speed
values for every high-speed and for every low-speed segment for every run. Unlike the proposal,
this analysis did not average these values for 5-mph increments. The resulting linear relationship
was used to correct the air speed measurements in the real-time data.
3.2.1.1.2.3 Yaw Angle Measurements
The agencies also received comments on the inclusion of yaw angle in the coastdown
procedure. The proposal assumed that the coastdown occurs at zero yaw, however this condition
can only occur in perfect headwind, perfect tailwind, or no wind; which are all extremely
Direction 1
Direction 2:

-------
unlikely. Though it is difficult to obtain the yaw curve (drag polar) from coastdown tests, we
can characterize a certain coastdown test at an average yaw angle. First, the yaw angle for every
run was calculated assigned to each CdA value. The air direction was measured onboard with an
anemometer that is accurate to ±2°, according the product specifications.10 Thus, the average
yaw angle for each run was calculated using trigonometric relations from the average parameters
from the high-speed segment. See Figure 3-3 above for variable references.
w ¦ sin(0 + 0veh)
i/>n = arctan
V + W ¦ COs(0 + 0veh)-
Equation 3-2
The effective yaw angle for the coastdown test, xpeff, was then calculated by averaging the
yaw angles from all the runs from that test. Because the opposite direction runs yield positive
and negative angles, the absolute value of the yaw angle from every run was used.
^eff = ^~—yVrun
'trnnc
Equation 3-3
3.2.1.1.2.4 Tire rolling Resistan ce Impacts
The agencies also commissioned a study on tire rolling resistance as a function of speed
on the tire models that were being used on the tractors and trailers tested in the coastdown
program. The agencies conducted a tire coastdown test using SAE J2452 at Smithers Rapra to
measure tire rolling resistance force at various speeds. The load and inflation test points were
modified slightly to accommodate operating conditions of tires on an empty tractor-trailer
configuration, as listed in Table 3-2.
Table 3-2 Test Points for Tire Rolling Resistance Stepwise Coastdowns
SAE J2452 (light truck)
EPA test
Load (% of max)
Inflation pressure
(% of max)
Load (% of max)
Inflation pressure
(% of max)
20
110
20
100
40
50
55
70
40
100
85
120
70
60
85
100
100
100
100
95
The result of each test was a regression equation relating the rolling resistance force to
load, inflation pressure, and speed, as described in Eq. 3 of SAE J2452.
Pa ¦ //(a + b ¦ v + c ¦ v2)
Equation 3-4

-------
This equation was used for each tire to develop the tire rolling resistance characteristics
with speed for each vehicle. Since the same tire model was installed on a given axle, the
calculation was done one axle at a time, assuming uniform load distribution over all the tires on a
given axle.
^TRRaxleO) = naxle ¦ P^fe ¦	¦ (aaxle + baxle ¦ V + Caxle ¦ V2)
^axle/
Equation 3-5
The tire rolling speed characteristic for the full vehicle is the sum of the three axles.
^TRR,veh (V) = ^TRR, drive (v) + ^TRR, steer (v) + ^TRR, trailer (v)
Equation 3-6
The change in tire rolling resistance between two speeds, AFtrr, can be calculated by
calculating the difference in /'trr.ycIi values at those two speeds.
3.2.1.1.2.5	Drive Axle Spin Loss Impacts
The proposed coastdown procedure included an assumption for drive axle spin loss as a
function of vehicle speed. It included fixed values at an average speeds of 20 mph and 65 mph.
However, the agencies obtained additional spin loss data, which indicated that spin losses can
vary significantly between axle models. This means that two identical tractors with differing
axle models could produce different drag area results even if tested in the same wind conditions.
The data we obtained showed a spin loss impact on the calculated drag area of up to 0.15 m2
As a result, the agencies used spin loss data specific to the axle model in the vehicle
being tested, where the data were available. Data were obtained as power loss as a function of
wheel speed and converted to force loss using estimates for wheel size. Consultation with one
axle manufacturer indicated that similarly sized axles from a given manufacturer could be
assumed to have similar spin losses as a function of speed. We did not have spin loss data for
4x2 configurations, so we estimated the spin loss for these vehicles to be half of a similar axle
model in the 6x4 configuration. The axle efficiency test discussed in Chapter 3.8 is the source for
such data, the zero-torque subset of which is applicable to the coastdown analysis.
3.2.1.1.2.6	Drag Area Calculation
The agencies proposed an iterative analysis method to determine drag area from a
coastdown test. While this analysis can be done for any pair of speed ranges, a low-speed range
of 25 to 15 mph and a high-speed range of 70 to 60 mph were proposed. Table 3-3 below
describes the analysis methodology in the NPRM step by step. This analysis involves analyzing
the coastdown over two separate speed ranges. A low-speed range is used to estimate
mechanical losses and subtract them out of a high-speed range to estimate aerodynamic drag.
This process is iterated until mechanical and aerodynamic drag forces converge. The force is not
calculated at each measurement, but instead the net force over each speed range is calculated by
measuring the time taken to decelerate through each speed range. We assumed a linear decrease
in speed (i.e. constant deceleration), because the speed ranges are small.

-------
Table 3-3 Drag Area Calculation Steps for High-Low Iteration Analysis in the Phase 2 Proposal
STEP
VARIABLES AND EQUATIONS
VARIABLE DEFINITIONS
Step 0: Find the times
bracketing the low-speed and
high-speed ranges
vioi ^hil> ^hi2
v = vehicle speed
\o\=15mph, lo2 =25mph
\a\=60mph, \A2=70mph
Step 1: Calculate acceleration
for each speed range.
V\02 ~ Viol &v\o
^lo2 — tlol At),-,
Vhi2 ~ Vhil A^lo
^hi2 — ^hil At),-,
a = vehicle acceleration
t = time
Step 2: Calculate average road
grade force over each speed
range.
IS (Ah\ h\02~Kl
^grade.lo — Mg ( J — Mg
\tiS/\0 ^lo2 Slol
fAh\ ** ^hi2 - hhil
^grade.hi — Mg ( J — Mg
\ZiS/hi Shi2 Shil
M = vehicle mass
h = elevation (relative)
s = travel distance
g = gravitational
acceleration = 9.81 m/s2
Step 3: Inertial and Effective
Mass
(Add 125 lbm per tire to
account for rotational inertia).
lbm
^inertial — 125 ^tires
tire
cch kg
= b6. /	nf irpq
tire
Me — M + Minertiai
Mmertmi = additional inertia
from rotating components
Me = effective mass
/?tires = total number of
tires in test configuration
Step 4: Road load force for
each speed range, also
accounting for rear axle loss
estimate ic).
Flo ~ ~MeCl\o + ^grade.lo — ^axle.lo
^hi — — ^e^hi ^grade.hi — ^axle.hi
II II
(U
Step 5: Air density during
each high speed section.
1000 x P
P ~ R(T + 273.15)
p = density of air
P= average ambient
pressure during high speed
run in kPa
T = average ambient
temperature during high
speed run in °C
R = gas constant for air
=287.058 J/(kg-K)
Step 6: Average relative air
speed over each speed range.
— v^lo2 vr — v^hi2 vr
vr,\0 - Zvl01— vm - %vhil —
vr = relative air speed
Step 7: Initial conditions (i=0).
Start with no aerodynamic
forces in the low speed range.
^aero,lo,0 — 0


-------
Step 8: Subtract low-speed
aerodynamic forces from low
speed forces to estimate
mechanical forces.

^mech.i — ^lo
^aero,lo,i

Step 9: Subtract mechanical
forces from high speed forces
to estimate aerodynamic
forces.

^aero,hi,i — ^hi
~ ^mech.i

Step 10: Adjust aerodynamic
forces by speed to estimate
low-speed aerodynamic
forces.
F ~F (^°\
raero,lo,i+l — raero,hi,i+l 1-2 /
\vr,hi/


Repeat steps 8-10 until:


Step 11: Repeat steps 8-10
until both high-speed
aerodynamic and low-speed
mechanical forces both
converge less than 1%.

^ ^aero,hi,i+l
< 0.01

and
^aero,hi,i



^ ^mech,lo,i+l
< 0.01



^mech,lo,i

Step 12: Calculate drag area.
r . 2Faerahi,i+i
L&A - _2
Pv r,hi

There are some advantages to using this method over the Phase 1 method. Focusing on
segmented speed ranges may open up more test locations, as less road or track space would be
required to collect a full data set. The middle range of speeds that would be eliminated contains
a higher proportion of rolling resistance forces and also sweeps through greater yaw angles, even
at modest crosswind conditions, which can increase the aerodynamic drag of certain runs and
subsequently increase the variability of a test.
Essentially, the proposed iteration method is attempting to solve two force equations, one
at the low speed and one at the high speed, were the drag area and mechanical forces (except
spin loss) are the same in the high speed and low speed.
1 2
^hi — ^mech 2 P^d-^^air.hi
1 2
F\o ~ ^mech ^P^d-^^air.lo
Equation 3-7

-------
This system of equations that the iteration method represents can be simplified into an
analytic equation that produces the same result and avoids the iteration process altogether.
CdA - j	
2 P ^r2hi - ^r,lo)
Equation 3-8
The iteration-based and analytical solutions were compared and shown to be identical
(within very small rounding errors), as shown in Figure 3-4.
<
"O
O
"re
o
TO
a
<
4
5
6
7
8
Iteration-based CdA
	1:1 line
Figure 3-4 Analytical Solutions from All Coastdown Runs are Identical to the Iteration Method.
Similarly, the inclusion of tire rolling resistance and drive axle spin loss as a function of
speed could also be incorporated into the analytical equation.
r . _ ^hi — fio — AFSpin — AFjrr
LdA -	j	
2"P"(^r2hi-^r,lo)
Equation 3-9
In this new equation, Fhi and F\0 include the drive axle spin loss (i.e. they are not
subtracted out), unlike the proposal. The \/''spin and AFtrr values are determined from the
average vehicle speeds in the low-speed and high-speed ranges, using the tire rolling resistance
and axle spin loss test procedures in the manner described above and in 40 CFR 1037.528.
As mentioned, the agencies proposed a low-speed range of 25-15 mph. With the
inclusion of yaw angle in the final rule, the agencies reviewed the appropriateness of this speed
range with respect to yaw characterization and the coastdown procedure generally. The agencies

-------
partnered with National Research Council Canada (NRC) to investigate coastdown and constant
speed testing. The ProStar sleeper cab tractor borrowed by the agencies from Environment
Canada and tested by SwRI, was tested by NRC at Transport Canada's Test and Research Centre
in Blainville, Quebec. One of NRC's conclusions from their study was to reduce the low-speed
range from 25-15 mph to 15-5 mph to reduce the contribution of aerodynamic forces to the road
load at low speeds, thus leading to drag area measurements with higher precision.11
The agencies analyzed the yaw characterization as a function of low-speed range.
Sleeper Cab 3 contained the greatest number of coastdown tests and was used for this purpose.
The drag area and yaw angle of every run that was conducted within the wind specifications was
calculated and plotted to evaluate the effect of yaw angle on the calculated drag area for the
various low-speed ranges.
Figure 3-5 shows that lowering the speed range shows a flatter yaw characterization.
Since we expect drag to increase with yaw angle, the lower speed ranges, particularly 15-5 mph,
better represent a realistic yaw curve. This aligns with the recommendation from NRC.
However, with average wind conditions allowed up to 10 mph, it would be possible to have a tail
wind "pushing" the vehicle at the low end of the 15-5 mph range. Testing at this low-speed
range without this tail wind effect would require that the tail wind not exceed 3 mph. The
agencies considered this constraint to be too restrictive to allow for enough available days to test.
As a result, the agencies are finalizing that the low-speed range be 20-10 mph with an added
constraint that the component of wind parallel to the direction of travel must not exceed 6 mph.
This value was chosen to be fully out of the low-speed coastdown speed range, which requires
coasting down to 8 mph for determining the coastdown ending time and ending speed points.
6.00
5.75
550
5.00
4.75
H	1	1	1	1	
0	12	3	4
|Yaw angle| [deg]
Low-speed range [mph] 	15-5 	 20-10 	 25-15
Figure 3-5 Drag Area as a Function of Yaw Angle Calculated for Different Low-Speed Ranges.

-------
Further analysis using the method described above showed an unexpected difference in
the CdA results with respect to run direction. For example, Figure 3-6 shows that the average
CdA for the westbound runs is consistently higher than the eastbound runs despite different yaw
conditions.
5.4
53
pJ 5.2
<
S 5.1
5.0
4.9
1§Mo^201<1


-4	-2	0	2	4
Yaw angle [deg]
Run direction • West # East
Figure 3-6 Average CdA by Direction from Sleeper Cab 3 Shows Direction Bias for 5 Different Tests.
The vehicle experiences different air speeds and yaw angles depending on the direction
during testing, even if wind conditions remain stable. The equations described so far have used a
"matched pair" approach, where a high-speed segment is matched with its corresponding low-
speed segment, both of which are in the same direction. This approach assumes that aerodynamic
forces are constant in the low-speed range. In reality, these forces will vary given the varying
magnitude and orientation of the air speed between the two travel directions in the low-speed
range. Conditions associated with the test site may also cause some differences between the
directions that may not be related to aerodynamics. To account for these effects, the low-speed
air speed and force values were averaged by opposite direction pairs before applying them to the
analytical solution for each high-speed segment. The resulting equation is below. For tests
conducted with two consecutive high-speed segments and two consecutive low-speed segments
in the same direction, the averaging was done for every four low-speed segments.
r . _ ^hi — ^lo.pair — ^^spin — ^^TRR
LaA -	j	
2 " P " 0>£hi - ^lo.pair)
Equation 3-10
The results from this calculation, shown in Figure 3-7, shows mitigation of the bias, with
a more even distribution of results by direction.

-------
5.4
5.3
FT 52
<
E
<
5.0
49
Figure 3-7 Average CaA, Using Low-Speed Paired Means, by Direction from Sleeper Cab 3 for 5 Different
Tests
While the low-speed paired results for individual runs and directions are different from
the matched pair results, the overall mean C&A result is not significantly affected, as shown in
Figure 3-8 for the five tests on Sleeper Cab 3. Though the results are similar, the benefit in this
method is the reduced scatter in the results from individual runs, which helps to prevent the
presence of outliers and include more data when determining results for the reference tractors.
This process is discussed in greater detail in Chapter 3.2.2.2.1. The agencies are finalizing the
low-speed paired method for calculating C&A.











30Jun2015 19Nov201-1
12Aug2015

20Jun2015
® 30Jun2D15^


^11Jun201

11Jur2015
_19Nov2014
• •


20Jun2015
25Jun2015
_25Jun5
A12AJS
m







i	1	r
-4	-2	0	2	4
Yaw angle [deg]
# West # East
520

-------
The uncertainty of the coastdown result was characterized through the standard error. In
the test program, most tests were conducted with 14 to 16 runs. Several other tests were
conducted up to 32 runs. As shown in Figure 3-9, on average, the standard error of the tests
decreased as the number of valid runs increased, with the standard error trending below 1 percent
beyond 20 runs.
Number of runs
Figure 3-9 Standard Error of Coastdown Test Decreases with Increasing Number of Runs.
After conducting the analysis above, the agencies are finalizing the high-low analysis
method using 70-60 mph and 20-10 mph along with requirements to quantify the speed
dependence of tire rolling resistance and drive axle spin loss to determine the drag area from the
coastdown test. Determining the effective yaw angle of the coastdown test is also being
finalized. Additional requirements on the statistical validity of data points, which were not
applied for the data set discussed here, are being finalized for reference tractors tested to
determine Fait-aero. They are discussed later in Chapter 3.2.2.2.
3.2.1.1.3 Wind-averaged Drag Adjustment
We received comments in Phase 1 regarding the use of the wind-averaged drag since it
accounts for aerodynamic performance across a broader spectrum of wind conditions rather than
a pure headwind or tailwind. Consequently, the use of wind-averaged drag for aerodynamic
assessment may better reflect real-world aerodynamic performance and fuel consumption. We
assessed the use of wind-averaged drag for Phase 2 and the results are discussed below in this
section.
EPA and NHTSA recognize that wind conditions have a greater impact on real world
CO2 emissions and fuel consumption of heavy-duty tractor-trailers than light-duty vehicles. As
stated in the NAS report12, the wind averaged drag coefficient is about 15 percent higher than the

-------
zero degree coefficient of drag (Cd). The large ratio of the side area of a combination tractor and
trailer to the frontal area suggests that winds will have a significant impact on drag. One
disadvantage of the agencies' approach to aerodynamic assessment in Phase 1 is that the test
methods have varying but limited degrees of ability to assess wind conditions. Wind tunnels and
CFD are existing tools to determine wind-averaged drag. The coastdown test has limited ability
to assess yaw conditions. The constant speed test has the potential to determine wind-average
drag, but an industry standard for this does not exist. It is very possible that different tools
produce different drag results for the same vehicle.
In Phase 1, the agencies finalized the use of coefficient of drag values that represented
zero yaw (i.e., representing wind from directly in front of the vehicle, not from the side). The
agencies recognized that the results of using the zero-yaw approach will produce fuel
consumption results in the regulatory program which are slightly lower (i.e. predict better fuel
consumption results) than in-use, but we believed this approach was appropriate since not all
manufacturers were using wind tunnels for the aerodynamic assessment to the extent needed for
wind-averaged drag quantification purposes.
While the coastdown test yields a drag area and an effective yaw angle, the aerodynamic
input into GEM for GHG compliance is a wind-averaged drag area. This was chosen for its
representation of more real-world wind conditions. Therefore, the agencies needed to adjust the
coastdown drag area to a wind-averaged drag area using data other than a coastdown test. For
Phase 2, the agencies are continuing to require the use of an alternate method adjustment factor,
or Fait-aero, to relate alternate aerodynamic methods to coastdown results. However, for Phase 2,
Fait-aero will be based on the effective yaw angle of the coastdown instead of zero degrees, which
was the basis for Phase 1.
The agencies are finalizing a wind-average drag input based on a 65-mph vehicle speed,
instead of 55 mph, which was originally proposed. We had received comment that 65 mph was
more representative of tractor driving behavior. Also, the GEM result for tractors more heavily
weights the 65-mph cycle over the 55-mph and ARB Transient cycles. Requiring a drag input
based on 65 mph makes this more consistent with the overall GHG evaluation of the tractor.
We also received comments that a surrogate angle can be used to accurately determine
wind-averaged drag, as opposed to the full yaw sweep using SAE J1252 that the agencies had
proposed. A surrogate angle of 4.5° was suggested by industry commenters. The agencies
compared results between the full yaw sweep and the suggested surrogate angle and found that
4.5° could be an accurate representation of wind-average drag at 65 mph vehicle speed and 7
mph wind speed. This analysis is described in further detail in the scale wind tunnel and CFD
sections below.
The analysis in this section shows how alternate aerodynamic test methods were used to
develop wind-averaged drag area baselines and the acceptability of 4.5° as a surrogate yaw angle
for determining wind-averaged drag at 7 mph wind speed and 65 mph vehicle speed. A fourth-
order polynomial curve was used to estimate CdA at ±ipdr and ±4.5° with the alternate methods.
(CdA) alt = a0 + a1xl) + a2il>2 + a3xl)3 + a4xf)*
Equation 3-11

-------
The following equation was used to adjust coastdown results from the coastdown
effective yaw angle to the wind-averaged surrogate angle of 4.5° using the yaw data generated
from the alternate aerodynamic methods. Average results from positive and negative angles
were used at both the coastdown effective yaw angle and 4.5° where data were available.
rr	(CdA)
alt,±4.5° _	[(Cdi4)ai1:,4.5°	(Ql^)alt,-4.5°]
l^divwa — coast " rr A>. — (>divcoast " r, „ T
(Cd/lJalt,+i/)eff	[(^di4)alt,i/)eff	(^di4)alt,-i/)effj
Equation 3-12
For most tests, results from positive and negative angles were averaged to calculate this
value. This equation was used with three alternate methods - wind tunnel, CFD, and constant
speed testing - to develop a broad set of wind-averaged drag area values for a given tractor.
These values then informed the aerodynamic bin structure for Phase 2.
3.2.1.1.3.1 Scale Wind Tunnel
Two scale wind tunnels were used in the aerodynamics baseline determination. The
agencies conducted 1/8-scale wind tunnel tests at Auto Research Center (ARC) in Indianapolis.
The ARC wind tunnel is a closed single return tunnel with 3/4 open-jet working section and
moving ground plane (2.3 m wide x 2.1 m high x 5.5 m long). It is powered by an air-cooled
373kW variable speed DC motor that drives a 9-bladed fan with carbon fiber blades. Its speed
may be varied and set at any value from 0 to 610 rpm. The maximum wind speed is about 50
m/s. The wind tunnel can accommodate a model up to 50 percent scale (1/2 scale) for race car
applications down to 12.5 percent scale (l/8th scale) for Class 8 tractor and trailer combinations.
The wind tunnel is equipped with a moving ground plane {i.e., rolling road), four-stage boundary
layer suction system, and a top-mounting "Sting" system that allows for yawing of the model.
For model development, ARC has in-house model developers and can create highly detailed
scale models using original computer aided design and engineering (CAD/CAE) drawings or
using in-house scanning equipment to perform scanning and digitizing to create CAD/CAE
drawings (see Figure 3-10 below).

-------
Figure 3-10 l/8th Scale Tractor-Trailer Model in ARC Reduced Scale Wind Tunnel.
The testing was conducted with a tunnel speed of 50 m/s, equivalent to a Reynolds
number (Re) of 1.1 million, with Class 8 sleeper and day cab tractors equipped with
aerodynamics components sold on the full size version of the tractors. For our test program, we
assumed a base tractor-trailer gap of 45 inches and a bogey position of 40 feet (California
position) from the leading edge of the trailer.
To support Phase 2, we tested model year 2011 or later sleeper cab and day cab tractors.
The tractor models used in the reduced scale wind tunnel (RSWT) test matched the tractor
models used for the on-road testing to the extent feasible. Not every wind tunnel tractor was a
close match to the tractors tested on-road at SwRI. The wind tunnel tractors that were close
matches, based on model year, make/model, and general aerodynamic features, were used to
determine the yaw curve adjustment of the coastdown to 4.5°. The RunID numbers from the
ARC study that were used in this analysis are listed in Table 3-4.5
Table 3-4 ARC Wind Tunnel Runs Representing Tractor Configurations Tested at SwRI
Tractor
Run ID
Sleeper Cab 1
2013091224
Sleeper Cab 3
2015082651
Sleeper Cab 4
2014102906
Day Cab 30
2015082531 J
Day Cab 31
2015082413
The ARC tunnel data also confirmed that the use of 4.5° was an appropriate
approximation of wind-averaged drag. The yaw sweep data from each test was fitted to a fourth-
order polynomial. Wind-averaged drag area was then calculated using SAE J1252, and the
surrogate angle drag area was calculated from the average of the 4.5° and -4.5° predictions from
the polynomial fit. For the 373 tests analyzed, the error from the surrogate-angle drag area to the
J1252 drag area ranged from -1.0 percent to 3.0 percent, with a mean of 0.2 percent and a median

-------
of 0.3 percent. Only ten tests from a single tractor-trailer configuration had an error greater than
1.0 percent.
National Research Council Canada (NRC) also performed a scale wind tunnel study to
support Transport Canada's ecoTECHNOLOGY for Vehicles program. The testing was done at
their 9-meter wind tunnel at 30 percent scale to measure the aerodynamic performance of various
drag reduction technologies. While the tractor model used in this study was not identical to a
particular OEM tractor model available on the market, NRC did advise us that it was originally
based on a design similar to Sleeper Cab 3 from the coastdown test program. The data from this
study was used to inform the adjustment of the coastdown result for this tractor model.
Figure 3-11 l/8th Scale Tractor-Trailer Model in Canada Reduced Scale Wind Tunnel.
A few adjustments were made to the NRC data. NRC ran tests at tractor-trailer gaps of
36, 42, and 48 inches without any trailer aerodynamic devices (baseline). They also ran a test at
a tractor-trailer gap of 36 inches with "standard side-skirts," the skirt type in the study that most
closely matched the skirt tested by SwRI. SwRI conducted coastdown testing with Sleeper Cab
3 with a 47-inch gap. Interpolations were made to the NRC tunnel to estimate the yaw curve for
a configuration with 47-inch gap and standard side-skirts. NRC collected a full-sweep of yaw
angles (-12° to 12°, inclusive) for a tractor-trailer gap of 36 inches, but only collected data for
seven angles between -12° to 1°, inclusive, for the 42-inch and 48-inch configurations. As a
result, final curve was only calculated over the -12° to 1° range of yaw angles.
First, the drag area results for each of these seven angles from the 42-inch and 48-inch
baseline tests were linearly interpolated to estimate results for a 47-inch gap. Then, the ratios of
these 47-inch-gap results to the 36-inch-gap baseline results was applied to the 36-inch-gap
standard side-skirts results to estimate the yaw curve for the configuration tested by SwRI.
The results from each coastdown test were adjusted to a wind-averaged value from the
coastdown effective yaw angle, with the wind tunnel results, using Equation 3-12. A
fourth-order polynomial fit, described by Equation 3-11, was used to estimate C&A at ±ipes and
±4.5°.The numbers determined from both wind tunnels for the baseline calculations are
presented in Table 3-5 below. The number is an average where multiple coastdowns were
conducted.

-------
Table 3-5 Wind Tunnel Results and Baseline Calculations from ARC and NRC Studies; CaA in m2
Site
Tractor
( CdA ) coast
[°]
(C,A) lt +
^ d ssit,±i//eS
(Ccp4)alt,±4.5°
F alt-aero
(Cd^4)wa
ARC
Sleeper 1
5.32
0.60
5.23
5.81
1.02
5.91
Sleeper 3
5.15
2.44
5.15
5.44
1.00
5.44
Sleeper 4
5.63
1.88
5.15
5.64
1.09
6.16
Day Cab 30
5.80
0.81
5.46
5.94
1.06
6.32
Day Cab 31
5.37
1.65
5.61
6.05
0.96
5.79
NRC*
Sleeper 3
5.15
2.44
5.38
5.65
0.96
5.41
*Only negative angles were evaluated from the NRC tunnel, due to available data.
3.2.1.1.3.2 Computational Fluid Dynamics (CFD)
Computational Fluid Dynamics (CFD) capitalizes on today's computing power by
modeling a full size vehicle and simulating the flows around this model to examine the fluid
dynamic properties, in a virtual environment. CFD tools are used to solve either the Navier-
Stokes equations that relate the physical law of conservation of momentum to the flow
relationship around a body in motion or a static body with fluid in motion around it, or the
Boltzmann equation that examines fluid mechanics and determines the characteristics of discrete,
individual particles within a fluid and relates this behavior to the overall dynamics and behavior
of the fluid. CFD analysis involves several steps: defining the basic model structure or
geometry based on provided specifications; applying a closed surface around the structure to
define the external model shape (wrapping or surface meshing); dividing the model and the
surrounding environment control volume into smaller, discreet shapes (gridding); defining the
flow conditions in and out of the control volume and the flow relationships within the grid
(including eddies and turbulence); and solving the flow equations based on the prescribed flow
conditions and relationships.
The agencies commissioned a CFD a study through contractor ICF to study a number of
issues related to the Phase 2 rulemaking. Two CFD providers, ARC and Exa, were chosen to
perform a CFD evaluation of one of the tractors used in SwRI's on-road testing. ARC used
Elements, a Reynolds Averaged Navier-Stokes (RANS)-based model. Exa used PowerFLOW, a
Lattice Boltzmann-based model. Three trailer configurations were simulated with this tractor:
Phase 1 (no-control) trailer, trailer with skirts, and trailer with skirts and a tail. Attempts were
made through photographs and measurements to create a vehicle geometry as close as possible to
the on-road vehicle. Multiple Reynolds numbers and turbulence intensities were evaluated.
Details of the simulations can be found in the CFD report prepared by ICF. 7
Full yaw sweeps were run for the 5.1 million Reynolds number (65-mph), zero
turbulence simulations. These simulations showed that 4.5° is a viable surrogate for wind
averaged drag at 7/65 mph, with variations of under 1.6 percent between the surrogate angle
average and wind-averaged drag calculated per SAE J1252.

-------
Table 3-6 Surrogate Angle C&A Comparison with SAE J1252 Wind-Averaged Drag Calculation; C&A in m2
CFD source
Configuration
Wind-averaged CdA
(SAE J1252, 7/65 mph)
CdA average at ±4.5°
% error
Exa
Phase 1
5.58
5.60
0.4%
Skirts
5.01
5.04
0.6%
Skirts + Tail
4.38
4.36
-0.5%
ARC-
ELEMENTS
Phase 1
6.25
6.35
1.6%
Skirts
5.57
5.63
1.1%
Skirts + Tail
5.07
5.11
0.8%
The CFD results also show that a multiplicative adjustment is likely more appropriate to
adjust based on yaw angle. Multiple turbulence intensities were evaluated to understand the
effect of real-world air flow that exists during coastdowns compared to a controlled zero-
turbulence result. Two non-zero turbulence intensities, 3 percent and 6 percent, were evaluated
in the skirt configuration in both CFD environments. Turbulence intensity over the road can be
higher, but this is often due to traffic, which is minimized during coastdown testing. The ARC
results showed less than 1 percent effect from increased turbulence intensity. The Exa results,
however, showed the wind-averaged drag area increased with turbulence intensity by 4.4 percent
and 6.5 percent, respectively. Importantly, the increase in drag from 0° to 4.5° within each
turbulence intensity simulation is more consistent as a ratio than as a difference. This is also true
across the two CFD codes. The simulation results are shown in Table 3-7.
Table 3-7 Comparison of Scalar Difference (Increase) to Ratio across CFD Codes and Turbulence
Conditions
CFD Source
TI
CdA at 0°
CdA at 4.5°
CdA increase
CdA ratio
r%i
[m2]
[m2]
(0° to 4.5°) [m2]
(4.5° to 0°)
Exa
0
4.501
5.017
0.516
1.115

3
4.677
5.237
0.560
1.120

7
4.777
5.342
0.565
1.118
ARC-ELEMENTS
0
5.031
5.636
0.605
1.120

3
4.989
5.615
0.626
1.125

7
4.999
5.615
0.616
1.123
The scalar increase of drag area from 0° to 4.5° varies from 0.516 to 0.626 m2 The
multiplicative increase (ratio) varies from 1.115 to 1.125. For a hypothetical coastdown result of
5.000 m2, this results in a range of coastdown yaw-adjusted drag values of 5.516 to 5.626 m2
using the scalar approach and 5.575 to 5.625 m2 using the multiplicative approach. This shows
that the multiplicative approach has less variability when applied to the coastdown tests and is
the reason why the multiplicative approach is being used in this analysis and the test procedure
the agencies are finalizing.
In addition to the CFD study commissioned by the agencies, certain manufacturers
provided CFD data for models represented by Sleeper Cab 4, Sleeper Cab 5, and Day Cab 20.
The results from each coastdown test were adjusted to a wind-averaged value from the
coastdown effective yaw angle, with the CFD results, using Equation 3-12. A fourth-order
polynomial fit, described by Equation 3-11, was used to estimate CdA at ±i/>eff and ±4.5°. The

-------
numbers determined from both wind tunnels for the baseline calculations are presented in Table
3-8 below. The number is an average where multiple coastdowns were conducted.
Table 3-8 CFD Results and Baseline Calculations; C&A in m2
CFD Source
Tractor
(C d^4)coast
[°]
(C,A) lt +
V d -/alt,±y/eff
(Cd^4)alt,±4.5°
Falt-aero
(Cd^4)wa
ARC-
ELEMENTS
Sleeper 1
5.32
0.6
5.00
5.62
1.06
5.98
Exa
Sleeper 1
5.32
0.6
4.51
5.06
1.18
5.96
Sleeper 4
5.63
1.88
**
**
1.20
6.12
Sleeper 5
5.16
2.06
**
**
1.16
5.44
Day Cab 20*
5.38
2.31
**
**
1.13
5.81
Note:
*Only positive angles were evaluated with CFD for Day Cab 20, due to available data.
**CFD results provided confidentially by manufacturers. Only final (CdA)wa result shown.
3.2.1.1.3.3 Constant Speed Testing
Similar to the coastdown testing, constant speed testing is conducted on road and
measures road load forces acting on the tractor. In contrast to the coastdown test where the
vehicle is accelerated to a set speed and then allowed to coast to a lower speed in neutral, the
constant speed test is conducted by measuring torque along the driveline while the vehicle is
driven at various constant speeds. This helps to reduce measurement uncertainty due to potential
driveline vibration experienced during coastdown and better isolate the force contributions
between speed transitions over the speed range (e.g., aerodynamic drag dominance at high speed;
a mix of aero drag and mechanical/frictional forces in middle speeds; and mechanical/frictional
force dominance at low speeds). In addition, whereas the total force, and consequently the total
drag force, is derived based on the speed and time for the coastdown test, the constant speed test
measures the total force at the wheels using wheel hub torque meters and/or a driveshaft torque
meter. The constant speed test has the potential to reduce uncertainty compared to a coastdown
because it can collect data at a single speed for a sustained amount of time. For Phase 1, we
allowed the use of Constant Speed testing as an alternate aerodynamic method but did not
promulgate specific test procedure requirements. In lieu of this, a manufacturer would have been
required to develop its own test procedure for constant speed testing and submit it to the agencies
for approval.
Currently, there is no industry standard for conducting constant speed tests with heavy-
duty vehicles and no manufacturers have submitted alternative compliance test plans for
approval from EPA. The European Union did include constant speed testing in the aerodynamic
component of their greenhouse gas emissions monitoring and certification program, but it did not
include a calculation of wind-averaged drag.13 For Phase 2, we proposed specific requirements
for the constant speed test procedure to be used by manufacturers to certify their tractors.
Accordingly, we evaluated the constant speed testing using the same vehicles tested with the
coastdown test procedure. For our evaluations, we used several speeds to determine the optimal
speeds for constant speed testing. In addition, we performed the testing with both wheel hub
torque meters and a driveshaft torque meter to quantify the benefits and detriments of both
methods.

-------
The agencies conducted constant speed testing through SwRI along the same stretch of
roadway as the coastdown testing. Torque was measured at the driveshaft for all the tests and
also at each of the four wheel hubs for most of the tests. More details of the test setup and
procedure can be found in SwRI's coastdown and constant speed testing report.8 Each vehicle
configuration of interest was tested at least twice, once in winds within the SAE J1263
specifications, and once outside of the specifications.
Testing was performed at the following speeds and durations while recording torque and
engine data.
o	10 mph - 7.5 minutes in each direction
o	20 mph - 7.5 minutes in each direction
o	30 mph - 7.5-10 minutes in each direction
o	50 mph - 8-13 minutes in each direction
o	70 mph - 8-10 minutes in each direction.
If necessary, multiple passes were conducted to meet the time requirements. The 20-mph run
was eliminated partially through the test program in favor of the higher speeds. Cruise control
was used to maintain speeds, except for the lower one or two speeds for certain tests, where the
driver controlled the speed through pedal position and close monitoring of instantaneous vehicle
speed. The combination of multiple wind conditions and multiple high speeds created enough of
a yaw angle distribution to construct a yaw curve for a given configuration. This yaw curve
construction would then help adjust the coastdown result to 4.5°.
For analysis of the constant speed test procedure data, the 10-Hz data were split into 10-
second segments over which the torques, air speed, and air direction were averaged. For tractors
equipped with the driveshaft torque meter, the road load force was calculated for each 10-second
segment as follows:
^	_ Tshaft " ^eng , ^
^RL,shaft — Qg .	+ ''grade
Equation 3-13
For tractors also equipped with the wheel torque meters, the road load force was
calculated as follows:
_ ^wheel ' ^wheel ,
''RL,wheel —	~	¦" ''grade
Equation 3-14
Where:
tshaft = driveshaft torque
cosng = engine speed
GR = transmission gear ratio
-Frt,shaft = road load force calculated from the driveshaft torque
7wheel = wheel torque, sum of all four wheel torque measurements

-------
CO wheel = wheel speed, average of all four wheel speed measurements
Frl,wheel = road load force calculated from the wheel torque
F&ads = grade force
Since we received comments that the speed dependence of tire rolling resistance should
be incorporated into coastdowns, we applied a similar process for the constant speed data
analysis. Under the same tire test program that was done for coastdown analysis, the agencies
tested the same tires at Smithers Rapra using the SAE J1269 constant speed tire rolling resistance
test. SAE J1269 requires testing only at 80 km/h (50 mph), so the agencies also tested 16 km/h
(10 mph) and 113 km/h (70 mph) to align with three of the speeds tested in the constant speed
test program.19 The change in rolling resistance with speed, /l/'TRR.veh, was calculated from these
tests using the regression-based method described for the stepwise coastdown tests in Chapter
3.2.1.1.2.4 (Equation 3-4 through Equation 3-6).
Drag area C&A was calculated using a subtraction of the low-speed force from the high-
speed force. However, the low speed force that was used was the average 10-mph force and air
speed from the calmer wind day. This was to avoid low-speed points where unusually high
aerodynamic loads would be present. The high-speed values were individual 10-sec segment
wheel force and air speed averages from the 50-mph and 70-mph runs.
r . _ ^RL,hi — ^RL,lo — AFXRRveh
C"A-	1	;	
2 ' P ' Vr,hi,avg
Equation 3-15
For each high-speed point, the yaw angle was also calculated from the measured vehicle
speed and from the wind direction and wind speed measured by the roadside weather station,
using Equation 3-2. A fourth-order polynomial fit of C&A and yaw angle, described by Equation
3-11, was used to estimate the mean C&A values from constant speed at ±ipctr and ±4.5°. Figure
3-12 below shows the resulting yaw curve for one of the tractors.

-------
9
8
c\T
<
£ 7
<
T3
o
6
5
-20	-10	0	10
Yaw angle [deg]
~ 95% Confidence Limits 	 Regression
Figure 3-12 Yaw Curve Results from Sleeper Cab 4.
While the graph may appear to show a high level of scatter, the large number of data
points shows a relatively low level of uncertainty. Uncertainties were determined using the
statistics produced by the regression. The standard error was used because the objective of the
regression was to identify the yaw characteristic for the vehicle and not to predict an individual
test point. Standard errors ranged from 0.5 to 0.8 percent for C&A values at 4.5° and -4.5° for
the five configurations analyzed. It is possible that there are bias errors associated with constant
speed testing, but determining the relative yaw characteristic within a given test was the
objective of this analysis.
The results from each coastdown test were adjusted to a wind-averaged value from the
coastdown effective yaw angle, with the constant speed test results, using Equation 3-12.
The numbers determined from both wind tunnels for the baseline calculations are presented in
Table 3-9 below. The number is an average where multiple coastdowns were conducted.
• •
••

-------
Table 3-9 Constant Speed Results and Baseline Calculations; C&A in m2
Tractor
(C cu4)coast

-------
5.5
5.4
CM

-------
6.2
CM
<
s
<
"O
O
6.0
5.8
5 6
5.4
• Coastdown result
~ Alt. method adjustment to 4 t
X Mean at 4.5°
j"

x6317









•



5814
JT-















o
•



0	1	2	3	4	5
Yaw angle [deg]
Tractor B 20 ~ 30 ~ 31
Figure 3-16 Coastdown Results with Alternate Method Adjustments to 4.5° Yaw - Day Cabs
The mean wind-average drag results for all the tractors for each cab type were combined
to develop the C&A bin boundaries for Phase 2. To keep the bin levels consistent between Phase
1 drag results and Phase 2 wind-averaged drag area results, the agencies developed the bin
boundaries shown in Figure 3-17 for high-roof sleeper cab tractors and Figure 3-18 for high-roof
day cab tractors. As these tractors were in Bin III and Bin IV for Phase 1, their Phase 2 results
led to the numerical values for the Phase 2 bin boundaries for Bin III and Bin IV.

-------
CM
<
E,
<
"O
O
"O
0)
D)
CO
1—
a)
>
ro
i
T3
C
7.0
6.9-
6 8
6.7
6.6-
6.5 -
6.4
6.3
6.2
6.1	-
60-
5.9
5.8
5.7-
5.6-
5.5
5.4
5 3
5.2
5.1
5.0


Bin II

6.039

5.902
•

Bin III
5.490
9 ^5.398
5.446
o
Bin IV
Tractor £ 1 O 2 ® 3 4 5
Figure 3-17 High Roof Sleeper Cab Phase 2 Results and Bin Boundaries
In general, the tractors' Phase 2 wind-averaged results with respect to one another are
similar to their relative Phase 1 results (Figure 3-2). As a result, the agencies drew Phase 2 Bin
IV such that Sleeper Cabs 3 and 5 were near the center of that bin. Sleeper Cab 1 moved further
into Bin III, whereas it was near the Bin III/IV boundary for Phase 1. Sleeper Cab 2 moved just
within Bin IV, whereas it was also near the Bin III/IV boundary for Phase 1. It is not unusual to
see modest shifts like this because the addition of trailer skirts may have a varying influence for
different tractors designs, but the tractors' overall order of results relative to one another were
similar between Phase 1 and Phase 2.

-------
CM
<
E,
<
"D
O
"O
0)
D)
CD
a)
>
ro
i
T3
C
5.317
Bin III
5.S14
5.787
B n IV
Tractor 0 20 £ 30 Q 31
Figure 3-18 High Roof Day Cab Phase 2 Results and Bin Boundaries
The two day cabs that were tested using the Phase 1 procedure landed close to the Phase
1 Bin III/IV boundary (Figure 3-2). For Phase 2, these two tractors (30 and 31) diverged in their
results and the Phase 2 Bin III/IV boundary was drawn in between them (6.0 m2). A third day
cab, only tested using the Phase 2 procedures, is included here for reference.
For bin boundaries beyond the Bin III/IV boundary, the bin widths were drawn similar to
Phase 1 or slightly narrower, approximately 0.4 to 0.5 m2 wide, for both the sleeper cabs and day
cabs.
The analysis described in this section led to the creation of aerodynamic bins for high-
roof sleeper cab and high-roof day cab tractors described in Table 3-10. This table can also be
found in Section III.E(2)(a)(viii) of the Preamble along with the bin definitions for low and mid
roof tractors, which were not tested in this program.

-------
Table 3-10 Phase 2 Aerodynamic Input Definitions to GEM for High Roof Tractors

CLASS 7
CLASS 8

Day Cab
Day Cab
Sleeper Cab

High Roof
High Roof
High Roof
Aerodynamic Test
Results (CdA wad in m2)
Bin I
>7.2
>7.2
>6.9
Bin II
6.6-7.1
6.6-7.1
6.3-6.8
Bin III
6.0-6.5
6.0-6.5
5.7-6.2
Bin IV
5.5-5.9
5.5-5.9
5.2-5.6
Bin V
5.0-5.4
5.0-5.4
4.7-5.1
Bin VI
4.5-4.9
4.5-4.9
4.2-4.6
Bin VII
<4.4
<4.4
<4.1
Aerodynamic Input to GEM (CdA „ad in m2)
Bin I
7.45
7.45
7.15
Bin II
6.85
6.85
6.55
Bin III
6.25
6.25
5.95
Bin IV
5.70
5.70
5.40
Bin V
5.20
5.20
4.90
Bin VI
4.70
4.70
4.40
Bin VII
4.20
4.20
3.90
3.2.2 Final Aerodynamic Test Procedures for Phase 2 Tractors
3.2.2.1 Standard Trailer
The most widely implemented trailer aerodynamic devices in the market today are trailer
side skirts that extend in the gap between the fifth wheel and the trailer bogey, and trailer
treatments that extend from the rear of the trailer (e.g., boat tails). As discussed in Section
III.E(2)(a)(iii) and Section IV.D(2) of the Preamble, we estimate that even without the Phase 2
rulemaking, approximately 50 percent of the new trailers sold in 2018 will have trailer side
skirts.14,15 As the agencies are finalizing GHG rules for tractors for model year 2021 and
beyond, we believe that it is appropriate to update the standard box trailer definition to reflect the
technologies we project will be used on the majority of the trailers in the fleet during that
timeframe. Therefore, the agencies are finalizing a new standard trailer for Phase 2 tractor
certification by requiring the use of trailer skirts with dimensions specified in 40 CFR
1037.501(g)(l)(v). As there may not be a commercially available skirt that meets these exact
dimensions, the agencies were able to verify similar aerodynamic performance of two different
skirts, one purchased and one fabricated by SwRI to the same dimensions. In order to help
simplify any fabrication processes, the skirt mounting requirement is flush with the side of the
trailer and does not contain curves.
With the addition of the skirt in our coastdown testing came a need for SwRI to move the
trailer bogey rearward one notch, approximately 4 inches, as the edge of the skirt came very
close to the leading outside trailer tires. This made the bogey position one inch out of the Phase

-------
1 specifications. To avoid potential problems during certification, we are finalizing a minor
change to the bogey position requirement to allow for more clearance of the skirt with the trailer
tires. The Phase 1 bogey position was at 146 ± 4 inches from the rear of the trailer. We are
finalizing 144 ± 4 inches. This still allows for the nominal 146-inch position but also allows for
clearance if needed.
3.2.2.2 Coastdown
The agencies are requiring the use of coastdowns as the reference method with the high-
low analysis method with the analytical solution discussed above. The coastdown test procedure
for tractors is described in 40 CFR 1037.528. This section describes various changes to the
procedure compared to the proposal.
As described in Chapter 3.2.1.1.2.6, the agencies are finalizing changes to the low-speed
range required for the coastdown test as well as the wind constraints. The agencies are finalizing
a low-speed range of 20-10 mph to better account for the drag behavior as a function of yaw
angle. Also, the agencies are adding an additional wind constraint, that the average component
of the wind speed parallel to the coastdown road or track must not exceed 6 mph. This
additional constraint was finalized to be fully outside the new low-speed range, which requires
coasting the vehicle down to 8 mph. Variability will also be reduced by limiting wind speeds,
and the agencies believe that this can be done without sacrificing a significant number of
available days for testing. The data from SwRI showed that 97 percent of the runs that were
within the proposed wind constraints also had an average parallel wind component that was less
than 6 mph. This percentage may be different for other test locations, depending on the direction
of prevailing winds in those areas.
The agencies are requiring filtering of the wind speed, wind direction, air speed, yaw
angle, and vehicle speed using the procedure described in Chapter 3.2.1.1.2.1. This method was
developed with input from the manufacturers for the purposes of standardizing the condition of
coastdown data to be analyzed.
In addition to the tire models tested to support the coastdown testing, the agencies tested
more tire models to understand the variation of the speed dependence of tire rolling resistance.
In total, four steer tire models, four drive tire models, and two trailer tire models, all SmartWay-
verified, were tested, leading to 32 different combinations. The test procedure and calculations
described in Chapter 3.2.1.1.2.4 were applied to each combination to determine its speed
dependence. The rolling resistance increase, AFtrr, was determined for a vehicle weight of
36,000 lbs, distributed at 34 percent, 36 percent, and 30 percent over the steer, drive, and trailer
axles, respectively, and at 65 mph and 15 mph, the midpoints of the high-speed and low-speed
coastdown segments being finalized. Values for AFtrr ranged from 200 to 219 N over all the
various combinations, a spread of about 10 percent. Because of this variation and because tire
rolling resistance characteristics may change in the future when manufacturers will be
performing coastdown tests for the Phase 2 rule, the agencies are requiring measuring tire rolling
resistance as a function of speed according to 40 CFR 1037.528, similar to the method used by
the agencies for this rulemaking.

-------
As described earlier, the agencies are also requiring testing drive axle spin loss tests on
the axle model and configuration to determine spin losses as a function of speed.
The requirements to perform tire rolling resistance and drive axle spin loss tests are a
change from the proposal, where a default spin loss was assumed and no speed dependence for
tire rolling resistance was included.
3.2.2.2.1 Reference Tractors (Fait -aero Testing)
The provisions in this section are particular to the tests to be performed on the reference
tractors to determine /'ait-aem. Given the usefulness in collecting as many coastdown data points
as possible, the agencies are requiring that at least 24 valid runs be conducted to determine a
mean drag area and yaw angle for a given test. Validity is determined by the following:
1)	Runs have no known technical or instrumentation errors,
2)	The yaw angles of the runs lie in a range within ±1° of the median yaw angle of
all the runs collected in one testing period no greater than 12 hours, and
3)	The drag area values within this yaw range are within 2 standard deviations of the
mean drag area of the drag area values within the yaw range.
These criteria establish the important objectives of defining yaw angle limits over which
a mean drag area and yaw angle result can be characterized and eliminating statistical outliers.
These validity criteria were not applied to our coastdown data because the vast majority of the
tests had less than 20 runs. This was due to testing through a broader speed range to evaluate
other aspects of the test procedure, such as speed range segmentation and analysis methods. This
meant that each run took more time to conduct. However, a few tests were conducted with up to
28 runs, and this validity determination is demonstrated below for one of the tests. The CdA and
(absolute value of) yaw angle for every run from one of the tests is shown in Figure 3-19. None
of the runs had any known technical of instrumentation errors.

-------
65
6.0
^ 5.5
CM
i 5.0
<
T3
° 4 5
'-'A
I* •
4.0
3.5
3.0
00	0.5	1.0	1.5	2.0
Yaw angle [deg]
2.5
3.0
Figure 3-19 CdA vs Yaw Angle from One Test of Sleeper Cab 3 Consisting of 28 Runs.
The median of all the yaw angles, ipmed, is 1.45°, which makes the yaw angle range 0.45c
to 2.45°. Points outside of this range were eliminated, as shown in Figure 3-20.
CN
<
E
<
"O
O
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
I
I
I
I




I
I
I




I
I
I




1 vk

J .!

1
1
1
1


!
•
1
1
1




1
1
1
1 1
j
41 med


0.0	0.5	1.0	1.5	2.0
Yaw angle [deg]
2.5
3,0
Figure 3-20 Yaw Angle Limits, Shown By the Dashed Lines - Eliminate the Points in Red From the Final
CdA Result.
Out of the remaining points (blue) the mean and the standard deviation of the C&A values
were calculated to determine the C&A outlier boundaries. With a mean of 5.033 m2 and standard

-------
deviation of 0.143 m2, the boundaries were drawn at 4.747 m2 and 5.319 m2, as shown in Figure
3-21.
7.0
6.5
6.0
„ 5.5
C\l
i 5.0
<
T3
° 4.5
4.0
3.5
3.0
Figure 3-21 The Points In Blue Within The Middle Rectangle Are The Remaining Valid Points To Determine
C&A For The Test.
After eliminating the outliers using the process described above, the mean C&A and mean
yaw angle were calculated from the remaining points to determine the result of the test. In this
case, the final result is (Cd^4)coast = 5.020 m2 at ipsfr = 1.5°.
3.2.2.2.2 Selective Enforcement Audits
The agencies will require manufacturers to perform selective enforcement audits (SEA)
on production tractors selected by the agencies. In general, the procedures will follow those for
the reference tractors. Compliance will be determined by comparing the certification C&A bin
with the bin determined from the SEA. Variability in the coastdown tests are addressed partially
through the implementation of a bin structure, as opposed to using the test result directly.
However, there may be tractors whose results are near the edge of a bin for which the SEA result
could be in the neighboring less aerodynamic bin.
To address this issue, the agencies are finalizing a confidence interval to apply to the top
of the C&A bin, within which an SEA result would be considered to be in compliance. The basis
for this confidence interval, z, is a ¦ + b, where is the standard error of the SEA result, a is
a t-value, and b is an offset to account for testing variability. Details of this approach and the
SEA process for aerodynamic performance are discussed in Section III.E(2)(a)(ix) of the
Preamble.

i
i
i
i

CdA coast
HJ_eff= 1.5°
= 5.020 mA2
i
i
i
i



i
i
i



i
i
i



i
i
i



i
i
i


:
—+
i
i
i
i i
1 • *
r«3
5w
i i
i i
*•
1 \ 1
i » i
.—Hi
1
1 1
1 1
1 1
1 1
1 1
1 #1
:

i
i
i



1
1
1



i
i
i



1
1
1



i
i
i



1
1
1










0.0	0.5	1.0	1.5	2.0	2.5	3.0
Yaw angle [deal

-------
The agencies determined that a value of 1.5 was appropriate for a. This critical t-value
for a failure of 1.5 means that, from the precision error alone, the agencies must have a
confidence level of 93 percent that the test results is above the boundary of the bin declared for
that tractor configuration. This comes from the (one-tailed) probability of approximately 7
percent that a result falls in the tail of a normal distribution for a t-value of 1.5.
In addition to the precision component, the agencies are allowing an offset, b, to be
applied to account for test-to-test variability. The variability of multiple tests of the same tractor
was used to consider value b. As mentioned earlier, Sleeper Cab 3 was tested on multiple days.
Wind conditions varied between each of these tests, causing different effective yaw angles. To
compare the tests with each other, the wind-averaged CdA values were used, after adjustment to
4.5° as described in Section 3.2.1.1.3. For a given alternate method used for the yaw adjustment,
the wind-averaged CdA values varied by a range of 0.11 m2
The coastdown testing at NRC was used to investigate site-to-site variability to inform
the b value. While the agencies anticipate that the manufacturers would use the same test
facilities that they used for their reference tractor tests, they could choose a different site based
on availability or other factors. The coastdown analysis process the agencies are finalizing could
not exactly be used on the NRC data because wind conditions were not always favorable, and an
unequal numbers of runs were conducted in each direction. A matched pair analysis (instead of a
low-pair mean) was used along with the alternate method adjustments that were performed for
the SwRI data in order to compare all results in the wind-averaged drag domain. The wind-
average CdA estimated using the NRC data differed by 0.15 m2 from that using the SwRI data.
As shown in Figure 3-9, the standard error of test decreases as the number of runs in a
test increases. At 24 runs, the standard error is on average, approximately 0.84 percent. For a
given distribution, increasing the number of runs to 100 would roughly halve the standard error
to 0.42 percent, as the standard error decreases with the square root of the number of runs. With
an a value of 1.5, the contribution to the confidence interval, z, of the precision error at the Bin
TTT/TV boundary of 5.6 m2 is approximately 0.04 m2
Since the bin boundaries are expressed to one decimal place, the SEA provision also
allows for rounding, which provides an additional 0.049 m2 Finally, the agencies selected a b
value of 0.03 m2 Combining the selected a and b values, the estimated standard error after 100
tests, and the rounding margin; the estimated confidence interval for a tractor at the Bin III/IV
boundary is 0.12 m2 This in the 0.11-0.15 m2 range estimated by the repeat tests done on
Sleeper Cab 3 at SwRI and NRC and is around 30 percent of the width of Bin IV. The agencies
are finalizing a confidence interval of z = 1.5 ¦ a% + 0.03, which would be applied to the SEA
result when determining compliance as per SEA test procedures in 40 CFR 1037.305.
3.2.2.3 CFD
For Phase 1, we established CFD procedures based on our results and industry
collaboration since there were no standardized practices at the time. In addition, to ensure data
consistency, a minimum set of characteristics and criteria was included for CFD analysis to
ensure that the boundary and surface conditions are not too coarse and thus not representative of
the real tractor and environmental conditions.

-------
For Phase 2, we are finalizing the requirement to use the Society of Automotive
Engineering (SAE) standard for CFD, SAE J2966, as the basis for our CFD procedures.16 We
included a few exceptions and clarifications to SAE J2966 to align with various requirements in
Phase 1 along with new provisions for Phase 2:
•	SAE J2966 contains provisions for both open road and wind tunnel simulations.
We are requiring that the CFD runs must simulate the open road condition.
•	The Reynolds number must be 5.1 million and vehicle speed must be 65 mph in a
full-scale environment. This is to harmonize with various other aspects of the
rulemaking, such as coastdown testing being done at a speed range around 65
mph and GEM GHG results being heavily weighted toward the 65-mph drive
cycle.
•	The output of the CFD must be drag area, not drag coefficient. This is to
harmonize with coastdown testing and GEM inputs, which are in the drag area
domain. This also eliminates the need to determine frontal area for the vehicle.
•	We are retaining Phase 1 grid size requirements for Phase 2, which may be finer
than what is recommended in SAE J2966.
•	Turbulence intensity must be 0.0 percent.
As discussed earlier, the agencies are requiring results from surrogate angle of ±4.5°.
However, CFD simulations may be performed at either +4.5° or -4.5°, but the manufacturer is
responsible for compliance with the average result, as would be determined from on-road
confirmatory and selective enforcement audit (SEA) testing combined with the alternate
aerodynamic methods.
3.2.2.4	Wind tunnel
The agencies are not making any major changes to the wind tunnel specifications from
the proposal. However, as discussed earlier, the agencies are requiring results from surrogate
angles of ±4.5°, instead of the SAE J1252 yaw sweep that was proposed. Also, the test for
Reynolds effects described in Section 7.1 of SAE J1252 will not be required. The CFD
simulations performed by Exa and ARC showed that Reynolds effects are very small in the range
of the Reynolds numbers that are allowed, which is required to be at least 1.0 million. The use of
Fait-aero to adjust back to a coastdown test also mitigates most of these effects; a change in
Reynolds number would require a recalculation of Fait-aero.
3.2.2.5	Aerodynamic Method Adjustment Factor (Fait-aero)
As the agencies showed in Phase 1, and in the various results shown in this Phase 2
analysis, aerodynamic test methods differ in their predictions of drag coefficient.17 On-road
methods, such as coastdown and constant speed tests, are performed in uncontrolled real-world
environments, whereas wind tunnel testing is performed in constrained, controlled conditions,
and CFD is a simulation that attempts to replicate complex aerodynamic events. Different test
methods have differences with regard to environmental conditions, assumptions for non-
aerodynamic drag forces, tunnel geometry, boundary conditions, and simulation characteristics.

-------
These differences can lead to different results, even though they are used to measure or calculate
the same parameter. The agencies acknowledged that there will never be perfect alignment
between the predicted drag area values from the aerodynamic methods even with full,
appropriate correction for every factor, but wanted to allow the use of these methods, which are
currently being used by the manufacturers, to limit test burden for certification.
As a result, for Phase 1, we employed the use of an aerodynamic method adjustment
factor, or /'ait-aem to relate the results from the reference method, a coastdown test, to the results
from the alternative method as a ratio of the coastdown result to the alternate method result for
selected Class 8 high roof sleeper cabs. The Fait-aero is then multiplied by the results generated
using the alternative method for all other OEM configurations. This allowed manufacturers the
convenience and lower test burden of using existing aerodynamic protocols rather than pursuing
extensive data correction to produce equivalent results across the aerodynamic methods.
For Phase 2, we will require the use of data from alternate aerodynamic test methods, and
subsequently the aerodynamic method adjustment factor. An important change is that we are
requiring that this factor be determined at the average yaw angle from the coastdown, not at zero
yaw. This is to recognize that coastdowns are not conducted in zero yaw conditions and
assuming such conditions would add error to the drag area determination. Furthermore, using
the average yaw angle provides manufacturers more flexibility to test in various wind conditions
(within the limits required in the regulation) without the risk of a zero-yaw assumption causing
an incorrect adjustment to the surrogate yaw angle.
3.2.2.6 Certification Calculation Steps
Table 3-11 describes, through a sample calculation, how to calculate the drag area for a
certification tractor using a coastdown reference tractor and an alternate method. This is the
most common way the agencies expect tractor manufacturers to certify their tractors.
Table 3-11 Sample Calculations of Drag Area for Certification Tractor
STEP
VARIABLE
EXAMPLE VALUE OR CALCULATION
Coastdown of reference tractor
(C 'd- I )co(isl
5.208 m2
eff
1.6°
Drag area of reference tractor from
alternate method at positive and negative
effective yaw angle from coastdown
(Ci4)ait at
±1.6°
5.002 m2
Alternate Method Factor
F alt-aero
Falt-aero = (C'd^)coast / (Cd^)alt,±1.6° = 5.208/5.002 = 1.041
Wind-averaged drag area of certification
tractor from alternate method
(Cd^)alt at
±4.5°
5.614 m2
Adjustment to wind-averaged drag; round
final value to one decimal place
(CdA)wa
(Cd^)alt,±4.5° X Falt-aero = 5.614 X 1.041 = 5.8 Hi2
Using the value of 5.8 m2, a manufacturer will then identify the appropriate bin for that
value and use the associated aerodynamic GEM input for determining CO2 emissions and fuel
consumption. If this tractor were a high-roof sleeper cab tractor, it would fall into Bin III, as per
Figure 3-17.

-------
The example described above uses a /'ait-aor<> value from a single reference tractor.
However, the CFD results in Table 3-8 show modest variation of Fait-aero, from 1.13 to 1.20,
among the four tractors evaluated using the Exa software. As a result, the agencies are finalizing
a requirement to test at least one high-roof sleeper cab and one high-roof day cab from each of
model years 2021, 2024, and 2027. The Fait-aero value will be determined using data from these
tractors and any data from selective enforcement audits, as described in 40 CFR 1037.525.
3.2.3 Aerodynamic Test Procedures for Trailers
For Phase 2, the agencies are finalizing CO2 standards reflecting CO2 and fuel
consumption reductions from trailers. Aerodynamic improvements are among the technologies
on which those standards are predicated. New aerodynamic technologies have been
implemented on box vans to improve their aerodynamic efficiency and lower overall tractor-
trailer fuel consumption. In addition, as discussed in Chapter 3.2.2.1, the agencies have assessed
the extent that some of these technologies may migrate to the trailer sector without regulation,
and the extent these improvements should be reflected in the standard trailer used in tractor
certification testing.
Consistent with the tractor regulations, our aerodynamic assessment of different trailer
configurations (applicable to coastdown, constant speed and reduced-scale wind tunnel testing)
and trailer types (applicable to reduced scale wind tunnel testing only) was limited to dry box
van type trailers of several lengths. Specifics on the applicable trailer types and certification
protocols are discussed further in Section IV.D.2, of the Preamble.
The trailer program is also based on wind-averaged drag area. However, unlike the
tractor program, trailer manufacturers will generate A to B test values where the "A" represents a
baseline test and "B" represents the certification trailer; both tests performed using the same test
method and same standard tractor. Subsequently, the trailer manufacturer will input their
specific ACdA value in the GEM-based equation, which will determine the appropriate (\\A
value, based on the analysis discussed in Chapter 2.10 of the RIA. GEM subtracts the kC&A
value from the default C&A value before running to determine the greenhouse gas emissions for
this configuration.
While the aerodynamic test procedures for trailers are based on the same procedures
outlined above for tractors, we have made several simplifications for the trailer program. As
discussed in the following sections, this rulemaking includes default values for tire rolling
resistance effects and axle spin losses in the coastdown test procedures, additional wind
restrictions to ensure consistency between A and B coastdown tests, and interim provisions that
allow manufacturers to use test results without correction to reference method (i.e., no Fait-aero).
3.2.3.1 Standard Tractor Definition for Trailer Testing
Similar to the standard trailer definition for tractor aerodynamic assessment, the agencies
finalized standard tractor definition for trailer aerodynamic assessment. The standard tractor
definition is based on attributes of a high-roof tractor equipped with, at a minimum, a roof
fairing, cab side extenders and fuel tank/chassis skirts. This tractor must meet a Bin III or better
tractor aerodynamic level under either Phase 1 or Phase 2. We believe the majority of tractors in

-------
the U.S. trucking fleet will be Bin III or better in the timeframe of this rulemaking and trailer
manufacturers have the option to choose higher-performing tractors in later years as tractor
technology improves. As with the standard trailer's test article specification for the tractor
program, the aerodynamic specification for the standard tractor here is strictly for the purpose of
certifying trailers beginning in model year 2018. Because the trailer program begins in model
year 2018, before the Phase 2 tractor program, a tractor meeting either the Phase 1 or Phase 2
aerodynamic Bin III or better can be used.
Accordingly, we are finalizing that trailer manufacturers will use this standard tractor
definition with their trailers to conduct A to B testing to capture the ACdA for their trailers that
are either: equipped with aerodynamic devices to meet the trailer standards or are designed to be
more aerodynamic than current, standard trailers.
For trailer OEMs that are certifying a trailer where devices are added to an existing OEM
trailer design, the trailer used for both "A" and "B" tests uses a trailer meeting our standard
trailer definition for 53' dry box vans shown above in Chapter 3.2.2.1, without any trailer
devices installed (i.e., no skirts); with the same standard reference tractor used for both tests.
In contrast, for trailer OEMs that certify a completely new trailer design, the "A" test
uses a trailer meeting our standard trailer definition for 53' dry box vans shown above in Chapter
3.2.2.1, without any trailer devices installed and the "B" test will be the new, OEM trailer
design; with a standard reference tractor used for both tests. In summary, the standard reference
tractor will be used for all trailer OEM "component" level testing; where the "component" in the
B test can range from an add-on trailer device up to a completely different trailer design.
To assist in defining the standard tractor for different trailer types, Table 3-12 shows the
trailers modeled in GEM. As mentioned in Section IV of the Preamble to this rulemaking, the
trailer program will use a GEM-based equation for compliance with is equivalent to using GEM.
Table 3-12 Description of Baseline Tractor-Trailers Used In GEM from Section IV.D(2)(b)(ii), of the
Preamble
TRAILER SUBCATEGORY
FEATURES
Dry van 50 feet and shorter
Class 7 or 8 high-roof day cab, pulling solo 28' dry van
CdA = 5.6, Cit = 6.0 kg/ton
Dry van longer than 50 feet
Class 8 high-roof sleeper cab pulling a solo 53' dry van
CdA = 6.0, Cit = 6.0 kg/ton
Refrigerated van 50 feet and shorter
Class 7 or 8 high-roof day cab pulling a solo 28' ref van
CdA = 5.6, Cit = 6.0 kg/ton
Refrigerated van longer than 50 feet
Class 8 high-roof sleeper cab pulling a solo 53' ref van
CdA = 6.0, Cit = 6.0 kg/ton
Based on this table, we are finalizing standard tractor definitions based on tractor type
and attributes that reflect the types of tractors used for trailers in each of these subcategories.
Specifically, we are finalizing that tractors for all trailers longer than 50 feet shall use a
standard tractor meeting the following criteria for A to B testing: a Class 8, high-roof sleeper
cab, tandem axle tractor that meets a Phase 1 or Phase 2 Bin III or better Class 8 high roof
sleeper cab tractor aerodynamic level. For all trailers 50 feet and shorter, a standard tractor

-------
meeting the following criteria shall be used for A-B testing: Class 7 or 8, high-roof day cab, 4x2
drive axle configuration tractor that meets a Phase 1 or Phase 2 Bin III or better Class 7 or 8 high
roof day cab tractor aerodynamic level.
Table 3-13 Characteristics of Standard Tractor for Aerodynamic Assessment of Trailers
TRAILER LENGTH
STANDARD TRACTOR FEATURES
Box trailers 50 feet and longer
Class 8 high roof sleeper cab
Dual-axle (6x4)
Bin III or better tractor (Phase 1 or Phase 2)
Cab Side extenders
Fuel tank cover/Chassis Skirts
Roof Fairing
Box trailers shorter than 50 feet
Class 7 or 8 high roof day cab
Single drive axle (4x2)
Bin III or better tractor (Phase 1 or Phase 2)
Cab Side extenders
Fuel tank covers/Chassis Skirts
Roof Fairing
3.2.3.2 Aerodynamic Methods
The comprehensive testing program and analysis of the various test procedures described
in the tractor aerodynamics sections above led to finalization of the various aerodynamic test
procedures. The trailer program will use the same coastdown, wind tunnel, and CFD test
procedures, with very minor differences.
To reduce test burden for trailer manufacturers, we are not considering coastdown as the
reference method for the trailer aerodynamic test program. Instead we expect manufacturers will
use wind tunnel or CFD for their aerodynamic assessment. Analysis from RIA 2.10 showed that
there were not drastic differences between these two aerodynamic methods for measuring wind-
averaged drag. As a result, we are finalizing interim provisions allowing methods that meet the
wind tunnel and CFD requirements in 40 CFR 1037.527 and 1037.529 to be used to calculate the
appropriate ACdA without correcting to a reference method. See 40 CFR 1037.150(x)
Coastdowns will still be an allowable method for the trailer program. In particular,
coastdown tests may be useful for technologies that cannot be modeled with sufficient fidelity in
scale wind tunnels or CFD simulations. Additionally, coastdowns will also be options for
confirmatory testing or Selective Enforcement Audits (SEA) due to the complications associated
with requiring scale models or CFD simulations.
The agencies considered using coastdown as a reference method, similar to the tractor
program. However, the use of ACdA was found to amplify some of the variability from the
found in the full-scale coastdown procedure. With the standard error of the C&A result from
coastdowns around 1 percent, this error can propagate significantly when determining ACdA
from two coastdown tests. For example, a coastdown with particular trailer technology measures
a drag area of 5.7 m2 compared to a baseline of 6.0 m2 for a ACdA of 0.3 m2 Assuming a 1
percent standard error on both the baseline and test configurations yields a 0.060 m2 and 0.057
m2 standard errors, respectively. The standard error of the ACdA value is the root mean square of
the two uncertainties, or 0.08 m2, which is about 27 percent of the AC a A value. This relative

-------
standard error of the ACdA would tend to decrease as ACdA increases. However, it could require
many additional coastdown runs to significantly reduce the uncertainty, which could
significantly increase test burden on trailer manufacturers.
The uncertainty propagation, additional test burden, and our results from RIA Chapter
2.10 showing that the near-zero yaw angles of coastdown testing will likely cause an
underestimation of ACdA for some drag-reducing technologies, made us reconsider coastdown as
a reference method for the tractor program.
3.2.3.2.1	Simplifications to the Coastdown Test Procedures for the Trailer
Program
The trailer coastdown test procedure must meet the requirements in 40 CFR 1037.526,
which are very similar to the coastdown procedure for tractors. However, trailer manufacturers
will not be responsible for including the tire rolling resistance or spin loss corrections required
for tractors. Instead, the agencies have developed default values that trailer manufacturers must
apply to their coastdown results. Unlike the tractor tests, default values are reasonable for the
trailer program because the same tires and axle (and tractor) must be used between the baseline t
and the certification tests, which means the same losses would be subtracted out of the
calculations in each case.
As described in Chapter 3.2.2.2, the agencies found a variation of 200 to 219 N in tire
rolling resistance increase from the low-speed range to the high-speed range for 53' box vans.
The same analysis showed a range of 140 to 155 N using the approximate weight distribution for
an empty single 28' box van pulled by a 4x2 high-roof day cab tractor. This was a total weight
of 25,000 lbs, distributed at 38 percent, 37 percent, and 25 percent over the steer, drive, and
trailer axle, respectively. As a result, the agencies are finalizing a default tire rolling resistance
force increase of 215 N for long box vans and 150 N for short box vans using the coastdown
procedure. Though these are default values to be used in the ACdA determination for trailers,
they must be adjusted for ambient temperature, which is based on the temperature correction in
ISO 28580 for truck and bus tires with higher load indices. The temperature correction is
necessary because the ambient temperature could be significantly different between the baseline
test and the certification configuration test.
The agencies are also finalizing a single default drive axle spin loss increase, AFspin for
trailer coastdown procedures. Our default value of 110 N is based on a linear extrapolation of
the proposed value (100 N) to the lower low-speed range that we are finalizing. No temperature
adjustment is required for the drive axle spin loss.
3.2.3.2.2	Wind Considerations in the Coastdown Test Procedures for the
Trailer Program
It should be noted that coastdown tests, as described in the tractor program discussions
above, do not measure wind-averaged drag. As a result, the ACdA from coastdown tests may
understate the aerodynamic improvements of some devices compared to other methods, given
that many trailer technologies are effective at higher yaw angles. The agencies are not requiring
manufacturers to adjust their coastdown results to a wind-averaged result, as described in

-------
Chapter 3.2.1.1.3 for tractors. Instead, our interim trailer provisions (40 CFR 1037.150(x)) allow
manufacturers to choose to adopt their near-zero yaw ACdA value from testing for compliance,
or correct their test result to a wind-averaged result using good engineering judgment.
Yaw effects are also important in terms variability between baseline (A) and certification
(B) results from coastdown tests. Manufacturers performing coastdown tests would follow
similar procedures as those outlined in Chapter 3.2.2.2.1 to determine the validity of their
coastdown runs. In an effort to reduce variability, we are limiting the difference in the effective
yaw angle, xpsff, between the baseline and test configurations to ±1.0 degrees.
3.3 Tire Rolling Resistance
The agencies are finalizing the use of the ISO 28580 test method to determine rolling
resistance and the coefficient of rolling resistance. A copy of the test method can be obtained
through the American National Standards Institute.18 Note that because measurement of rolling
resistance is a continuation of the Phase 1 structure and the Phase 1 requirements serve as the
baseline for Phase 2, the agencies are not including any additional compliance margins in our
analysis of the feasibility of lower rolling resistance tires.
3.3.1 Reason for Using ISO 28580
EPA's SmartWay Partnership Program started to identify equipment and feature
requirements for SmartWay-designated Class 8 over-the-road tractors and trailers in 2006. In
order to develop a tire rolling resistance specification for SmartWay-designated commercial
trucks, EPA researched different test methods used to evaluate tire rolling resistance, reviewing
data and information from tire manufacturers, testing laboratories, the State of California, the
Department of Transportation, tractor manufacturers, and various technical organizations. After
assessing this information, EPA determined that its SmartWay program will use the SAE J126919
tire rolling resistance method until the ISO 2858020 method (at that time under development) was
finalized, at which time the Agency will consider moving to this method for its SmartWay
program.
During this same time period, the National Highway Traffic Safety Administration
(NHTSA) conducted an evaluation of passenger vehicle tire rolling resistance test methods and
their variability.21 Five different laboratory test methods at two separate labs were evaluated.
The NHTSA study focused on passenger tires; however, three of the four test methods evaluated
can be used for medium-duty and heavy-duty tractor tires. The methods evaluated were SAE
J1269, SAE J245222 (not applicable for medium-duty or heavy-duty tractor tires), ISO 1816423
and ISO 28580. The NHTSA study showed significant lab to lab variability between the labs
used. The variability was not consistent between tests or types of tire within the same test. The
study concluded that a method to account for this variability is necessary if the rolling resistance
value of tires is to be compared (NHTSA, 2009). Because of laboratory variability, NHTSA
recommended that the use of ISO 28580 is preferred over the other test methods referenced.
ISO 28580 is preferred because the test method involves laboratory alignment between a
"reference laboratory" and "candidate laboratory." The ISO technical committee involved in
developing this test method also has the responsibility for determining the laboratory that will

-------
serve as the reference laboratory. The reference laboratory will make available an alignment tire
that can be purchased by candidate laboratories. The candidate laboratory will identify its
reference machine. However, at this time, the reference laboratory and alignment tires have not
been identified.
3.3.2	Measurement Method and Results
The ISO 28580 test method includes a specific methodology for "light truck, commercial
truck and bus" tires, and it has 4 measurement methods, force, torque, deceleration, and power,
all of which appear to be suitable for use.
The results of the ISO 28580 test are intended for use in vehicle simulation modeling,
such as the model used to assess the effects of various technology options for national
greenhouse gas and fuel economy requirements for commercial trucks (see Chapter 4). The
results are usually expressed as a rolling resistance coefficient and measured as kilogram per
metric ton (kg/metric ton) or as dimensionless units (1 kg/metric ton is the same as the
dimensionless unit 0.001). The results are corrected for ambient temperature drum surface and
drum diameter as specified in the test method.
3.3.3	Sample Size
The rolling resistance of tires within the same model and construction are expected to be
relatively uniform. In the study conducted by NHTSA, only one individual tire had a rolling
resistance value that was significantly different from the other tires of the same model. The
effect of production variability can be further reduced by conducting three replicate tests and
using the average as the value for the rolling resistance coefficient. Tire models available in
multiple diameters may have different values of rolling resistance for each diameter because
larger diameter tires can produce lower rolling resistance than smaller diameters under the same
load and inflation conditions. If the size range within a tire model becomes large enough that a
given tire size is no longer "substantially similar" in rolling resistance performance to all other
tire sizes of that model, then good engineering judgment should be exercised as to whether the
differently-sized tire shall be treated, for testing and vehicle simulation purposes, as a distinct tire
model. For Class 8 tractors that typically use tires that fit on 22.5" or 24.5" wheels, this situation
might occur with 17.5" tires, more commonly used on moving vans and other applications that
require a low floor.
3.3.4	Tire Size
In Phase 2, the agencies will require manufacturers to enter tire revolutions per mile as a
GEM input. While this rulemaking does not include tire size among the technologies applied to
improve fuel efficiency, this measurement is among the driveline parameters necessary for GEM
to calculate a vehicle speed for a given engine speed. Because there is a wide range of possible
measurements for revolutions per mile, the agencies are specifying a measurement procedure. In
the first 100-200 miles of a tire's useful life, there will be a break-in process during which a
commercial tire can "grow" one to two percent, up to 18 mm. Because this growth affects the air
pressure in the tire, it's important to specify the air pressure under which the measurement is
performed. The Society of Automotive Engineers (SAE) has published recommended practice

-------
J1025 for determining the revolutions per mile of new truck tires.24 Consistent with that
recommended practice, the agencies are finalizing that manufacturers will quantify the
revolutions per mile of the drive tire, NIST traceable within ±0.5 percent uncertainty, by
measuring the number of revolutions of the loaded tire installed on the vehicle per unit distance
to the surface on which it is rolling. Load the tire to the maximum load capacity specified by the
manufacturer, at the corresponding air inflation level. See 40 CFR 1037.520(c).
3.4 Duty Cycle
Certification duty cycles have a significant impact on the GHG emissions from a truck
and how technologies are assessed. Every truck has a different duty cycle in-use. Therefore, it is
very challenging to develop a uniform duty cycle which accurately assesses GHG improvements
and fuel efficiency from technologies relative to their performance in the real world.
The duty cycle attributes that impact a vehicle's performance include average speed,
maximum speed, acceleration rates, deceleration rates, number of stops, road grade, power take-
off operation, and idling time. Average and maximum speeds are the attributes which have the
greatest impact on aerodynamic technologies. Vehicle speed also impacts the effect of low
rolling resistance tires. The effectiveness of extended idle reduction measures is determined by
the amount of time spent idling. Lastly, hybrid technologies demonstrate the greatest
improvement on cycles which include a significant amount of stop-and-go driving due to the
opportunities to recover braking energy. In addition, the amount of power take-off operation will
impact the effectiveness of some vocational hybrid applications.
The ideal duty cycle for a line-haul truck will account for a significant amount of time
spent cruising at high speeds. A pickup and delivery truck duty cycle will contain a combination
of urban driving, some number of stops, and limited highway driving. Finalizing an ill-suited
duty cycle for a regulatory subcategory could drive technologies where they may not see in-use
benefits. For example, requiring all trucks to use a constant speed highway duty cycle will drive
significant aerodynamic improvements. However, in the real world a pickup and delivery truck
may spend too little time on the highway to realize the benefits of aerodynamic enhancements.
In addition, the extra weight of the aerodynamic fairings will actually penalize the GHG
performance of that truck in urban driving and may reduce its freight carrying capability.
3.4.1 Duty Cycles Considered
In Phase 1, the agencies selected three duty cycles for certification testing: the Transient
portion of the California Air Resource Board (CARB) Heavy Heavy-Duty Truck 5 Mode Cycle,
55 mph cruise (without grade), and 65 mph cruise (without grade).
For Phase 2, the agencies carefully considered which duty cycles are appropriate for the
different regulatory subcategories. We considered several duty cycles in the development of the
rulemaking including EPA's MOVES model; the Light-Duty FTP75 and HFET; Heavy-Duty
UDDS; World Wide Transient Vehicle Cycle (WTVC); Highway Line Haul; Hybrid Truck User
Forum (HTUF) cycles; and California CARB's Heavy-Heavy-Duty Truck 5 Mode Cycle.

-------
MOVES Medium-Duty and Heavy-Duty schedules were developed based on three
studies. Eastern Research Group (ERG) instrumented 150 medium and heavy-duty vehicles,
Battelle instrumented 120 vehicles instrumented with GPS, and Faucett instrumented 30 trucks
to characterize their in-use operation.25 ERG then segregated the driving into freeway and non-
freeway driving for medium and heavy-duty vehicles, and then further stratified vehicle trips
according the predefined ranges of average speed covering the range of vehicle operation.
Driving schedules were then developed for each speed bin by creating combinations of idle-to-
idle "microtrips" until the representative target metrics were achieved. The schedules developed
by ERG are not contiguous schedules which will be run on a chassis dynamometer, but are made
up of non-contiguous "snippets" of driving meant to represent target distributions. This gives
MOVES the versatility to handle smaller scale inventories, such as intersections or sections of
interstate highway, independently.
The FTP75 and HFET duty cycles are used extensively for Light-Duty emissions and
CAFE programs. Our assessment is that these cycles are not appropriate for HD trucks for two
primary reasons. First, the FTP has 24 accelerations during the cycle which are too steep for a
Class 8 combination tractor to follow. Second, the maximum speed is 60 mph during the
HWFEC, while the national average truck highway speed is 65 mph.
The Heavy-Duty Urban Dynamometer Driving Cycle was developed to determine the
Heavy-Duty Engine FTP cycle. The cycle was developed from CAPE-21 survey data which
included information from 44 trucks and 3 buses in Los Angeles and 44 trucks and 4 buses in
New York in 1977. The cycle was computer generated and weighted to represent New York
non-freeway (254 sec), Los Angeles non-freeway (285 sec), Los Angeles freeway (267 sec),
New York non-freeway (254 sec) to produce a nearly 50/50 weighting of highway cruise and
urban transient. We believe this cycle is not appropriate for our program for several reasons.
The maximum speed on the UDDS is 58 mph which is low relative to the truck speed limits in
effect today. The 50/50 weighting of cruise to transient is too low for combination tractors and
too high for vocational vehicles and the single cycle does not provide flexibility to change the
weightings. Lastly, the acceleration rates are low for today's higher power trucks.
The World Harmonized WTVC was developed by the UN ECE GRPE group. It
represents urban, rural, and motorway operation. The cycle was developed based on data from
20 straight trucks, 18 combination tractors, and 11 buses total from Australia, Europe, Japan, and
the US. EPA has a desire to harmonize internationally, however, we believe that this single
cycle does not optimally cover the different types of truck operation in the United States and
does not provide the flexibility to vary the weightings of a single cycle.
The Highway Line Haul schedule was created by Southwest Research Institute, using
input from a group of stakeholders, including EPA, Northeastern States for Coordinated Air Use
Management (NESCAUM), several truck and engine manufacturers, state organizations, and
others, for a NESCAUM heavy truck fuel efficiency modeling and simulation project. The cycle
is 103 miles long and incorporates grade and altitude. This cycle is a good representation of line
haul operation. However, the altitude changes cannot be incorporated into a chassis
dynamometer or track test and the cycle is also too long for a typical chassis dynamometer test.

-------
The Calstart-Weststart Hybrid Truck Users Forum is developing cycles to match the
characteristics of truck applications which are expected to be first to market for hybrids. The
cycles include the Manhattan Bus Cycle, Orange County Bus Cycle, Class 4 Parcel Delivery,
Class 6 Parcel Delivery, Combined International Local and Commuter Cycle (CILCC),
Neighborhood Refuse, Utility Service, and Intermodal Drayage cycles. The cycles are very
application-specific and appropriately evaluate each vocation. However, the use of these types
of application specific cycles in a regulatory scheme will lead to a proliferation of cycles for
every application, an outcome that is not desirable.
The CARB 5 Mode cycle was developed by California CARB from heavy-duty truck
data gathered from 1997 through 2000.26 Data was collected from real world driving from
randomly selected vehicles. The data was gathered from 140 heavy-duty trucks by Battelle and
from 31 heavy-duty trucks in a study conducted by Jack Faucett and Associates. The final data
set included 84 of these heavy duty trucks covering over 60,000 miles and 1,600 hours of
activity. The cycles were developed to reflect typical in-use behavior as demonstrated from the
data collected. The four modes (idle, creep, transient, and cruise) were determined as distinct
operating patterns, which then led to the four drive schedules. The cycle is well accepted in the
heavy-duty industry. It was used in the CRC E55/59 Study which is the largest HD chassis
dynamometer study to date and used in MOVES and EMFAC to determine emission rate inputs;
EPA's biodiesel study which used engine dynamometer schedules created from CARB cruise
cycle; the HEI ACES Study: WVU developed engine cycles from CARB 4-mode chassis cycles;
CE/CERT test; and by WVU to predict fuel efficiency performance on any duty cycle from
CARB 5 mode results. The modal approach to the cycles provides flexibility in cycle weightings
to accommodate a variety of truck applications. A downside of the cycle is that it was developed
from truck activity in California only.
3.4.2 Duty Cycles
3.4.2.1 Highway Cruise Cycles
The agencies analyzed the average truck speed limit on interstates and other freeways to
identify the appropriate speed of the highway cruise cycles. State speed limits for trucks vary
between 55 and 75 mph, depending on the state.27 The median urban and rural interstate speed
limit of all states is 65 mph. The agencies also analyzed the speed limits in terms of VMT-
weighting. The agencies used the Federal Highway Administration data on Annual Vehicle
Miles for 2008 published in November 2009 to establish the vehicle miles travelled on rural and
urban interstates broken down by state. The VMT-weighted national average speed limit is 63
mph based on the information provided in Table 3-14. The results of this analysis led to the
adoption of the High Speed (65 mph) and Low Speed (55 mph) Cruise duty cycles in Phase 1.

-------
Table 3-14 VMT-Weighted National Truck Speed Limit
STATE
RURAL
INTERSTATE
SPEED LIMIT
URBAN
INTERSTATE
SPEED LIMIT
RURAL
INTERSTATE
MILES
URBAN
INTERSTATE
AND OTHER
FREEWAYS
MILES
U.S.
WEIGHTED
VMT
FRACTION
RURAL
U.S.
WEIGHTED
VMT
FRACTION
URBAN
VMT
WEIGHTED
SPEED
LIMIT
AL
70
65
5,643
7,950
0.6%
0.8%
0.968
AK
55
55
803
662
0.1%
0.1%
0.086
AZ
75
65
6,966
13,324
0.7%
1.4%
1.474
AR
65
55
4,510
4,794
0.5%
0.5%
0.591
CA
55
55
17,681
123,482
1.9%
13.1%
8.242
CO
75
65
4,409
11,745
0.5%
1.2%
1.161
CN
65
55
715
13,485
0.1%
1.4%
0.837
DE
55
55
-
1,694
0.0%
0.2%
0.099
DC
55
55
-
813
0.0%
0.1%
0.047
FL
70
65
9,591
37,185
1.0%
3.9%
3.279
GA
70
55
9,433
21,522
1.0%
2.3%
1.958
HA
60
60
110
2,403
0.0%
0.3%
0.160
ID
65
65
2,101
1,250
0.2%
0.1%
0.231
IL
65
55
8,972
23,584
1.0%
2.5%
1.996
IN
65
55
7,140
10,850
0.8%
1.2%
1.126
IA
70
55
4,628
2,538
0.5%
0.3%
0.492
KA
75
75
3,242
5,480
0.3%
0.6%
0.694
KE
65
65
6,566
6,834
0.7%
0.7%
0.925
LA
70
70
5,489
7,708
0.6%
0.8%
0.981
ME
65
65
2,207
958
0.2%
0.1%
0.218
MA
65
65
3,484
18,792
0.4%
2.0%
1.537
MS
70
70
1,257
20,579
0.1%
2.2%
1.623
MI
60
60
5,245
20,931
0.6%
2.2%
1.667
MN
70
60
4,150
12,071
0.4%
1.3%
1.077
MS
70
70
4,103
4,004
0.4%
0.4%
0.602

-------
STATE
RURAL
INTERSTATE
SPEED LIMIT
URBAN
INTERSTATE
SPEED LIMIT
RURAL
INTERSTATE
MILES
URBAN
INTERSTATE
AND OTHER
FREEWAYS
MILES
U.S.
WEIGHTED
VMT
FRACTION
RURAL
U.S.
WEIGHTED
VMT
FRACTION
URBAN
VMT
WEIGHTED
SPEED
LIMIT
MO
70
60
5,972
16,957
0.6%
1.8%
1.524
MT
65
65
2,350
343
0.2%
0.0%
0.186
NE
75
65
2,590
1,653
0.3%
0.2%
0.320
NV
75
65
1,826
5,286
0.2%
0.6%
0.510
NH
65
65
1,235
2,574
0.1%
0.3%
0.263
NJ
65
55
1,609
25,330
0.2%
2.7%
1.590
NM
75
65
4,530
2,667
0.5%
0.3%
0.545
NY
65
55
6,176
37,306
0.7%
4.0%
2.604
NC
70
70
5,957
19,216
0.6%
2.0%
1.871
ND
75
75
1,394
374
0.1%
0.0%
0.141
OH
65
65
9,039
27,830
1.0%
3.0%
2.544
OK
75
70
5,029
7,223
0.5%
0.8%
0.937
OR
55
55
4,109
5,734
0.4%
0.6%
0.575
PA
65
55
10,864
21,756
1.2%
2.3%
2.020
RI
65
55
404
2,948
0.0%
0.3%
0.200
SC
70
70
7,355
6,879
0.8%
0.7%
1.058
SD
75
75
1,960
648
0.2%
0.1%
0.208
TN
70
70
8,686
13,414
0.9%
1.4%
1.642
TX
70
70
15,397
71,820
1.6%
7.6%
6.481
UT
75
65
3,117
6,165
0.3%
0.7%
0.674
VT
65
55
1,216
443
0.1%
0.0%
0.110
VA
70
70
8,764
18,907
0.9%
2.0%
2.056
WA
60
60
4,392
15,816
0.5%
1.7%
1.287
WV
70
65
3,195
3,175
0.3%
0.3%
0.456
WI
65
65
5,197
9,139
0.6%
1.0%
0.989
WY
75
75
2,482
474
0.3%
0.1%
0.235
In establishing the highway cruise cycles in Phase 1, we did not address the effect of road
grade on emissions. For Phase 2, where road grade-sensitive technologies like transmission and
driveline improvements are expected to be key technologies utilized for compliance, we have
altered the High Speed Cruise and Low Speed Cruise modes to reflect road grade. Based on
input from trucking fleets and truck manufacturers, we believe this is representative of in-use
operation, wherein truck drivers use cruise control whenever possible during periods of sustained
higher speed driving and variable road grade.
The U.S. Department of Energy and EPA partnered on a project aimed at evaluating,
refining, and developing an appropriate road grade profile for the cruise duty cycles that could be
used in the certification of heavy-duty vehicles to the GHG emission and fuel efficiency Phase 2

-------
standards. The National Renewable Energy Laboratory (NREL) led the project which resulted in
a refinement of the existing highway cruise duty cycles. In the course of their work, NREL
developed several activity-weighted road grade profiles which are representative of U.S. limited-
access highways using high-accuracy road grade and county-specific data for vehicle miles
traveled. This analysis resulted in a single distance-based road grade profile that is
representative of the nation's limited-access highways. To build on the NREL work, the
agencies have incorporated data from the NREL analysis into a different methodology, and have
developed a road grade profile for use with the 55 mph and 65 mph highway cruise cycles.
This following section describes the development of candidate nationally representative,
activity weighted road grade profiles by the agencies as alternatives to the profiles developed by
the National Renewable Energy Laboratory (NREL) and described in report entitled "EPA GHG
Certification of Medium- and Heavy-Duty Vehicles: Development of Road Grade Profiles
Representative of US Restricted Access Highways."28
The agencies' profile is based on the same national half hill database that was used by
NREL, but relies on a different methodology of defining the parameters of its constituent half
hills. It is a goal of the agencies to select the most appropriate road grade profile(s) on the basis
of being nationally representative as well as reasonably similar to real-world driving conditions.
The agencies' profile relies on direct characterization of the whole activity weighted
national half hill population, not on random sampling from that population. This profile consists
of half hills representing unique, yet contiguous segments of that population. Any half hill in
this profile is associated with a single such segment and vice versa. The activity assigned to any
one of those segments, defined in the NREL report as vehicle miles travelled (VMT) by medium-
duty and heavy-duty vehicles on restricted access highways, is calculated as the sum of activities
of its constituent half hills. The parameters of each half hill in the profile, such as length,
average grade, maximum grade or grade distribution, are based on parameters of the half hills
constituting the particular segment. This provides a clear interpretation of why a particular half
hill in the profile is associated with a particular length and grade distribution.
The whole national half hill population is split into segments in such a way that the
lengths of all half hills in the nationally representative profile are directly proportional to the
share of activity their segments represent. This enables proper activity weighting of all profile
parameters and characteristics. For the half hill length the process, illustrated in Figure 3-22,
starts with defining the length of the longest half hill, as this parameter establishes the total
length of the profile. In this particular example, the desired length of the profile was 11 to 12
miles. All half hills spanning the 55 to 75 mph range of truck speed limits in the NREL database
were used. Their lengths ranged from 0.01 to 24.98 miles.

-------
Mean length of half hill in segmefit
>>
>
o
<
1ft
+
I I
¦M
if
4^777
- rfi
Ltf1
.l?1
fi:
n» i i
fi"
%
it$

+
Vt
V!+

4§U.
17j
18

Segment

Segment 20


i i i
0.0
0.5
1.0	1.5	2.0
Half Hill Length, mi
2.5
3.0
Figure 3-22 Segmentation of the Half Hill Datasct
Based on prior experience with profile designs, it was estimated that the longest half hill
in the profile should not exceed 3.0 miles if the above requirement was to be met. After some
iteration, the agencies settled on 2.78 miles in this particular example. This half hill length was
the arithmetic mean length of the 8,402 longest half hills in the database representing 25 percent
of activity. This defined the bounds of Segment 20 (1.77 and 24.98 miles) as shown in Figure
3-22. The ratio of half hill length to normalized activity identified for Segment 20 (namely
2.87/0.25 = 11.48), was subsequently used as the main criterion in defining the lengths of all the
remaining half hills of the profile. Specifically, starting with the lower bound of Segment 20 as
the upper bound of Segment 19, the lower bound of Segment 19 was shifted left until the ratio of
mean half hill length for this segment to normalized activity reached 11.48. This process was
successfully repeated until the whole half hill dataset was exhausted in Segment 1. Detailed
results of this process are provided in Table 3-15.

-------
Table 3-15 Segmentation of the Half Hill Dataset.
Segment
Activity
Activity
(A)
Mean
Half Hill
Distance
(D)
D/A
%
-
mi
-
1
0.1
0.001
0.01
16.43
2
0.3
0.003
0.04
11.49
3
0.7
0.007
0.08
11.48
4
1.1
0.011
0.13
11.48
5
1.5
0.015
0.17
11.48
6
1.8
0.018
0.21
11.48
7
2.1
0.021
0.24
11.48
8
2.4
0.024
0.28
11.48
9
2.7
0.027
0.31
11.48
10
3.1
0.031
0.35
11.48
11
3.4
0.034
0.39
11.48
12
3.9
0.039
0.44
11.48
13
4.4
0.044
0.50
11.48
14
5.0
0.050
0.57
11.48
15
5.7
0.057
0.65
11.48
16
6.6
0.066
0.76
11.48
17
7.8
0.078
0.90
11.48
18
9.6
0.096
1.10
11.48
19
12.7
0.127
1.46
11.48
20
25.0
0.250
2.87
11.48
The activity weighted distribution of half hills identified in the segmentation process
described above is compared in Figure 3-23 to the national activity weighted, cumulative
distribution of half hill length. At first look, the two distributions do not match. For example,
the first one attains the 100 percent of cumulative activity at 2.87 miles and the other at 24.98
miles. However, the half hills identified in the segmentation process represent ranges of half hill
length associated with the respective segments of the national half hill population. The
cumulative activity associated with any of those segments is therefore not represented by the
mean half hill length but by the upper bounds of the respective segments. This is illustrated in
Figure 3-23 using the example of Segment 19. More specifically, the horizontal line
representing cumulative activity associated with Segment 19 intersects the upper bound of this
segment at a point located on the line representing the national distribution of half hill length.
The same is true of all remaining half hills constituting this profile.

-------
100
-I—I
>
-t—I
o
<

-t—i
Z3
E
ZJ
O
Segment 19
Mean
Lower bound
Segment 20
Range: 1.77- 24.98 mi
Upper bound
>
National cumulative
activity distribution
Activity weighted half
distribution in profile
Half Hill Length, rni
3.0
Figure 3-23 Cumulative, Activity Weighted Distributions of Half Hill Lengths in the NREL Database and In
the Candidate Profile.
Once the methodology of defining the lengths of half hills in the nationally
representative, activity weighted profiles was established, a method of designing road grade
contours for the individual half hills was developed. To this end, NREL was requested to
generate road grade data in 0.01 mile increments for the half hill population of each segment of
the profile. This was done to ensure that the contours of each half hill in the profile would
accurately represent the finer details of road grade characteristic of that segment. Activity data
were then applied to the grades of those 0.01 mile roadway sections and cumulative distributions
of road grade were created for each half hill of the profile. One such distribution is shown in
Figure 3-24.

-------
Road Grade, %
Figure 3-24 Example of Cumulative, Activity Weighted Distribution of Road Grade
These distributions were subsequently applied to the respective half hills of the profile.
More specifically, the length of each half hill was split in half and the cumulative distribution of
the 0.01 mile road grade sections was superimposed symmetrically onto each half in such a way
that activity was now represented by the distance driven along the half hill. The symmetrical
arrangement was employed to simulate the shape of half hill contours encountered on roadways
and to ensure smooth transition to and from zero slope at each end of the half hill. This
arrangement enabled half hill specific, activity weighted road grade to be applied individually to
each half hill of the profile. The progression of road grade and the corresponding change in
elevation along the length of a 556 m long half hill are illustrated in Figure 3-25 and Figure 3-26.
The data are plotted in 2 m increments of half hill distance, a format used in the GEM.

-------
Distance Along Half Hill, m
Figure 3-25 Progression of Road Grade along the Length of A Half Hill.
Distance Along Half Hill, m
Figure 3-26 Change in Elevation along the Length of A Half Hill.
While accurately representing the distributions characteristic of the respective segments
of the half hill population, the road grade contours incorporated in the half hills of the profile
included high peaks in the middle section. These peaks were softened by capping them at the
98th percentile of the segment's grade distribution. Hence, grades < 98th percentile were kept
unchanged, while grades > 98th percentile were set equal to the 98th percentile. Capping of the
grade had an insignificant impact on the overall elevation change. An example of such a
modified contour is illustrated in Figure 3-27 and Figure 3-28 for the half hill whose original
parameters were shown in Figure 3-25 and Figure 3-26.

-------

(D
"C
T3
O
CC
3.5
3
2.5
2
1.5
1
0.5
0


lllllllll





t ;
f ^





j
\

























100
200	300	400
Distance Along Half Hill, m
500
600
Figure 3-27 Progression of Road Grade along The Length of a Half Hill (98th percentile version).
Distance Along Half Hill, m
Figure 3-28 Change in Elevation along the Length of a Half Hill (98th percentile version).
Once the lengths and road grade contours of the half hills were defined, they were used to
construct various versions of the profile. In the process, the signs of road grade in the
consecutive half hills were alternated, though this is not a firm requirement, and the half hills
were sequenced in such a way as to ensure that the profile starts and ends at the same elevation.
In fact, in neither of the developed profiles did that overall elevation change exceed 10 cm. In
all, the following four road grade profiles were constructed:

-------
•	Profile A: A 20 km asymmetric profile representing US restricted access
highways with truck speed limits of 55 to 75 mph and its reversed version. The
progressions of road grade and elevation along the length of this profile are shown
in Figure 3-29 and Figure 3-30, respectively.
•	Profile B: A 20 km asymmetric profile representing US restricted access
highways with truck speed limits of 55 to 60 mph and its reversed version.
•	Profile C: A 20 km asymmetric profile representing US restricted access
highways with truck speed limits of 65 to 75 mph and its reversed version.
•	Profile D: A 20 km symmetric profile representing US restricted access highways
with truck speed limits of 55 to 75 mph consisting of a 10 km segment and its
reversed twin. The progressions of the road grade along the length of this profile
is shown in Figure 3-31.
Test Cycle Distance, m
Figure 3-29 Progression of Road Grade along the Length of Profile A.

-------
Test Cycle Distance, m
Figure 3-30 Change in Elevation along the Length of Profile A.
All of the above profiles were evaluated in several powertrains and in the GEM. The
observed effect of the truck speed limit specific profiles on fuel economy proved to be
insignificant both at 55 mph and 65 mph. The asymmetric profiles consistently produced
somewhat lower fuel economy results if the longest half hill was driven up the grade, while the
symmetric profile D approximated the average fuel economy of the two versions of profile A.
Consequently, profile D was selected for use in the regulation. A detailed numerical
representation of this profile is provided in metric units in file
EPASyntheticRoadGradeProfile.xlsx available in the docket.
At proposal the agencies analyzed the effect of different road grade profiles on vehicle
performance as simulated in GEM and described these in a memorandum to the docket titled,
"Possible Tractor, Trailer, and Vocational Vehicle Standards Derived from Alternative Road
Grade Profiles."

-------
-6
Test Cycle Distance, m
Figure 3-31 Progression of Road Grade along the Length of Profile D.
In addition to the agencies completing a thorough analysis in simulation and powertrain
testing, the cycles were shared with manufacturers to evaluate. The summary of this feedback
can be boiled down to three main points. The first was that in-use data from thousands of
tractors with road grade sensors confirm activity weighted road grade distribution of profile D, as
can be seen in Figure 3-32. The second was that compressing the road grade distribution into a
12.5 mile cycle caused unrepresentative rates of change in road grade with distance. The final
comment is that with the addition of profile D and the defined vehicle mass for high-roof sleeper
cabs the engine operation time at peak torque is unrepresentative of in-use engine operation. To
respond to these comments the agencies made the following changes to profile D. The first was
to limit the change in grade versus change in distance to 0.015 percent per meter as shown in
Figure 3-33. This change had a small effect on the long hills but significantly reduced the peak
grade of the shortest half-hills. The second change that was made was adding an additional 1.5
miles at grade equal to or less than 0.5 percent. By doing this the percent time at peak torque
better matched the in-use data reported by manufacturers. With these two changes to profile D,
the road grade distribution was shifted from the activity weighted road grade distribution shown
in Figure 3-34, but this was justified to better align engine operation on the regulatory cycles
with actual engine operation. The final road grade profile and elevation can be seen in Figures 3-
35 and 36.

-------
70%
60%
50%
4%
Figure 3-32 Comparison to Road Grade Distribution of Synthetic Cycles to Volvo In-Use Data of Over 8,000
Trucks.
=5
-Q
~s_
"53
D
>
=3
E
ZJ
O
0.005
Profile D
Profile D mod
0.01 0.015 0.02 0.025 0.03
d(grade(% ))/d(di stance(m))
0.035 0.04
Figure 3-33 Comparison of the Original Profile D Cumulative Road Grade Distribution to the Modified
Profile D.

-------
Road Grade Distribution
100%
90%
80%
70%
activity-weighted 55 mph roads
60%
	activity-weighted 65 roads
50%
Profile D
40%
Profile D mod
30%
20%
10%
0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
Absolute Road Grade
Figure 3-34 Progression of Road Grade along the Length of Profile D.















V \ f Vooro \/o,i
joo\ 15 J
m W Wi
300










Test Cycle Distance, m
Figure 3-35 Progression of Road Grade along the Length of Modified Profile D.

-------
Test Cycle Distance, m
Figure 3-36 Change in Elevation along the Length of Modified Profile D.
3.4.2.2	Transient Cycle
The Phase 1 rule requires use of the Transient portion of the CARB's Heavy Heavy-Duty
Truck 5 Mode Cycle. The agencies have found that this cycle reasonably represents transient
operation of many heavy-duty vehicles, though it is a very short test cycle - less than 3 miles -
and can be driven in roughly 11 minutes. We are not making any changes to that cycle in this
final rule, and will continue to use it when certifying vehicles to the Phase 2 standards.
The agencies launched a project at the National Renewable Energy Laboratory (NREL)
to determine the extent to which the Transient mode of the CARB Heavy-Duty Truck 5 Mode
Cycle is representative of transient operation of Class 2b-8 vocational vehicles. This analysis
was performed using NREL's extensive vehicle activity database and a variety of metrics such as
average driving speed, kinetic intensity, idle time, maximum driving speed and standard
deviation of speed. Although the analysis resulted in the development of a possible new
transient duty-cycle, the Preamble Section V.B explains the reasons why the agencies are not
adopting the new duty-cycle in this rulemaking. Therefore the agencies will finalize the
continued use of the Transient mode of the CARB cycle. The report documenting NREL's
vocational duty cycle work, including the development of a possible new transient cycle, is
available to the public in the docket.29
3.4.2.3	Idle Cycle
We are also finalizing the addition of drive and parked idle-only cycles to determine both
fuel consumption and CO2 emissions when a vehicle is idling in both drive and park in order to
recognize technologies that either reduce the fuel consumption rate or shut the engine off (and
restart) during short-term idle events during the workday. These cycles will not recognize

-------
technologies that allow the main engine to remain off during stationary vehicle operation with a
PTO engaged and performing work. Those technologies are recognized over the Hybrid-PTO
test procedure defined in 40 CFR 1037.540. In these idle-only cycles, based on user inputs
generated through engine testing, GEM will calculate CO2 emissions and fuel consumption at
both zero torque (neutral idle) and with torque set to 100 Nm for use in the CO2 emission
calculation in 40 CFR 1037.510(b). GEM will also calculate reduced CO2 and fueling for stop-
start systems, based on an assumption that the effectiveness will represent a 90 percent reduction
of the emissions that will occur if the vehicle had operated at Curb-Idle Transmission Torque
over the drive idle cycle. This cycle is applicable only for vocational vehicles using either the
Regional, Multi-Purpose, or Urban composite duty cycles. GEM will also calculate reduced
CO2 and fueling for automatic engine shutdown systems, based on an assumption that the
effectiveness will represent an 80 percent reduction of the emissions that will occur if the vehicle
had operated at Neutral Idle over the parked idle cycle.
3.4.3 Weightings of Each Cycle per Regulatory Subcategory
Table 3-16 presents the Phase 1 final GEM duty cycle composite weightings for
vocational vehicles and tractors.
Table 3-16 Phase 1 Vehicle Duty Cycle Composite Weightings
VEHICLE
CATEGORY
PHASE 1 COMPOSITE WEIGHTINGS OF DUTY
CYCLE MODE
Transient
55 mph Cruise
65 mph Cruise
Vocational
42%
21%
37%
Vocational Hybrid
Vehicles
75%
9%
16%
Day Cabs
19%
17%
64%
Sleeper Cabs
5%
9%
86%
The agencies received a comment from American Trucking Associations regarding the
drive cycle weightings. The agencies believe that the study cited by ATA includes weightings of
speed records, which represent the fraction of time spent at a given speed. However, our drive
cycle weightings represent the fraction of vehicle miles traveled (VMT). The agencies used the
vehicle speed information provided in the ATA comments and translated the weightings to
VMT, as shown below in Table 3-17.

-------
Table 3-17 VMT Weighting of Spot Speed Records

SPEEDS > 55
MPH
SPEEDS < 55
MPH
time fraction
57%
43%
total driving hours per day
8
8
hours in a day traveling in this speed
range
4.6
3.4
assumed speed in that speed range
64
30
miles per day in the speed range
292
103
VMT fraction
74%
26%
Based on our assessment, their findings produce weightings that are approximately 74
percent of the vehicle miles traveled are at speeds greater than 55 mph and 26 percent less than
55 mph. In addition, the study cited by ATA represents "Class 8 trucks" which would include
day cab tractors, sleeper cab tractors, and heavy heavy-duty vocational trucks. Based on this
assessment, the agencies do not believe this new information is significantly different than the
drive cycle weightings that were proposed.
3.4.3.1 Phase 2 Vocational Vehicles
3.4.3.1.1 Derivation of the Composite Weightings of the Vocational Driving
Cycles
The U.S. Department of Energy and EPA partnered on a project aimed at identifying
possible segments of vehicles with different driving patterns within the vocational vehicle sector,
for use in identifying regulatory subcategories as part of the certification of heavy-duty vehicles
to the GHG emission and fuel efficiency Phase 2 standards. The National Renewable Energy
Laboratory (NREL) led the project which resulted in identification of three distinct clusters of
vehicles, each with characteristic driving patterns. In the course of their work, NREL developed
distributions of miles accumulated at different speeds by vehicles whose driving statistics most
closely matched the medioid of each cluster. The distance histograms for the 50 best matching
vehicles in each cluster are summarized in Table 3-18. The development of these histograms is
documented in NREL's 2016 vocational drive cycle report.29
Table 3-18 Distance Histograms for Vocational Driving Cycles
SPEED BIN
CLUSTER 1
TOP 50
AVERAGE
CLUSTER 2
TOP 50
AVERAGE
CLUSTER 3
TOP 50
AVERAGE
0+ - 2 mph distance (%)
0.20
0.10
0.03
2+ - 4 mph distance (%)
0.69
0.33
0.11
4+ - 6 mph distance (%)
1.18
0.55
0.19
6+ - 8 mph distance (%)
1.64
0.77
0.21
8+ - 10 mph distance (%)
2.16
0.91
0.26
10+ - 12 mph distance (%)
2.66
1.03
0.30
12+ - 14 mph distance (%)
2.98
1.12
0.34

-------
14+ - 16 mph distance (%)
3.22
1.20
0.36
16+ - 18 mph distance (%)
3.48
1.34
0.40
18+ - 20 mph distance (%)
3.82
1.41
0.44
20+ - 22 mph distance (%)
4.26
1.60
0.53
22+ - 24 mph distance (%)
4.48
1.84
0.56
24+ - 26 mph distance (%)
4.75
2.11
0.65
26+ - 28 mph distance (%)
5.06
2.40
0.77
28+ - 30 mph distance (%)
5.63
2.58
0.91
30+ - 32 mph distance (%)
5.98
2.77
1.04
32+ - 34 mph distance (%)
6.29
3.11
1.13
34+ - 36 mph distance (%)
6.11
3.41
1.16
36+ - 38 mph distance (%)
5.69
3.50
1.19
38+ - 40 mph distance (%)
5.11
3.52
1.31
40+ - 42 mph distance (%)
4.45
3.51
1.45
42+ - 44 mph distance (%)
3.94
3.67
1.55
44+ - 46 mph distance (%)
3.45
3.69
1.59
46+ - 48 mph distance (%)
2.57
3.58
1.68
48+ - 50 mph distance (%)
2.28
3.60
1.82
50+ - 52 mph distance (%)
1.79
3.69
2.01
52+ - 54 mph distance (%)
1.77
4.57
2.69
54+ - 56 mph distance (%)
1.48
5.98
4.01
56+ - 58 mph distance (%)
1.02
7.07
6.16
58+ - 60 mph distance (%)
0.83
7.65
9.19
60+ - 62 mph distance (%)
0.65
7.24
10.03
62+ - 64 mph distance (%)
0.30
4.75
16.96
64+ - 66 mph distance (%)
0.06
3.80
23.61
66+ - 68 mph distance (%)
0.01
1.33
4.63
68+ - 70 mph distance (%)
0.00
0.24
0.55
70+ - 72 mph distance (%)
0.00
0.04
0.11
72+ - 74 mph distance (%)
0.00
0.00
0.03
74+ mph distance (%)
0.00
0.00
0.04
3.4.3.1.2 Composite Weightings of the Vocational Cycles
In order to properly weight the driving time of each vehicle subcategory, the distance
histograms above have been applied to the agencies' regulatory test cycles. For class 2b-7
Multipurpose vehicles and all Regional vehicles, miles accumulated up to 50 mph have been
counted in the weighting for the ARB Transient cycle, miles accumulated between 50 and 60
mph have been counted in the weighting for the 55 mph cycle, and miles accumulated above 60
mph have been counted toward the weighting of the 65 mph cycle. Volvo's data showed that
more miles are accumulated in the 55 mph range for class 8 vehicles than were observed by

-------
NREL. Although both NREL and Volvo data showed vehicles whose behavior would logically
be classified as Urban, accumulating some miles (from one to seven percent) in the 65 mph
range, the agencies are applying a zero weighting factor to the 65 mph cycle for all Urban
vehicles for certification purposes. For class 8 Urban vehicles, miles accumulated up to 48 mph
have been counted in the weighting for the ARB Transient cycle, and miles accumulated above
48 mph have been counted in the weighting for the 55 mph cycle. For classes 2b-7 Urban
vehicles, miles accumulated up to 50 mph have been counted in the weighting for the ARB
Transient cycle, and miles accumulated above 50 mph have been counted in the weighting for
the 55 mph cycle. For class 8 Multipurpose vehicles, we have applied judgment along with
consideration of the weightings that would result from applying cutoffs at 50 mph and 60 mph
and data from Volvo from over 12,000 vehicles. Volvo's data showed the class 8 vehicles they
believe would likely be classified as Multipurpose accumulate an equal amount of distance in the
range of 55 mph as in the range of 65 mph, and an average of transient driving very similar to
that observed by NREL for other multipurpose vehicles. If we applied the weightings as
calculated using the NREL distance histograms for Multipurpose using the 48 mph and 58 mph
cutoffs, the resulting weight of the transient cycle of 50 percent would have been too low
compared to Volvo's data (59 percent), and the 55 and 65 weightings would be equal at 25
percent, but this would be too high compared to Volvo's data showing 21 percent each of those
cycles. Thus we kept the 54 percent of transient and applied an even 23 percent to both the 55
mph cycle and 65 mph cycle to the class 8 Multipurpose vehicles.
In addition to the miles accumulated while driving, NREL provided data on total zero-
speed operation for each cluster of vehicles, as well as percent of a workday spent in out-of-gear
parked idle. The final weightings of the drive idle cycle have been adjusted to account for idling
that occurs over the transient cycle, which includes 15.6 percent zero speed time. In the Phase 1
rule the duty cycles were weighted by distance to properly reflect the vehicle miles traveled by
each category. To incorporate both drive and parked idle emissions, the equation has been
modified to allow for the idle emissions to be time weighted with the driving cycles. The result
of this is that the weighting factors for the driving cycles will still add up to 100 percent while
the total idle weighting factor will be less than 100 percent, reflecting the actual idle time of the
vehicles by category. The agencies have modified the equation in 40 CFR 1037.510(b) to
accommodate both the distance (non-idle) and time based (drive and parked idle) weighting
factors.
The duty cycle weightings for each vocational vehicle test cycle are included in Table
3-19.

-------
Table 3-19 Phase 2 Duty Cycle Mode Composite Weightings
VEHICLE
CATEGORY
DUTY CYCLE MODE
Transient
55 mph
Cruise
65 mph
Cruise
Drive Idle
Parked
Idle
Non-Idle
Vocational
Regional
20%
24%
56%
0%
25%
75%
Vocational Multi-
purpose (2b-7)
54%
29%
17%
17%
25%
54%
Vocational Multi-
purpose (class 8)
54%
23%
23%
17%
25%
54%
Vocational Urban
92%
8%
0%
15%
25%
67%
Vocational Urban
(class 8)
90%
10%
0%
15%
25%
67%
3.5 Tare Weights and Payload
We will continue defining the total weight of a truck as the combination of the truck's
tare weight, a trailer's tare weight (if applicable), and the payload; as it was defined in the Phase
1 rule. The total weight of a truck is important because it in part determines the impact of
technologies, such as rolling resistance, on GHG emissions and fuel consumption. As the HD
program is designed, it is important that the agencies define weights which are representative of
the fleet while recognizing that the final weights are not representative of a specific vehicle. The
sections below describe the agencies' approach to defining each of these weights.
3.5.1 Truck Tare Weights
The tare weight of a truck will vary depending on many factors, including the choices
made by the manufacturer in designing the truck (such as the use of lightweight materials, the
cab configuration (such as day or sleeper cab), whether it has aerodynamic fairing (such as a roof
fairing), and the specific options on the truck.
The Class 8 combination tractor tare weights were developed based on the weights of
actual tractors tested in EPA's coastdown program. The empty weight of the Class 8 sleeper
cabs with a high roof tested ranged between 19,000 and 20,260 pounds. The empty weight of the
Class 8 day cab with a high roof tested was 17,840 pounds. The agencies derived the tare weight
of the Class 7 day cabs based on the guidance of truck manufacturers. The agencies then
assumed that a roof fairing weighs approximately 500 pounds. Based on this, the agencies are
finalizing the tractor tare weights as shown in Table 3-20.

-------
Table 3-20 Tractor Tare Weights
MODEL
TYPE
CLASS 8
CLASS 8
CLASS 8
CLASS 8
CLASS 8
CLASS 7
CLASS 7
Regulatory
Subcategory
Sleeper
Cab High
Roof
Sleeper
Cab Mid
Roof
Sleeper
Cab Low
Roof
Day Cab
High Roof
Day Cab
Low Roof
Day Cab
High Roof
Day Cab
Low Roof
Tractor Tare
Weight (lbs)
19,000
18,750
18,500
17,500
17,000
11,500
11,000
The agencies developed the empty tare weights of the vocational vehicles based on the
EDF report30 on GHG management for Medium-Duty Fleets. The EDF report found that the
average tare weight of a Class 4 truck is 10,343 pounds, of a Class 6 truck is 13,942 pounds, and
a Class 8 truck is 23,525 pounds. The agencies will continue to use the following tare weights:
•	Light Heavy (Class 2b-5) = 10,300 pounds
•	Medium Heavy (Class 6-7) = 13,950 pounds
•	Heavy Heavy (Class 8) = 23,500 pounds
3.5.2	Trailer Tare Weights
We will continue to define the trailer tare weights used in the tractor program based on
measurements conducted during EPA's coastdown testing and information gathered by ICF in
the cost report to EPA, as adopted in the Phase 1 rule.31
A typical 53 foot box (or van) trailer has an empty weight ranging between 13,500 and
14,000 pounds per ICF's findings. The box trailer tested by EPA in the coastdown testing
weighed 13,660 pounds. Therefore, the agencies are defining the empty box trailer weight as
13,500 pounds.
A typical flatbed trailer weighs between 9,760 and 10,760 per the survey conducted by
ICF. EPA's coastdown work utilized a flatbed trailer which weighed 10,480 pounds. Based on
this, the agencies are defining a flatbed trailer weight of 10,500 pounds.
Lastly, a tanker trailer weight typically ranges between 9,010 and 10,500 pounds based
on ICF findings. The tanker trailer used in the coastdown testing weighed 9,840 pounds. The
agencies are defining the empty tanker trailer weight of 10,000 pounds.
3.5.3	Payload
The amount of payload by weight that a tractor can carry depends on the class (or
GVWR) of the vehicle. For example, a typical Class 7 tractor can carry fewer tons of payload
than a Class 8 tractor. Payload impacts both the overall test weight of the truck and is used to
assess the "per ton-mile" fuel consumption and GHG emissions. The "tons" represent the
payload measured in tons.

-------
M.J. Bradley analyzed the Truck Inventory and Use Survey and found that approximately
9 percent of combination tractor miles travelled empty, 61 percent are "cubed-out" (the trailer is
full before the weight limit is reached), and 30 percent are "weighed out" (operating weight
equal 80,000 pounds which is the gross vehicle weight limit on the Federal Interstate Highway
System or greater than 80,000 pounds for vehicles traveling on roads outside of the interstate
system).32 The Federal Highway Administration developed Truck Payload Equivalent Factors to
inform the development of highway system strategies using Vehicle Inventory and Use Survey
(VIUS) and Vehicle Travel Information System (VTRIS) data. Their results, as shown in Table
3-21, found that the average payload of a Class 8 truck ranged from 29,628 to 40,243 pounds,
depending on the average distance travelled per day.33 The same results found that Class 7
trucks carried between 18,674 and 34,210 pounds of payload also depending on average distance
travelled per day.
Table 3-21 National Average Payload (lbs.) per Distance Travelled and Gross Vehicle Weight Group
(VIUS)34

CLASS 3
CLASS 4
CLASS 5
CLASS 6
CLASS 7
CLASS 8
< 50 miles
3,706
4,550
8,023
10,310
18,674
29,628
51 to 100 miles
3,585
4,913
6,436
10,628
23,270
36,247
101 to 200 miles
4,189
6,628
8,491
12,747
30,180
39,743
201 to 500 miles
4,273
7,029
6,360
10,301
25,379
40,243
> 500 mile
3,216
8,052
6,545
12,031
34,210
40,089
Average
3,794
6,234
7,171
11,203
26,343
37,190
The agencies are prescribing a fixed payload of 25,000 pounds for Class 7 tractors and
38,000 pounds for Class 8 tractors for their respective test procedures. These payload values
represent a heavily loaded trailer, but not maximum GVWR, since as described above the
majority of tractors "cube-out" rather than "weigh-out."
NHTSA and EPA will continue to use the payload requirements for each regulatory
subcategory in the vocational vehicle category that were finalized in the Phase 1 rule. The
payloads were developed from Federal Highway statistics based on the averaging the payloads
for the weight classes of represented within each vehicle category.35 The payload requirement is
5,700 pounds for the Light Heavy trucks based on the average payload of Class 3, 4, and 5 trucks
from Table 3-21. The payload for Medium Heavy trucks is 11,200 pounds per the average
payload of Class 6 trucks as shown in Table 3-21. Lastly the agencies are defining 38,000
pounds payload for the Heavy Heavy trucks based on the average Class 8 payload in Table 3-21.

-------
3.5.4 Total Weight
In summary, the total weights of the combination tractors are shown in Table 3-22.
Table 3-22 Combination Tractor Total Weight
MODEL
TYPE
CLASS 8
CLASS 8
CLASS 8
CLASS 8
CLASS 8
CLASS 8
CLASS 7
CLASS 7
CLASS 7
Regulatory
Subcategory
Sleeper
Cab High
Roof
Sleeper
Cab Mid
Roof
Sleeper
Cab Low
Roof
Day Cab
High
Roof
Day Cab
Mid Roof
Day Cab
Low Roof
Day Cab
High
Roof
Day Cab
Mid Roof
Day Cab
Low
Roof
Tractor Tare
Weight (lbs)
19,000
18,750
18,500
17,500
17,100
17,000
11,500
11,100
11,000
Trailer
Weight (lbs)
13,500
10,000
10,500
13,500
10,000
10,500
13,500
10,000
10,500
Payload
(lbs)
38,000
38,000
38,000
38,000
38,000
38,000
25,000
25,000
25,000
Total
Weight (lbs)
70,500
66,750
67,000
69,000
65,100
65,500
50,000
46,100
46,500
The total weights of the vocational vehicles are shown in Table 3-23.
Table 3-23 Vocational Vehicle Total Weights
REGULATORY
LIGHT
MEDIUM
HEAVY
SUBCATEGORY
HEAVY
HEAVY
HEAVY
Truck Tare Weight
10,300
13,950
27,000
(lbs)



Payload (lbs)
5,700
11,200
15,000
Total Weight (lbs)
16,000
25,150
42,000
3.6 Powertrain Test Procedures
In the Phase 1 rule the agencies introduced a powertrain test procedure to allow
manufacturers to generate credits for selling advanced powertrains that reduced CO2 emissions
and fuel consumption. In Phase 2 we bring the powertrain test procedure into the main program
and project that 15 to 30 percent of the vocational vehicles (including both hybrid and non-
hybrid applications) will certify using this method. To accommodate this change we are
finalizing a number of improvements to the test procedure in 40 CFR 1037.550 and reducing the
test burden by only requiring testing of the powertrain that is to be certified. The agencies are
also finalizing modifications to 40 CFR 1037.550 to separate out the hybrid specific testing
protocols.
3.6.1 Reason Behind Use Of Powertrain Test Method for Conventional and Hybrid
Powertrain Certification
The agencies are finalizing a powertrain test option to afford a robust mechanism to
quantify the benefits of CO2 reducing technologies that are a part of the powertrain
(conventional or hybrid), that are not captured in the GEM simulation. Among these
technologies are integrated engine and transmission control and hybrid systems. The largest
change from the Phase 1 powertrain procedure is that only the advanced powertrain will need to

-------
be tested - as opposed to the Phase 1 approach that calculated an improvement factor from the
powertrain results of both the advanced powertrain and a conventional powertrain (often called
A-to-B testing). This change is possible because the GEM simulation tool has been modified to
use powertrain test results in place of the engine fuel map and torque curve of the vehicle that is
to be certified, and thus it can simulate absolute performance of the advanced powertrain.
3.6.2 Use of Generic Vehicles to Apply Measurements Broadly Across All Vehicles
That the Powertrain Will Be Installed In
To limit the amount of testing under this rule, powertrains can be divided into families
and are tested in a limited number of simulated vehicles that will cover the range of vehicles in
which the powertrain will be used.
A matrix of 8 to 9 tests (6 for heavy haul) will be needed per vehicle cycle, to enable the
use of the powertrain results broadly across all the vehicles in which the powertrain will be
installed. The individual tests differ by the vehicle that is being simulated during the test. Table
3-24,
Table 3-25, and Table 3-26 define the unique vehicles being finalized that will cover the
range of coefficient of drag, coefficient of rolling resistance, vehicle mass and axle ratio of the
vehicles that the powertrain will be installed in.
To allow for a generic tire size definition that will cover the tires and axles installed on
the certified vehicles, the agencies are finalizing that each tire radius will be set so that when the
vehicle is cruising at 65 mph the engine speed will equal the corresponding minimum NTE
exclusion speed as defined in 40 CFR part 86.1370(b)(1), intermediate test speed (A, B, or C), or
maximum test speed defined in 40 CFR part 1065. To calculate the tire radius, use the equation
in 40 CFR 1037.550. In case the manufacturer knows the minimum and maximum powertrain
rotational speed to vehicle speed, we are finalizing that the manufacturer may use these known
tire sizes and axle ratios along with one or two equally spaced intermediate points instead of the
predefined tire sizes and axle ratios that are based on engine speed.

-------
Table 3-24 Generic Vehicle Definitions for Class 2b-7 Vehicles

TEST 1
TEST 2
TEST 3
TEST 4
TEST 5
TEST 6
TEST 7
TEST 8
Mass (kg)
7,257
11,408
7,257
11,408
7,257
11,408
7,257
11,408
CdA
6.2
7.7
6.2
7.7
6.2
7.7
6.2
7.7
Tire ('»
(kg/ton)
6.4
7.7
6.4
7.7
6.4
7.7
6.4
7.7
Rotating
Inertia (kg)
340
340
340
340
340
340
340
340
Axle Gear
Efficiency
(%)
95.5
95.5
95.5
95.5
95.5
95.5
95.5
95.5
Axle ratio or
tire radius CI
engines at
engine speed
A
A
B
B
C
C
Maximum
engine
speed
Maximum
engine
speed
Axle ratio or
tire radius SI
engines at
engine speed
Minimum
NTE
exclusion
speed
Minimum
NTE
exclusion
speed
A
A
B
B
C
C
Table 3-25 Generic Vehicle Definitions for Tractors and Class 8 Vocational Vehicles—General Purpose

TEST 1
TEST 2
TEST 3
TEST 4
TEST 5
TEST 6
TEST 7
TEST 8
TEST 9
Mass (kg)
31,978
25,515
19,051
31,978
25,515
19,051
31,978
25,515
19,051
CdA
5.4
4.7
4.0
5.4
4.7
4.0
5.4
4.7
4.0
Tire ('„
(kg/ton)
6.9
6.9
6.9
6.9
6.9
6.9
6.9
6.9
6.9
Rotating
Inertia
(kg)
1,021
794
794
1,021
794
794
1,021
794
794
Axle Gear
Efficiency
(%)
95.5
95.5
95.5
95.5
95.5
95.5
95.5
95.5
95.5
Axle ratio
or tire
radius at
engine
speed
Minimum
NTE
exclusion
speed
Minimum
NTE
exclusion
speed
Minimum
NTE
exclusion
speed
B
B
B
Maximum
engine
speed
Maximum
engine
speed
Maximum
engine
speed

-------
Table 3-26 Generic Vehicle Definitions for Class 8 Combination— Heavy-Haul Vehicle

TEST 1
TEST 2
TEST 3
TEST 4
TEST 5
TEST 6
Mass (kg)
53,751
31,978
53,751
31,978
53,751
31,978
CdA
5.0
5.4
5.0
5.4
5.0
5.4
Tire (
(kg/ton)
6.9
6.9
6.9
6.9
6.9
6.9
Rotating
Inertia
(kg)
1,021
1,021
1,021
1,021
1,021
1,021
Axle Gear
Efficiency
(%)
95.5
95.5
95.5
95.5
95.5
95.5
Axle ratio
or tire
radius at
engine
speed
Minimum
NTE
exclusion
speed
Minimum
NTE
exclusion
speed
B
B
Maximum
engine
speed
Maximum
engine
speed
The main outputs of this matrix of tests is grams of fuel, the average transmission output
shaft speed divided by the average vehicle speed, and positive work measured at the output shaft
of the powertrain. This matrix of test results will then be used to calculate the vehicle's CO2
emissions in GEM taking the work per ton-mile from the GEM simulation and multiplying it by
the interpolated work specific CO2 mass emissions from the powertrain test.
3.6.3 Measurement Method and Results
The agencies are expanding upon the test procedures defined 40 CFR 1037.550 for Phase
1. The Phase 2 expansion will migrate the current Phase 1 test procedure to a new 40 CFR
1037.555 and will modify the current test procedure in 40 CFR 1037.550, allowing its use for
Phase 2 only. The Phase 2 modifications to 40 CFR 1037.550 include the addition of the
rotating inertia of the driveline and tires, and the axle efficiency. This revised procedure also
requires that each of the powertrain components be cooled so that the temperature of each of the
components is kept in the normal operation range.
In addition to changing the vehicle model, we are finalizing changes to the drive model.
The first of these changes is to compensate for the powertrain getting ahead or falling behind in
the duty cycle. Use of this compensation algorithm will ensure that every powertrain drives the
complete distance of the cycle, regardless of whether or not it can maintain the target speed of
the cycle at a given moment in time. The second change that we are finalizing is to allow
overspeeding of the cruise cycle's target speed by 3 mph when the grade is negative. This
change aligns the driver model in GEM with the driver model required for powertrain testing.
Lastly, we are extending the use of the powertrain procedure to PHEV powertrains in
response to comments requesting a defined pathway for demonstration of PHEV emission
reductions. When using this procedure, prior approval of the utility factor curve is required, due
to the diversity of heavy-duty vehicle duty cycles, including miles driven per day. The utility
factor curve must be representative of the daily distance traveled by the vehicles that the PHEV

-------
powertrain will be installed in. The procedure references SAE J2711A, for determining when to
stop testing, and for the determination of the split between charge-depleting and charge-
sustaining operation.
Although detailed equations for the vehicle and driver models can be found in 40 CFR
1037.550, the agencies are recommending that manufacturers use the MATLAB and Simulink
models provided by the agencies. These models can be found at
http://www3.epa.gov/otaq/climate/gem.htm.
Powertrain Test Setup
Powertrain testing contains many of the same requirements as engine dynamometer
testing. The main differences are where the test article connects to the dynamometer and the
software that is used to command the dynamometer and operator demand setpoints. The
powertrain procedure finalized in Phase 2 allows for the dynamometer(s) to be connected to the
powertrain either upstream of the drive axle or at the wheel hubs. The output of the transmission
is upstream of the drive axle for conventional powertrains. In addition to the transmission, a
hydraulic pump or an electric motor in the case of a series hybrid may be located upstream of the
drive axle for hybrid powertrains. If optional testing with the wheel hub is used, two
dynamometers will be needed, one at each hub. Beyond these points, the only other difference
between powertrain testing and engine testing is that for powertrains, the dynamometer and
throttle setpoints are not set by fixed speed and torque targets prescribed by the cycle, but are
calculated in real time by the vehicle model. The powertrain test procedure requires a forward
calculating vehicle model, thus the output of the model is the dynamometer speed setpoints. The
vehicle model calculates the speed target using the measured torque at the previous time step, the
simulated brake force from the driver model, and the vehicle parameters (tire rolling resistance,
drag area, vehicle mass, rotating mass, and axle efficiency). The operator demand that is used to
change the torque from the engine is controlled such that the powertrain follows the vehicle
speed target for the cycle instead of being controlled to match the torque or speed setpoints of the
cycle. The emission measurement procedures and calculations are identical to engine testing.
Conventional Powertrain Test Results
The agencies have performed internal test programs, contracted with outside labs, as well
as collaborated with manufacturers to test out the improvements to the powertrain test procedure.
The following paragraphs summarize some of that work.
The data presented in Figure 3-37 is from a conventional powertrain that consisted of a
Cummins ISX engine and Eaton 10 speed automated manual transmission that was tested in one
of these test programs. This data summarizes the results from three different types of tests. The
first set of data, labeled "Engine Only," was collected from engine tests where the speed and
torque setpoints were determined by GEM. The simulations were done with 9 different vehicle
configurations over the three duty cycles that are being finalized as the certification duty cycles
(55 mph with grade, 65 mph with grade and ARB transient cycle). The "GEM Model" data
A SAE J2711, Recommended Practice for Measuring Fuel Economy and Emissions of Hybrid-
Electric and Conventional Heavy-Duty Vehicles, issued September 2002.

-------
contains the CO2 emissions as determined by GEM using the engine's fuel map and the
transmission's gear ratios using the default shift strategy. The x-axis defines the Powertrain test
results. The data shows that across all three test cycles the powertrain test procedure produces
2.5 percent less CO2 emission than the GEM simulation predicted. One must, however, take into
account the fact that the GEM simulation was done using the engine steady-state fuel map; thus,
the GEM results don't fully take into account the effect of transient fueling on CO2 emissions.
This is evident when looking at the data collected when operating over the transient test cycles
(highest CO2 g/ton-mile results). Here you see that the engine consumed greater than 3 percent
more fuel than GEM predicted. When taking the transient test results into account, the
powertrain performed 5 to 8 percent better than GEM predicted.

140

130
0
1 j
0
s

0
no

CI

N
Q
100


_


90 -


-3

63
SO


S

O
0
70
c

tf

5
60

50

40 -
¦ y = x
~	Engine-Only
*	GEM Model
	Linear (y = x)
Linear (Engine-Only)
- Linear (OEM Model )
= 1.06235
= 11.979
if.'
*'
f
J#—-r.~
~~ „
w.
~ ~
iV®
'S *
y= I.025x
Rz = 0.9924
40 50 60 70 80 90 100 110
Powertrain CO, [g/toii-mile]
120
130
140
Figure 3-37 Engine only and GEM CO2 Results vs. Powertrain.
Since the proposal, the engine and powertrain testing at Oakridge National Laboratory
(ORNL) has been completed. A 2012 Cummins ISX was tested as part of this work using the
engine fuel mapping procedures finalized in this rule; 40 CFR 1036.535 and 1036.540. In
addition to the engine testing, the same engine was paired with an Eaton 10 speed Ultra ShiftPlus
automated manual transmission and an Allison TC10 automatic transmission and tested using the
powertrain procedure in 40 CFR 1037.550. The engine was tested with both the parent rating of
450 Hp and a child rating for the engine of 400 Hp. In addition to the vehicles defined in 40 CFR
1036.540, the powertrains and engine were tested with additional vehicles to test the fit of the

-------
cycle-average and powertrain fuel map. From these results the following conclusions were made
in the final report:6
1.	The powertrain test procedure as defined in 40 CFR 1037.550 is an efficient way to use a
limited amount of test data to predict fuel consumption from many vehicles including
vehicles incorporating child ratings of the engine.
2.	The powertrain procedure constrains variations in drive behavior and dynamometer speed
control to produce representative and repeatable results with a coefficient of variation of
less than 0.5 percent for measured fuel consumption.
3.	The linear fit of fuel, as a function of powertrain N/V and work, fits the powertrain data
well with low error.
4.	The use of a generic powertrain in GEM rather than using the engine's actual torque
curve and transmission's actual gear ratios, has negligible effect on the N/V and work
used to calculate fuel from the powertrain map.
5.	The maximum torque from a powertrain test over the regulatory cycles, is less than half
the theoretical maximum torque determined by multiplying the first gear ratio by the
maximum torque of the engine.
3.6.4	Powertrain Family Definition
To complement the agencies powertrain procedures we are finalizing criteria for defining
a powertrain family. The specifics of these criteria can be found in 40 CFR 1037.231 but
nominally a powertrain family is made up of one engine family and one transmission family.
3.6.4.1	Criteria for Powertrain Families
The regulations in 40 CFR 1037.231 outline the criteria for grouping transmission models
into powertrain families sharing similar emission characteristics. A few of these defining criteria
include the transmission's architecture (manual, automatic, automated manual, dual-clutch and
hybrid), number of gears in the front box, number of meshes in the back box and dry sump
versus regular sump. In addition to the criteria for the transmission, all of the engines in the
powertrain family have to be from the same engine family.
3.6.4.2	Emissions Test Powertrain
We are finalizing that manufacturers select at least one powertrain per powertrain family
for emission testing. The methodology for selecting the test powertrain(s) should be consistent
with 40 CFR 1037.231. The test powertrain(s) should consist of the engine and transmission
combination that results in the highest CO2 emissions.
3.6.5	Vehicle Certification with Powertrain Results in GEM
For manufacturers that choose to use the powertrain method when certifying a vehicle,
the powertrain results from the test will be input into GEM instead of the engine's fuel map,
B Oakridge National Laboratory July 2016, "Powertrain Test Procedure Development for EPA GHG Certification of
Medium- and Heavy-Duty Engines and Vehicles."

-------
torque curve, motoring curve and the transmissions gear ratios. GEM will use the default
powertrain inputs, as described in Table 3-27, and the inputs of the to-be certified vehicle to
calculate the cycle work (W) of the powertrain and the ratio of rotational speed over the vehicle
speed (N/V) as defined by the tire radius and drive-axle ratio.
Table 3-27 GEM Default Parameters for Vehicle Certification Using Powertrain Testing.
REGULATORY CLASS
ENGINE
TRANSMISSION
GEAR RATIOS
Class 8
Combination
Heavy-Haul
2017 MY 15L
Engine with 600
HP
13 speed
Automated Manual
Transmission
12.29, 8.51,
6.05, 4.38, 3.20,
2.29, 1.95, 1.62,
1.38, 1.17, 1.00,
0.86, 0.73
Sleeper Cab - High Roof
Sleeper Cab - Mid Roof
Sleeper Cab - Low Roof
Day Cab - High Roof
Day Cab - Mid Roof
Day Cab - Low Roof
2017 MY 15L
Engine with 455
HP
10 speed
Automated Manual
Transmission
12.8, 9.25, 6.76,
4.9, 3.58,2.61,
1.89, 1.38, 1,
0.73
Class 7
Combination
Day Cab - High Roof
Day Cab - Mid Roof
Day Cab - Low Roof
2017 MY 11L
Engine with 350
HP
HHD
Vocational
Regional Duty Cycle
2017 MY 15L
Engine with 455
HP
Multi-Purpose Duty Cycle
Urban Duty Cycle
2017 MY 11L
Engine with 350
HP
5 speed HHD
Automatic
Transmission
4.6957,2.213,
1.5291, 1,
0.7643
MHD
Vocational
Regional Duty Cycle
Multi-Purpose Duty Cycle
Urban Duty Cycle
2017 MY 7L
Engine with 270
HP
5 speed MLHD
Automatic
Transmission
3.102, 1.8107,
1.4063, 1,
0.7117
LHD
Vocational
Regional Duty Cycle
Multi-Purpose Duty Cycle
Urban Duty Cycle
2017 MY 7L
Engine with 200
HP
In GEM the cycle work from the powertrain testing will be corrected for the electrical
and mechanical accessory power according to the following equation. The accessory power is
defined for each vehicle category in Chapter 4 of this RIA.
T/IA	_ t/it	n +¦ ^trans.out or wheel hub(+)
''"powertrain corrected — ''''test — 'acc ' ^test '	TTj
*'''engine(+)

-------
GEM will use the calculated cycle work and N/V of the powertrain for the to-be certified
vehicle to interpolate the powertrain input table. For vehicle configurations that have cycle work
or N/V outside of the powertrain input table, we are finalizing that the closest end points of the
table be used instead of extrapolating. GEM will then use the following equation to calculate the
CO2 g/ton-mile result per cycle before any technology inputs are applied. Finally the technology
inputs are applied, all the cycles are weighted and the gallons of fuel are then calculated from the
mass of CO2.
e - e [^fuel]	ly	\	mco2
L/cVK/l-linterpolated	TTlileSgem ' P^ylotld ?TlfUel
3.7 Hybrid Powertrain Test Procedures
As discussed in Section V of the Preamble, the agencies see an opportunity to help drive
the technology's advancement by predicating the vocational vehicle standards on a small
adoption rate of hybrid powertrains in this rulemaking. However, since the projected
effectiveness of this technology over the Urban vocational duty cycle is up to 25 percent, the
agencies believe it is no longer appropriate to provide a 1.5 multiplier for credits generated by
vehicles applying this technology. EPA and NHTSA are finalizing two methods to demonstrate
benefits of a hybrid powertrain - powertrain and engine testing.
3.7.1	Measurement Method and Results
The agencies are finalizing that hybrid powertrains be tested just like conventional
powertrains, with the dynamometer connected at either the input shaft of the drive axle or the
input shaft to the wheels, using the powertrain method described in Chapter 3.6 with some
additional requirements for the rechargeable energy storage systems (RESS) net energy change
(NEC) over the test.
We are finalizing the testing of hybrids using the procedures outlined in 40 CFR
1066.501 to determine End-of-Test for charge-depleting operation. The NEC of the RESS must
meet the requirements of SAE J2711 for each test.
3.7.2	Engine Hybrid Method
To address hybrid powertrain system performance for hybrids that recover energy
between the engine and transmission, the agencies will retain the engine hybrid procedures
defined in 40 CFR 1036.525. The control volume for these hybrids is drawn so as to include the
battery, battery support and control systems, power electronics, the engine, motor generator and
hybrid control module. The performance of this system is an engine based evaluation in which
emission rates are determined on a brake-specific work basis. As such, the duty cycles being
finalized to assess this system performance are engine speed and torque command cycles that are
similar but not identical to the cycles used for criteria pollutant standards. In addition to the
cycles being slightly different between the test for GHG emissions and the test for criteria
emissions, the system boundary of the engine for the criteria emission test will remain

-------
unchanged and will not include the hybrid components. It is expected that, parallel engine
hybrids will be the most likely choice for engine-based hybrid certification. Details related to
engine hybrid test procedures may be found in 40 CFR 1036.525.
HytJrid
Cont'd
Module
T rantmtsei on
Rear Wheel Drive
Motor
' Genw-atsr
Internal
Combustion
Engne
Figure 3-38 Engine Hybrid Test Configuration
3.7.3	Removal of the Chassis Test Option for Hybrids
In the Phase 1 rule the agencies finalized a powertrain and chassis test option for hybrid
testing. The agencies will remove the chassis test option for the Phase 2 program because it
appears to be incompatible with the changes regarding use of results from the hybrid test
procedure. In the procedure, the output of the hybrid test is brake specific CO2 emission where
the positive work is measured at the output shaft of the hybrid powertrain. Since work cannot
be measured at this location on a chassis dynamometer without modifying the vehicle, the
agencies are finalizing the removal of the chassis testing option. Another reason for the removal
of the chassis test option is that there are a number of additional sources of variability when
testing a vehicle on a chassis dynamometer. These include electrical and mechanical accessory
load, tire temperature and driver variability to name a few.
3.7.4	Electrified PTO Test Method
A power take off (PTO) is a system on a vehicle that allows energy to be drawn from the
vehicle's drive system and used to power an attachment or a separate machine. Typically in a
heavy-duty truck, a shaft runs from the transmission of the truck and operates a hydraulic pump.
The operator of the truck can select to engage the PTO shaft in order for it to do work, or
disengage the PTO shaft when the PTO is not required to do work. The pressure and flow from
this hydraulic fluid can be used to do work in implements attached to the truck. Common
examples of this are utility trucks that have a lift boom on them, refuse trucks that pick up and
compact trash, and cement trucks that have a rotating barrel. In each case the auxiliary
implement is typically powered by a PTO that uses energy from the truck's primary drive engine.

-------
In most PTO equipped trucks, it is necessary to run the primary drive engine at all times
when the PTO might be needed. This is an unoptimized configuration. Typical PTO systems
require no more than 19 kW at any time, which is far below the optimal operation range of the
primary drive engine of most trucks. Furthermore, in intermittent operations, the primary drive
engine is kept running at all times in order to ensure that the PTO can operate instantaneously.
This results in excess GHG emissions and fuel consumption due to idle time. Additionally,
idling a truck engine for prolonged periods of time while operating auxiliary equipment like a
PTO could cause the engine to cycle into a higher idle speed, wasting even more fuel.
Hybridization and changing the operation of a conventional PTO equipped truck are two
viable means to lower the GHG emissions and fuel consumption in the real world. The test
procedures will allow for manufacturers to quantify the reduction of CO2 emissions and fuel
consumption from more efficient PTO systems.
In Phase 1, hybrid PTO testing was performed either via chassis or powertrain testing of
both the conventional and hybrid systems over the PTO duty cycles described in Appendix II of
40 CFR 1037, in addition to the vehicle duty cycles. An improvement factor was then generated
as described in 40 CFR 1037.615 and applied to the g/ton-mile CO2 emission rate resulting from
the GEM output for the hybrid vehicle as described in 40 CFR 1037.540.
EPA and NHTSA will continue the Phase 1 testing methodology outlined in 40 CFR
1037.540 where A to B testing is used to generate an improvement factor either via powertrain or
chassis testing, but with two changes. The first is how the results are used to calculate the
vehicle's emission result. For Phase 2, the agencies are finalizing that the reduction in emissions
from the electrified PTO system versus the conventional PTO system be subtracted from the
composite emissions result. The second change to the procedure for both Phase 1 and 2, is that
the agencies are now allowing plug-in hybrids to use the results from both the charge sustaining
tests and charge depleting tests to calculate the fuel consumed by the electrified PTO system.
Specifics on the applicability of testing for improved PTOs is discussed further in Chapter V.C
of the Preamble.
With the expansion of the PTO procedure for PHEV PTO systems, NREL and EPA
partnered to develop a utility factor curve to weight fuel consumption from charge-sustaining
tests and charge-depleting tests.c The utility factor curve was developed by analyzing driving,
idling and PTO operation from 85 vehicles over 11 months, resulting in greater than 1500
vehicle days of operation and greater than 70k miles. Once the operation was broken up into
driving, idling and PTO operation, a cumulative distribution of PTO hours per day was created.
Since this distribution contained about ten percent of work days when the PTO was not used,
those days were removed before creating the utility factor curve. Removing the days without
PTO operation was justified because the utility factor curve is only intended to represent the
daily PTO operating time. The second justification for this is that plug-in PTO systems only
provide reduction in fuel consumption when the PTO system is used. Figure 3-39 is a plot of the
utility factor fraction versus charge-depleting PTO operating time that has been finalized in the
Appendix of 40 CFR 1037.
c National Renewable Energy Laboratory July 2016, "Characterization of PTO and Idle Behavior for Utility
Vehicles" NREL/TP-5400-66747.

-------
PTO Utilty Factor Curve
0.9
0.8
s= 0.7
o
o
m 0.6
LL
2 0.5
o
ra
D
0.3
0.2
0.1
0
5
10
15
Charge Depleting Time (Hr)
Figure 3-39 Utility Factor Curve for PTO Operating Time
3.8 Axle Efficiency Test
The agencies are also finalizing a test procedure to measure axle efficiency. See 40 CFR
1037.560. This procedure was developed in part using the draft JRC method incorporated into
their CO2 monitoring of HD vehicles procedure and incorporates modifications based on
consultations with the axle manufacturers. This procedure ultimately provides for the
determination of torque loss versus input speed and input torque for use in the GEM simulation
tool. The procedure prescribes dynamometer set ups for axles with and without lockable
differentials as well as drive-through axles. This procedure puts limitations on the test cell
ambient temperature, sump oil temperature, and requires the use of representative commercially
available axle lubricating oil. The mapping process requires that you map the axle by testing
with an input torque in the range of 0 to 4000 Nm in 1000 Nm steps for tractor and vocational
class 8 single drive axle applications (2000 Nm max for tractor tandem drive and vocational
class 2B through 7 single drive axles) at wheel speeds that range from 50 rpm to the maximum
wheel speed in 100 rpm steps. Statistical analysis of the results are performed based on a 95
percent confidence interval and the result of the analysis is compared against an error limit of
0.10 percent (loaded axle test) and 0.05 percent (unloaded axle test) to minimize testing
variability.

-------
3.9	Transmission Efficiency Test
The agencies are also finalizing a procedure for mapping transmission efficiency. See 40
CFR 1037.565. This procedure ultimately provides for the determination of transmission spin
and total power loss for use in the GEM simulation tool. The procedure prescribes a
dynamometer test set up for transmissions. This procedure puts limitations on the test cell
ambient temperature, sump oil temperature, and requires the use of representative commercially
available axle lubricating oil. Transmission spin loss is determined at transmission input shaft
speeds that include the maximum rated input shaft speed, 600 rpm, and three equally spaced
intermediate speeds up to maximum wheel speed as defined by 40 CFR 1065.510. Transmission
torque loss is determined at one loaded torque setpoint in the range of 75 percent to 105 percent
of the maximum transmission input torque and at one unloaded (zero-torque) setpoint. Statistical
analysis of the results are performed based on a 95 percent confidence interval and the result of
the analysis is compared against an error limit of 0.10 percent (loaded torque setpoint) and 0.05
percent (unloaded torque setpoint) to minimize testing variability.
3.10	HD Pickup Truck and Van Chassis Test Procedure
The agencies are finalizing that HD pickup trucks and vans continue to demonstrate
compliance using the 40 CFR part 1066 chassis test procedures. For each test vehicle from a
family required to comply with the GHG and fuel consumption requirements, the manufacturer
will supply representative road load forces for the vehicle at speeds between 15 km/hr (9.3 mph)
and 115 km/hr (71.5 mph). The road load force will represent vehicle operation on a smooth
level road, during calm winds, with no precipitation, at an ambient temperature of 20 °C (68 °F),
and atmospheric pressure of 98.21 kPa. Road load force for speeds below 9.3 mph may be
extrapolated.
The dynamometer's power absorption will be set for each vehicle's emission test
sequence such that the force imposed during dynamometer operation matches actual road load
force at all speeds. Required test dynamometer inertia weight class selections are determined by
the test vehicle test weight basis using adjusted loaded vehicle weight from which the
corresponding equivalent test weight is determined.
3.10.1 LHD FTP and HWFE Testing
The FTP dynamometer schedule consists of two tests, a "cold" start UDDS test after a
minimum 12-hour and a maximum 36-hour soak according to the provisions of 40 CFR
1066.801, 1066.815, and 1066.816, and a "hot" start test following the "cold" start after a 10
minute soak. Engine startup (with all accessories turned off), operation over the UDDS, and
engine shutdown constitutes a complete cold start test. Engine startup and operation over the
first 505 seconds of the driving schedule complete the hot start test. The driving schedule for
EPA's Urban Dynamometer Driving Schedule is contained in Appendix I of 40 CFR part 86.
The driving schedule is defined by a smooth trace drawn through the specified speed versus time
relationship. The schedule consists of a distinct non-repetitive series of idle, acceleration, cruise,
and deceleration modes of various time sequences and rates.

-------
The Highway Fuel Economy Dynamometer Procedure (HFET) consists of
preconditioning highway driving sequence and a measured highway driving sequence. The
HFET is designated to simulate non-metropolitan driving with an average speed of 48.6 mph and
a maximum speed of 60 mph. The cycle is 10.2 miles long with 0.2 stops per mile and consists
of warmed-up vehicle operation on a chassis dynamometer through a specified driving cycle.
The Highway Fuel Economy Driving Schedule is set forth in Appendix I of 40 CFR Part 600,
while the test is carried out according to 40 CFR 1066.840. The driving schedule is defined by a
smooth trace drawn through the specified speed versus time relationships.
Practice runs over the prescribed driving schedules may be performed, provided an
emission sample is not taken, for the purpose of finding the appropriate throttle action to
maintain the proper speed-time relationship, or to permit sampling system adjustment.
Both smoothing of speed variations and excessive accelerator pedal perturbations are to be
avoided. The driver should attempt to follow the target schedule as closely as possible. The
speed tolerance at any given time on the dynamometer driving schedules specified in Appendix I
of parts 86 and 600 is defined by upper and lower limits in 40 CFR 1066.425. The upper limit is
2 mph higher than the highest point on trace within 1 second of the given time. The lower limit
is 2 mph lower than the lowest point on the trace within 1 second of the given time. Speed
variations greater than the tolerances (such as may occur during gear changes) are acceptable
provided they occur for less than 2 seconds on any occasion. Speeds lower than those prescribed
are acceptable provided the vehicle is operated at maximum available power during such
occurrences.
3.10.2 LHD FTP and HWFE Hybrid Testing
Since LHD chassis certified vehicles share test schedules and test equipment with much
of Light-Duty Vehicle testing, EPA believes it is appropriate to continue to use the Phase 1 test
procedure which references SAE J1711 "Recommended Practice for Measuring the Exhaust
Emissions and Fuel Economy of Hybrid-Electric Vehicles, Including Plug-in Hybrid Vehicles"
instead of SAE J2711 "Recommended Practice for Measuring Fuel Economy and Emissions of
Hybrid-Electric and Conventional Heavy-Duty Vehicles."
3.10.2.1 Hybrid Charge Sustaining Operation - FTP or "City" Test and
HFET or "Highway" Test
The agencies will continue incorporating by reference SAE J1711 Chapters 3 and 4 for
definitions and test procedures, respectively, where appropriate, with the following exceptions
and clarifications.
The agencies are adopting the 1 percent of fuel energy NEC state of charge criteria as
expressed in SAE J1711 and described in 40 CFR 1066.501. The Administrator may approve
alternate NEC tolerances and state of charge correction factors.
Preconditioning special procedures are optional for traditional "warm" test cycles that are
now required to test starting at full RESS charge due to charge depleting range testing. If the
vehicle is equipped with a charge sustain switch, the preconditioning cycle may be conducted per
40 CFR 600. Ill provided that the RESS is not charged. Exhaust emissions are not taken in

-------
preconditioning drives. Alternate vehicle warm up strategies may be approved by the
Administrator.
State of Charge tolerance correction factors may be approved by the Administrator as
described in 40 CFR 1066.501. RESS state of charge tolerances beyond the 1 percent of fuel
energy may be approved by the Administrator.
Due to the nature of PHEV and EV operation, testing may require many more vehicle
miles than conventional vehicles. Furthermore, EVs and PHEVs either do not have engines or
may use the engine for only a fraction of the miles driven.
Electric Vehicles and PHEVs are to be recharged using the supplied manufacturer
method provided that the methods are available to consumers. This method could include the
electricity service requirements such as service amperage, voltage, and phase. Manufacturers
may employ the use of voltage regulators in order to reduce test to test variability with prior
Administrator approval.

-------
References
1	For more information, see CFR Title 40, Part 86.004-28.
2	Zhang, H., Sanchez, J., and Spears, M., "Alternative Heavy-Duty Engine Test Procedure for Full Vehicle
Certification," SAE Int. J. Commer. Veh. 8(2):364-377, 2015, doi: 10.4271/2015-01-2768.
3	Salemme, G., Dykes, E., Kieffer, D., Howenstein, M. et al., "An Engine and Powertrain Mapping Approach for
Simulation of Vehicle C02 Emissions," SAE Technical Paper 2015-01-2777, 2015, doi:10.4271/2015-01-2777.
4	Zhang, H., Sanchez, J., and Spears, M., "Alternative Heavy-Duty Engine Test Procedure for Full Vehicle
Certification," SAE Int. J. Commer. Veh. 8(2):364-377, 2015, doi: 10.4271/2015-01-2768.
5	U.S. E.P. A. Supplemental Aerodynamic Data from EPA Testing, Docket Identification Number EPA-HQ-OAR-
2014-0827-1624, Feb 2016.
6	McAuliffe, Brian. Improving the Aerodynamic Efficiency of Heavy Duty Vehicles: Wind Tunnel Test Results of
Trailer-Based Drag-Reduction Technologies. National Research Council Canada Report LTR-AL-2015-0272. July
2015.
7ICF International, Aerodynamic Trailer Component Assessment and Impact on the Greenhouse Gas Emissions
from Heavy-Duty Combination Vehicles - Computational Fluid Dynamics Simulation, July 2016.
8	Southwest Research Institute, Heavy Duty Tractor Coastdown and Constant Speed Testing - Final Report, July
2016.
9	National Renewable Energy Laboratory, 2014. Technical Report: Appending High-Resolution Elevation Data to
GPS Speed Traces for Vehicle Energy Modeling and Simulation. http://www.nrel.gov/docs/Iyl4osti/61109.pdf.
10	R.M. Young. Product specifications for Ultrasonic Anemometer Model 81000.
http://www.youngusa.eom/products/6/3.html.
11	McAuliffe, B. and Chuang, D. Coast-down and Constant-speed Testing of a Tractor-trailer Combination in
Support of Regulatory Developments for Greenhouse Gas Emissions. National Research Council Canada Report
LTR-AL-2016-0019. May 2016. (Pending publication).
12	2010 NAS Report. Finding 2-4 on page 39.
13	DG Clima, European Commission, 2014. Development and Validation of a Methodology for Monitoring and
Certification of Greenhouse Gas Emissions from Heavy Duty Vehicles through Vehicle Simulations.
http://ec.europa.eu/clima/policies/transport/vehicles/heavy/docs/final_report_co2_hdv_en.pdf.
14	Ben Sharpe (ICCT) and Mike Roeth (North American Council for Freight Efficiency), "Costs and Adoption Rates
of Fuel-Saving Technologies for Trailer in the North American On-Road Freight Sector", Feb 2014.
15	Frost & Sullivan, "Strategic Analysis of North American Semi-trailer Advanced Technology Market", Feb 2013.
16	SAE International, 2013, Guidelines for Aerodynamic Assessment of Medium and Heavy Commercial Ground
Vehicles Using Computational Fluid Dynamics, SAE J2966, 2013-09.
17	EPA, 2011, HD Phase 1 Regulatory Impact Analysis, Figure 3-16. EPA-420-R-11-901.
18	(http://webstore.ansi.org/RecordDetail.aspx?sku=ISO+28580%3a2009).
19	SAE International, 2006, Rolling Resistance measurement Procedure for Passenger Car, Light Truck, and
Highway Truck and Bus Tires, SAE J1269, 2006-09.
20	ISO, 2009, Passenger Car, Truck, and Bus Tyres - Methods of Measuring Rolling Resistance - Single Point Test
and Correlation of Measurement Results: ISO 28580:2009(E), First Edition, 2009-07-01.
21	NHTSA, 2009. "NHTSA Tire Fuel Efficiency Consumer Information Program Development: Phase 1 -
Evaluation of Laboratory Test Protocols." DOT HS 811 119. June, (www.regulations.gov, Docket ID: NHTSA-
2008-0121-0019).
22	SAE International, 1999, Stepwise Coastdown Methodology for Measuring Tire Rolling Resistance, SAE J2452,
1999-06.
23	ISO, 2005, Passenger Car, Truck, Bus, and Motorcycle Tyres - Methods of Measuring Rolling Resistance, ISO
18164:2005(E).
24	SAE International, 2012, Test Procedures for Measuring Truck Tire Revolutions per Kilometer/Mile, SAE J1025,
2012-08.
25	Based on MOVES analysis.

-------
26	Gautam, M., N. Clark, W. Riddle, R. Nine, S. Wayne, H. Maldonado, A. Agrawal, M. Carlock. "Development
and Initial Use of a Heavy-Duty Diesel Truck Test Schedule for Emissions Characterization." SAE Paper 2002-01-
1753. 2002.
27	Governors Highway Safety Association. Speed Limit Laws May 2011. Last viewed on May 9, 2011 at
http://www.ghsa.org/html/stateinfo/laws/speedlimit_laws.html.
28	NREL May 2015, "EPA GHG Certification of Medium- and Heavy-Duty Vehicles: Development of Road Grade
Profiles Representative of US Controlled Access Highways," NREL/TP-5400-63853.
29	NREL June 2016, "The Development of Vocational Vehicle Drive Cycles and Segmentation," NREL/TP-5400-
65921.
30	Environmental Defense Fund. "Greenhouse Gas Management for Medium-Duty Truck Fleets." Viewed at
http://edf.org/documents/10860_fleets-med-ghg-management.pdf. Page 6.
31ICF International. Investigation of Costs for Strategies to Reduce Greenhouse Gas Emissions for Heavy-Duty
On-Road Vehicles. July 2010. Pages 4-16. Docket Identification Number EPA-HQ-OAR-2010-0162-0044.
32	M.J. Bradley & Associates. Setting the Stage for Regulation of Heavy-Duty Vehicle Fuel Economy and GHG
Emissions: Issues and Opportunities. February 2009. Page 35. Analysis based on 1992 Truck Inventory and Use
Survey data, where the survey data allowed developing the distribution of loads instead of merely the average loads.
33	The U.S. Federal Highway Administration. Development of Truck Payload Equivalent Factor. Table 11. Last
viewed on March 9, 2010 at
http://ops.fhwa.dot.gov/freight/freight_analysis/faf/faf2_reports/reports9/s510_ll_12_tables.htm.
34	Excerpted from The U.S. Federal Highway Administration. Development of Truck Payload Equivalent Factor.
Table 11. Last viewed on March 9, 2010 at
http://ops.fhwa.dot.gov/freight/freight_analysis/faf/faf2_reports/reports9/s510_ll_12_tables.htm.
35	The U.S. Federal Highway Administration. Development of Truck Payload Equivalent Factor. Table 11. Last
viewed on March 9, 2010 at
http://ops.fhwa.dot.gov/freight/freight_analysis/faf/faf2_reports/reports9/s510_ll_12_tables.htm.

-------
Chapter 4: Vehicle Simulation Model
4.1 Purpose and Scope
In designing a regulatory GHG emission control and fuel consumption program, it is
necessary to estimate the performance of technologies, verify compliance with the regulatory
standards, and estimate overall benefits of the program. The agencies developed the Greenhouse
gas Emission Model (GEM) to serve these purposes for Phase 1, consistent with
recommendations by the National Academies of Sciences (NAS) to use vehicle simulation to
demonstrate compliance.1 GEM is currently being used to certify the fuel consumption and CO2
benefits of the Phase 1 rulemaking for all heavy duty vehicles except for HD pickups and vans,
which require a chassis dynamometer test for certification. While the version of GEM used in
Phase 1 contained most of the technical and mathematical features needed to run a vehicle
simulation, the model was limited. For example:
•	Only manual transmissions were used in the model for all tractor and vocational vehicle
simulations, which is not always the case for real world applications, especially for
vocational vehicle applications
•	The model did not include engine torque interruption during gear shifting
•	Engine controls were simplified, with no fueling cut-off features
•	Only the agencies' pre-specified engine fuel maps were used
The Phase 1 certification process only required up to five user inputs, and all other
vehicle parameters and their inputs were pre-specified by the agencies.7 Phase 1 GEM only
recognized the benefits of aerodynamics improvement, tire rolling resistance, vehicle speed
limiter, weight reduction, and idle reduction (only for high roof sleeper tractors).
Because the Phase 2 standards are predicated on the performance of a broader range of
technological improvements than Phase 1, including changes to transmissions and better
integration of engines and transmissions, a more comprehensive vehicle simulation model is
required to recognize these technologies' performance^ This chapter describes a new version of
this vehicle simulation model, referred to as Phase 2 GEM.
4.1.1 Summary of GEM Changes between Phase 1 and Phase 2 NPRM
Prior to the proposal, the agencies created an initial version of Phase 2 GEM referred
named "GEM P2vl .0." This version would require manufacturers to perform a new engine
"mapping" test procedure to generate steady-state and transient engine fuel consumption inputs
to represent the actual engine in a vehicle. It also would require entering into GEM new inputs
to describe the vehicle's transmission type and its number of gears and gear ratios. In order to
meet Phase 2 rulemaking requirements in recognizing most of the technologies that are measured
A Under Phase 1, these technologies could be innovative technology credited under that mechanism, but this
mechanism is not generally suited with respect to technologies on whose performance standards are predicated.
Since transmission, driveline, and engine-transmission integration are key parts of projected compliance pathways
for many of the Phase 2 standards, it is appropriate for GEM to recognize these technologies' performance.

-------
in both engine and chassis dynamometers, GEM has been considerably enhanced as opposed to
Phase 1 GEM. Specifically, the agencies implemented the following key technical features into
Phase 2 GEM:
An upgraded engine model, which includes engine fuel cut-off during braking and
deceleration as well as more realistic torque response.
•	Newly developed automatic and automated manual transmissions, with adaptive shifting
algorithms and the option of utilizing manufacturer supplied transmission loss data.
An upgraded driver model with a distance-compensated driver that will drive the
certification drive trace over a prescribed distance regardless of increased drive time due
to vehicle under-performance, for example.
•	New axle model featuring the option of utilizing manufacturer supplied loss data.
Simulation of start-stop, neutral idle, and automatic engine shutdown technologies in
applicable vocational vehicles.
Road grade on 55 and 65mph cruise speed cycles
4.1.2 Summary of GEM Changes between Phase 2 NPRM and Phase 2 FRM
The agencies have continued to make modifications to GEM since proposal. Many of these
iterations were made available for comment, in meetings2'3'4'5'6, and, most recently, viaNODA y.
The agencies received helpful comment on many of these iterations, which comments are
reflected in the promulgated version of GEM. The following summarizes the major changes of
GEM in response to those comments and data submitted to the agencies since the Phase 2
proposal:
•	Modified road grade profile for 55- and 65-mph cruise cycles
•	Revised idle cycles into overall vocational vehicles with new vocational cycle weightings
•	Made significant changes on the input file structures. Examples includes additions of
columns for axle configuration ("6x2," "6x4,""6x4D," "4x2"), and additions of a few
more technology improvement inputs, such as "Neutral Idle and Start/Stop."
•	Made significant changes on output file structures. Examples includes an option to allow
the user to output detailed results on average speed, average work before and after
transmissions, and the numbers of shift for each phase (55 and 65mph cycles and ARB
cycle).
•	Added input file for axle power losses (function of axle output speed and torque) and
replaced single axle efficiency in model with lookup table of torque loss
•	Added simulation of engine torque response with fast response region defined by engine
displacement, and slower torque increase in boosted region with fast falloff on available
torque
•	Added regression models for all certification cycles to allow the user to simulate vehicle
with cycle average approach
•	Added different fuel properties according to 1036.530.
•	Significantly improved shift strategy based on testing data
•	Adjusted transmission loss & inertia scale factors per regulatory subcategory

-------
•	Added optional input table for transmission power loss data
•	Added minimum torque converter lock-up gear input for AT
•	Retuned the default transmission mechanical efficiency based on the testing data
•	Added neutral idle and start/stop features during simulation
•	Adjusted shift and torque converter lockup strategy
4.2 Model Code Description
4.2.1	Engineering Foundations of the Model
EPA developed GEM to be a forward-looking Matlab/Simulink-based model for heavy-
duty (Class 2b-8) vehicle compliance in 2011.7 A more detailed description of this model and its
engineering foundation can be found in Reference 7. The underlying GEM code was originally
developed to simulate a broad range of vehicle speeds over essentially any in-use duty cycle.
However, the official version that is used for determining compliance with the Phase 1 standards
incorporates the regulatory duty cycles into the code. In other words, manufacturers cannot run
other duty cycles with the official version of GEM. We will continue this approach for Phase 2.
4.2.2	Model Components
The GEM architecture is comprised of four systems: Ambient, Driver, Powertrain, and
Vehicle as seen in Figure 4-1. With the exception of Ambient and Driver, each system consists
of one or more subcomponents. The function of each system and its respective component
models, wherever applicable, is discussed in this chapter. Many changes and modifications
described in this chapter have resulted from numerous constructive comments from both public
comments8'9'10,11,12:13 and GEM peer reviews.14 The model has been upgraded to improve its
fidelity and better match the function of the simulated vehicles, which also meets our primary
goal to accurately reflect changes in, and performance of, technology for both stringency
standard development and compliance.
As part of this effort, the agencies devoted substantial effort to accurately track and audit
power flows through the model to ensure conservation of energy. This is critical because this
can allow the user to understand how the energy is balanced across entire vehicle system, and
also help the user to understand which component of the vehicle system contribute the most and
the least energy loss, so that a systematic optimization on a total vehicle can be conducted.

-------
GEM Vehicle Model
[vehicle]
[system_bus] ^
vehicle
GEM_CVM powertrain
vehicle
[driver]
driver

Scope
driver
[powertrain;
[driver]
[ambient]
[vehicle]
[system_bus]
[ambient]
Memory
ambient
n

sy stem_bus
bus_out
sy stem_bus
bus_out
veh_spd_mps
sy stem_bus
mass_out_kg
force_out_N
bus_out
force_in_N
sy stem_bus
massinkg
veh_spd_mps
bus_out
ambient
Figure 4-1 GEM model structure
4.2.2.1	Ambient Subsystem
This system defines ambient conditions such as pressure, temperature, and road gradient,
where vehicle operations are simulated. Just as in Phase 1 GEM, the ambient conditions have
been maintained in accordance with standard SAE practices. The road gradient has been
modified to accept a road grade that varies as a function of distance traveled.
4.2.2.2	Driver Subsystem
The driver model in Phase 2 GEM has been substantially reorganized to simplify
operation and to add support for distance compensated drive cycles. The result is a purely
proportional-integral control driver that features a small look ahead to anticipate the drive cycle,
especially useful at launch where the vehicle response may be delayed due to the large effective
inertia in low gears. The target drive cycle consists of a road grade versus distance and a vehicle
speed target as a function of the time required to achieve those speeds as a function of distance
(i.e. desired cycle time). The drive cycle speed can be converted to a target speed versus
distance travelled; however, such a conversion involves the complication of tracking vehicle stop
times separately since they necessarily occur over zero distance.
Because the simulation itself is time-based, we consider the driver to be distance-
compensated rather than distance-based. The driver always operates in the time domain. To
implement the distance compensated driver, the cycle position is tracked separately from
simulation time, based on the ability of the target vehicle to meet the target speed trace. If the

-------
vehicle meets the target speed trace then cycle position is equivalent to simulation time as there
is no difference in the distance travelled. If the vehicle under-performs the drive cycle, then
cycle position proceeds more slowly than simulation time, forcing the vehicle to drive for a
longer amount of time in order to cover an equivalent distance.
In terms of implementation, to apply distance compensation at each time step, the current
model vehicle speed is divided by the target speed from the drive cycle. This value is integrated
to produce the current cycle time and an updated speed target. The result is that if a simulated
vehicle is traveling at half the drive cycle speed, the simulation will progress through the drive
cycle at half the rate. This behavior is disabled at speeds below 1 meter per second to provide
reasonable launch behavior (which necessarily occurs over short distances), and to maintain
vehicle stop times independent of small discrepancies in total distance travelled.
The addition of distance compensation allows all simulated vehicles to complete an
equivalent trip such as traveling from point A to point B. Without distance compensation, under-
powered vehicles might complete the drive cycle by time but not distance and would have done
less work, as measured in ton-miles, than higher powered vehicles. Distance compensation also
allows for the variation in road grade to be kept in synchronization with the drive cycle speed
trace.
The driver behavior during the steady state cruise cycles has also been modified. To be
more representative of in-use operation for vehicles on descending grades, the modified driver
model will no longer apply the brakes immediately to maintain the speed target. Instead, the
vehicle is allowed to exceed the speed target by 3 mph before the brakes are applied. This
allows the vehicle to carry additional momentum into the next hill.
4.2.2.3 Powertrain Subsystem
The engine, transmission, electric accessories, and portions of the vehicle models from
Phase 1 GEM have been upgraded and merged into a conventional vehicle powertrain system as
shown in Figure 4-2. The conventional powertrain system contains sub-models representing the
engine, transmission, electric accessories, and driveline. Only conventional powertrains are
modeled in Phase 2 GEM, and thus no hybrid power systems are modeled and certified with
GEM. Rather, hybrid powertrains will be certified through the powertrain dynamometer tests
described in Chapter 3 of the RIA.

-------
GEM Conventional Vehicle Powertrain
CD-
sysI«Hi_bLB
»	bus_ou - [control] |
*	syst«n_bus
*	1rq_*i_Nm
~ ref_n_Kgrn2
i —[engine] |
GEM enghe and accessories


trqjn_Nm
1rq_c»Jt_Nm
n«_ou_kgm2
spd_om_iKJps
spdjrwadps

bus out
sy stwnjxjs


trq_o*l_Nm
spdjn_rac£s
met oU kgm2


—frlectilc] |
[engine]
» syslcm_bu$
» tiqjn_Nm
~ ri3 _n _kgm2
CD-
force, oui_N
mass_out_kg
Spdjii_r*JpS
Figure 4-2 GEM Powertrain Model
KjD
-•CD
fOfce_om_N
-~CD
4.2.2.3.1 Engine Subsystem
The engine model is based on a steady-state fuel map and a cycle average fuel map
covering all engine speed and torque conditions with torque curves for wide open throttle (full
load) and closed throttle (no load). The engine fuel map in Phase 2 is the input provided by
users. The steady-state engine fuel map features three sets of data: engine speed, torque, and
fueling rate at pre-specified engine speed and torque intervals, and is being used for 55 and
65mph cruise speed cycles with the road grade. The cycle-average engine fuel map features
three sets of data: ratio of average engine speed to average vehicle speed, work, and fueling rate
over the ARB transient cycle at pre-specified vehicle configurations, and it is only applied to
ARB cycle. As an option, the cycle average map could be also applied to 55 and 65mph cruise
speed cycles (i.e. steady state). In-cylinder combustion processes are not modelled. The engine
speed at a given point in the drive cycle is calculated from the physics of the downstream speeds.
The quantity of torque required is calculated from the driver model accelerator demand, an idle
speed governor, and requests from the transmission during shifts. The torque request is then
limited by a torque response model which is calculated from engine displacement and the
maximum torque curve of the parent engine. Then, torque is limited by the maximum torque
curve for the particular engine calibration provided. The resulting engine torque and speed are
used to interpolate a fuel rate from the fuel map.
The engine model also includes a constant power loss to simulate mechanical accessories.
Most vehicles run a number of accessories that are driven via mechanical power from the engine.
Some of these accessories are necessary for the vehicle to run, like the engine coolant pump or
power steering, while others are only used occasionally and at the operator's discretion, such as
the air conditioning compressor. Some heavy-duty vehicles also use Power Take Off (PTO) to
operate auxiliary equipment, such as refuse compactors or lift forks. These will also be modeled

-------
as a mechanical accessory. The mechanical accessory load is fixed for all vehicles based on
regulatory subcategory, as shown below in Table 4-8, Table 4-9, and Table 4-13. The actual
power consumed for this loss would differ for actual vehicle configurations, but the agencies will
not allow users to change this value in GEM. If a manufacturer uses a hybrid system for power
take-off devices, it may make use of the hybrid-PTO test procedure. See 40 CFR 1037.540.
4.2.2.3.2	Electric Subsystem
The electric subsystem is modeled as a constant power loss. The power consumed for
this loss is based on the vehicle subcategory. It represents the power loss associated with the
starter, electric energy system, alternator and the electrically driven accessories. The
simplification has a negligible impact on the fuel consumption and CO2 emissions results. The
power losses for different vehicles are shown in tables from Table 4-7 to Table 4-15.
4.2.2.3.3	Transmission Subsystem
The transmission subsystem features three different variants representing the major types
of transmissions that are currently in use in the heavy-duty sector, which are the transmission
types on whose performance the various standards are partially predicated. The different
transmission models are built from similar components, but each features a unique control
algorithm matching behaviors observed during vehicle testing.15
4.2.2.3.3.1	Transmission Gear Selection
All of the transmission models use a dynamic shift algorithm to determine the operating
gear over the cycle.15 This employs a rule based approach utilizing the engine torque curve and
fuel map to select gears that optimize efficient engine operation and provide a torque reserve as a
traditional transmission calibration would. The algorithm in GEM attempts to select the
minimum fuel consumption gear after applying constrains on engine speed and torque reserve. It
also allows downshifts due to high driver demand. A detailed description on the shifting strategy
can be seen in the cited article.15
4.2.2.3.3.2	Clutch
The clutch model in Phase 2 GEM replaces the simplified model found in Phase 1 GEM.
The original clutch model had no transition between the fully engaged and fully disengaged
states and provided no commensurate torque impulse to the driveline. In the new clutch model,
engagement and disengagement occur over time, torque is conserved across the clutch, and the
inertial effects of accelerating and decelerating the upstream inertias are captured.
4.2.2.3.3.3	Gearbox
The gearbox model has also been substantially revised in Phase 2 GEM to provide more
realistic operation when shifting. The gearbox contains gear ratios, which are provided by the
user in the transmission input file. In order to model power loss through transmission, a look-up
table that contains the power loss as function of gear number, input speed and input torque is
used. If users provide their own power loss table, this table will incorporate the user-provided
data. The power loss table can be obtained by following the test procedure 40 CFR 1036.565. If

-------
users do not specify the power loss, a default power loss table will be selected within GEM based
on the transmission type, number of gears, gear ratios and engine torque rating. GEM assumes a
higher efficiency for direct drive than in any other gear.
Shifting behavior is more realistic than in Phase 1 GEM with appropriate delays provided
by a synchronizer clutch model. The layout of the gearbox model is most similar to a manual
transmission, but the application for a planetary gearbox is a reasonable approximation as this
type of gearbox can utilize a variety of topologies.
The gearbox rotational inertias are split between a common input inertia, common output
inertia and a gear specific inertia. The common inertias represent rotational inertia always
coupled to the input or output shafts. The gear specific inertias are added or removed as gears
are engaged or disengaged and incur additional losses.
4.2.2.3.3.4	Hydrodynamic Torque Converter
The torque converter model in Phase 2 GEM simulates a lockup-type torque converter.
The torque multiplication and resulting engine load are calculated via torque ratio and K-factor
curves that vary as a function of speed ratio. A base torque ratio curve is used for all simulations
and the K-factor curve is scaled based on the engine torque curve to provide a good match
between the torque converter stall speed and the engine's speed at maximum torque. This
approximation could result in some simulation differences for highly specialized vehicles
equipped with torque converters matched to their specialized duty cycle, but for the vast majority
of vehicles the effect of this approximation on simulated CO2 emissions is negligible.
The lockup behavior of the torque converter is accomplished by integrating a clutch
model similar to the one discussed in Chapter 4.2.2.3.3.2. The torque converter model also
contains a pump loss torque that varies with input speed and torque to simulate the power
required to operate the pump on an automatic transmission.
4.2.2.3.3.5	Automatic Transmission & Control
The automatic transmission (AT) is composed of the torque converter and gearbox
systems discussed above. The gearbox gear specific inertias and spin loss torques are higher as
would be expected from a conventional planetary automatic transmission gearbox. The AT is
allowed to shift under load. During upshifts and torque converter lockup the engine output
torque is slightly reduced to minimize the resultant torque pulse encountered by decelerating the
engine inertia.
The torque converter lockup clutch command is determined based on transmission gear
and gearbox input speed. The threshold at which lockup and unlock are triggered is calculated
from the engine torque curve. In the transmission file the user may specify the minimum gear in
which torque converter lockup may occur. If the value is not specified 3rd gear will be assumed
as the default.

-------
4.2.2.3.3.6	Automated Manual Transmission & Control
The automated manual transmission (AMT) is composed of the clutch and gearbox
systems discussed above with the addition of an inertia brake to slow the gearbox input inertia
during upshifts. The AMT features a low speed clutch engagement routine that feathers the
clutch to get the vehicle moving. The available torque in each gear is constrained based on data
provided by the user in the transmission input file.
Upshifts in tractors and HHD vocational vehicles are handled by limiting the engine load,
disengaging the clutch and shifting the gearbox to neutral. The inertia brake is then applied to
slow the transmission input inertia before the gearbox engages the new gear. With the new gear
engaged the clutch is reengaged and the engine is again allowed to operate at full load. For LHD
and MHD vocational vehicles the clutch is not disengaged during upshifts and the inertia brake
decelerates the inertia of both the engine and transmission input shaft. Downshifts are handled
by shifting the gearbox to neutral and accelerating the gearbox input up to a speed matching the
desired gear using the engine.
4.2.2.3.3.7	Manual Transmission
The results for Manual Transmission vehicles are calculated from the GEM AMT model.
After simulation a 2 percent penalty is applied to non-idle test cycles to match the relative
benefits observed in fleet data, as discussed in RIA Chapter 2.4.
4.2.2.3.4 Driveline
The driveline system contains all of the components that convert the torque at the
transmission output to force at the wheels. This includes drive shafts as well as driven axles,
consisting of a differential, brakes and tires.
4.2.2.3.4.1	Drives haft
The driveshaft is a simple component for transferring torque while adding additional
rotational inertia.
4.2.2.3.4.2	Final Drive
The final drive is modeled as a gear ratio change with power loss table provided in the
optional axle input file. The input can be generated by following the test procedure 40 CFR
1036.560. If users do not provide the power loss table, a default table will be selected based on
the vehicle category and axle ratio.
4.2.2.3.4.3	Brakes
The brake system on each axle applies a torque to the axle proportional to the brake pedal
position from the driver model. The brake model is scaled to match the requirements of the
vehicle.

-------
4.2.2.3.4.4 Tires
The tire component model transfers the torques and rotational inertias from upstream
components to a force and equivalent mass that is passed to the vehicle model. This conversion
uses the loaded tire radius and adds the tire's rotational inertia. The force associated with the tire
rolling resistance is also applied when the vehicle is moving. The magnitude of this force is
determined by the coefficient of rolling resistance, vehicle static mass and current grade.
The new version of GEM will make tire size a manufacturer-specified input rather than
use a predefined value as was done for Phase 1. Manufacturers will specify tire size in terms of
tire revolutions per mile per SAE. Other than this, tires are being modeled the same as in Phase
1.
4.2.2.4 Vehicle
The vehicle system consists of the chassis, its mass and forces associated with
aerodynamic drag, rolling resistance, and changes in road grade. The aerodynamic force is
calculated from the air density, vehicle speed, coefficient of drag, and frontal area. The vehicle
system also contains the vehicle speed integrator that computes acceleration from the input force
and equivalent mass which is integrated to generate vehicle speed and distance traveled.
4.2.3 Capability, Features, and Computer Resources
GEM is a flexible simulation platform that can model a wide variety of vehicles with
conventional powertrains from Class 2b to Class 8. The key to this flexibility is the component
description files that can be modified or adjusted to accommodate vehicle-specific information.
Parameters such as vehicle weight, engine fuel map, transmission gear ratios, tire radius, or axle
ratio can all be changed as inputs by the user in this fashion. The Phase 2 GEM predefines all
drive cycles (the Transient mode defined by the California Air Resources Board (CARB) in their
Highway Heavy-Duty Diesel Transient (HHDDT) cycle, and EPA GEM highway cruise cycles
at 65 mph and 55 mph, each with varying road grade). The agencies also pre-defined many key
parameters, since those parameters are either hard to quantify due to lack of certified testing
procedures or difficult to obtain due to proprietary barriers. One of the examples of this is the
transmission shifting strategies. The transmission shifting strategies were developed as a result
of substantial testing as discussed in the Southwest Research Institute Report16, as well as
confidential discussions with engine, chassis and component manufacturers.
During simulation the GEM tracks the status of many components and the status of all of
the modeled losses. This information provides an energy audit to ensure the model conserves
energy. The fuel consumed and vehicle speed traces are immediately available in the generated
report, while the larger data set is available in a Matlab .mat file or a comma-separated values
(CSV) file.
4.2.3.1 GEM Executable
The final rule requires that vehicle manufacturers use the Phase 2 GEM executable
version, which does not require the use of Matlab or Simulink software, for demonstrating

-------
compliance with the Phase 2 CO2 and fuel consumption standards. In this form, a precompiled
executable format is used for certification. Its computational requirements are minimal. When
using the minimum recommended 2 GHz processor and 4 GB of RAM, a single simulation
should complete in 10 seconds and generate 100 MB output files. Inputs from the manufacturers
are provided in a text file, and the results are available in a generated report.
4.2.3.2 GEM Matlab /Simulink Model
The Matlab/Simulink version of the GEM source code will be released for users that
desire a more detailed look at the inner workings of the model. The system requirements for the
Matlab /Simulink version of GEM include Matlab, Simulink and StateFlow software from
Mathworks (version 2014a or later) and a compatible compiler.17 The recommended hardware
for the Matlab release of GEM is 2+ GHz processor and 4 GB of RAM. The output data from a
GEM simulation into Matlab is approximately 500 MB, depending on the simulation
configuration and outputs selected. Simulations inside Matlab /Simulink using the source code
take approximately 2 to 3 minutes. Although the source code is available to users, all of the
component initialization files, control strategies and the underlying Matlab /Simulink/StateFlow-
based models may not be used for determining compliance. Only the executable version can be
used when producing official HD vehicle certification results. Also, it should be pointed out that
EPA will not provide any technical support for the use of the GEM source code because it is
beyond the scope of the agency's responsibilities and resources.
4.2.4 Peer Review of Phase 2 GEM
Before proposal, Phase 2 GEM was the subject of peer review.14 The peer review was
conducted by an independent contractor and includes four reviewers. Additional details
regarding the peer review and EPA's responses to the peer review comments can be found in the
docket.14
The agencies also met with and received comments from the Engine Manufacturers
Association, along with other industry stakeholders, during the development of Phase 2 GEM,
which identified some areas of concern with GEM. In response, the agencies made significant
improvements that are summarized in Chapter 4.2.1 above.
The agencies released two versions of GEM for public comment with the NPRM (GEM
P2vl .0 and GEM P2vl. 1). After making revisions in response to comments, the agencies
released a few more versions for public comment with the NODA (GEM P2v2.1) GEM P2v2.2,
GEM P2v2.3, and GEM P2v2.4.
4.3 Validation of Phase 2 GEM Simulations
This chapter presents the results of an engineering evaluation of the ability of the
computer model in GEM to accurately simulate actual engine and vehicle performance. Note
that this version differs from the compliance version in that it was possible to use actual values
for vehicle parameters that are locked in the compliance version of GEM. For example,
validations used actual vehicle curb weights. They also incorporated actual shift strategies where
available. This is appropriate because the purpose of the validations was to evaluate the

-------
engineering basis of the model, rather than to evaluate whether the policy of locking certain
parameters is appropriate.
4.3.1 Experimental Tests for GEM Validation
Working with Southwest Research Institute (SwRI), EPA has invested significantly in
various truck tests in order to collect data to validate Phase 2 GEM. The technical research
workshop held at SwRI, San Antonio, TX, December 10-11, 2014, details all of these tests18.
The following truck tests were carried out by SwRI for the purpose of model validation:
Class 6 Kenworth T270 vocational box truck with AT
Class 6 Ford F-650 vocational tow truck with AT
• Class 8 Kenworth T700 line haul truck with AMT
Class 8 Autocar refuse truck with AT
The key specifications for those trucks are listed in Table 4-1.
Table 4-1 Vehicle Specifications of Heavy-Duty Trucks Tested at Southwest Research Institute
Truck
2013 Kenworth
T700
2012 Kenworth
T270
2011 Ford
F-650 Tow truck
2012 Autocar
Refuse
Engine /Rated
Power (hp)
Cummins ISX
455
Cummins I SB
240
Cummins
I SB 270
Cummins
ISL 345
Transmission
Eaton
F016E310C-LAS
Allison
2100
Allison
2200 RDS
Allison 4500
Series
In order to fully validate the model, each truck was tested over six different driving
cycles including regulatory cycles and non-regulatory cycles. They are the EPA GEM 55mph
(with and without grade), EPA GEM 65mph (with and without grade), the transient portion of
the CARB Heavy-Duty Diesel Truck (HDDT) cycle, the World Harmonized Vehicle Cycle
(WHVC), the High-efficiency Truck Users Forum (HTUF) Class 6 Parcel Delivery Cycle, and
the National Renewable Energy Laboratory (NREL) Combined International Local and
Commuter Cycle (CILCC) cycle (which is a utility vehicle cycle). The inclusion of driving
cycles in addition to those used for Phase 1 certification was done to expand the range of
operation, and also to include cycles that are used in the Phase 2 certification process (e.g.
CARB transient, and 55mph and 65mph cruise speed cycles). Some of the cycles are very
aggressive (especially for Class 8 trucks), such as the CILCC and Parcel Delivery cycles, with
many stops and rapid accelerations. EPA evaluated the results from these additional cycles to
improve the modeling capability and its response to highly transient conditions, thus providing
additional confidence in model fidelity. All trucks were tested on a chassis dynamometer. In
addition, the engine and transmission from the F-650 tow truck were tested in a powertrain
dynamometer cell. More information on the vehicle chassis and powertrain dynamometer setups
and tests can found in the Southwest Research Institute Report19.

-------
Considering that procurement of trucks for model validations would be time consuming
and expensive, EPA developed a comprehensive approach to quantify variants of vehicles in
order to maximize testing efficiency. This was done by varying aerodynamic drag and tire
rolling resistance, as well as using weights to simulate different trucks, affording coverage of a
wide range of vehicles. For tractors, varying these parameters also reflects the effects of pulling
different types of trailers, which would impact the combined drag, rolling resistance, and weight
of the vehicle. In this sense, this simultaneously provides validation data for both tractors and
trailers.
Three vehicles were selected for this portion of the test program and they are: the
Kenworth T270 box truck, the Kenworth T700 truck with a 53 foot box trailer, and the F-650
tow truck. The first two trucks were tested on a chassis dynamometer, while the third one was
tested on a powertrain system dynamometer. A total of six drive cycles were tested: EPA GEM
55 mph, EPA GEM 65 mph, CARB HHDDT, WHVC, NREL CILCC, and HTUF Parcel
Delivery cycle. An additional set of six tests were run for each driving cycle listed above to
evaluate the impact of various vehicle characteristics on CO2 emissions and fuel efficiency. The
characteristics of the six test modifications are listed below:
1.	Adding 800 to 1,000 pounds to the vehicle's tare weight depending on the vehicle class
2.	Adding 15 percent to the vehicle-specific constant value representing the vehicle's
frictional load to simulate higher rolling resistance tires
3.	Reducing the vehicle-specific constant value representing the vehicle's frictional load by
15 percent to simulate lower rolling resistance tires
4.	Increasing the vehicle-specific coefficient representing aerodynamic effects by 15 percent
to simulate a higher aerodynamic drag vehicle
5.	Decreasing the vehicle-specific coefficient representing aerodynamic effects by 15
percent to simulate a lower aerodynamic drag vehicle
6.	Running a new set of road load coefficients, to represent a vehicle configuration
optimized for fuel efficiency for each vehicle that was tested, which consists of the
lowest rolling resistance as well as the lowest aerodynamic drag coefficient
Three valid replicate tests were conducted for each vehicle and characteristic over each
driving cycle. A valid replicate was defined as a successful test run in which all data was
collected without regeneration of the diesel particulate filter. The following parameters were
measured or recorded during all tests:
•	Vehicle speed as a function of time
•	Engine fuel rate as a function of time
•	Engine speed as a function of time

-------
•	Gear number as a function of time
•	Engine load (Nm) as a function of time
•	Emissions (NOx, HC, CO, CO2, N2O, CH4) as a function of time in g/s
•	Measured cycle fuel economy (MPG) and emissions (NOx, HC, CO, CO2, PM, N2O,
CH4)
•	Grade as function of time for the cycle with road grade if tested
In addition to all these tests mentioned above, two identical trucks have been used to
validate GEM in a real-world driving route. The specification of these trucks are defined as
follows.
¦	Kenworth T700 with 2012 Cummins ISX15
-	450 hp @ 1800 RPM
-	1550 lb-ft @ 1100 RPM (calibration verified with INSITE)
-	Engine family CCEXH0912XAP
-	Controlled parts list (CPL) 3719
-	Fuel rate code FR10993, Cal P/N CL10135.25
-	Eaton F016E310C-LAS UltraShift 10 speed automated manual
-	Axle ratio 3.36
-	Tire size 295/75R22.5
¦	Trailers ballasted for GCVW of approximately 61,000 lbs
¦	Test weight without aero = 60,440 lbs (27,415 lbs)
¦	Test weight with aero = 61,240 lbs (27,778 lbs)
4.3.2 Results of the GEM Validations
The validation process was comprehensive, featuring three levels of validations. The first
level of the validation is the modeling using the exact same engine fuel maps and transmission
shifting tables obtained from manufacturers when GEM is used to model these vehicles. This
level of the comparisons between testing and simulations are the most critical among the three,
because this level of validation can directly point out the fidelity or issues of the model. The
second level is modeling of the vehicle with the agency's pre-defined shifting strategy, called

-------
auto-shift, when simulation results are compared to the testing results obtained only from
powertrain tests.15 The only difference between the first and second level validations is the
shifting strategy. The third level of the validation is modeling of a real-world driving route with
two well specified trucks, followed by relative comparisons between simulations and testing
results.
4.3.2.1 Validations Using the Exact Engine and Transmission Information
This section describes the first level of validations. Taking into account all of the vehicles
and test configurations mentioned above, more than 130 vehicle variants were tested, allowing
GEM to be comprehensively validated against a very well-defined and robust set of test data.
The results displayed in through Figure 4-3 through Figure 4-6 show the of comparison between
the GEM simulations and testing data of the Class 8 Kenworth T700 truck, Class 6 Ford F-650
tow truck, Class 6 Kenworth T270 box truck, and Autocar refuse truck respectively. In all
figures shown here for 55 and 65mph cycles, road grade is not included.
li
10
9
~ 55MPH
8
7
9?
a CARB
Q.
x Parcel
6
x Utility
5
WHVC
4
-•-1:1-5%
3
-••1:1+5%
2
2
4
6
8
10
Chassis Dynamometer [MPG]
Figure 4-3 GEM Validation against Class 8 Kenworth T700 Truck Chassis Tests

-------
16
15
14
~ 55MPH
12

-------
/
/
/• /
Q.
~ /
s s
/
55MPH
65MPH
CARB
x Utility
WHVC
	1:1
	1:1-5%
— ••1:1+5%

~ /

V s
's"
S/'
¦ s '
•Z''
4	5
Chassis Dynamometer [MPG]
Figure 4-6 GEM Validation against Autocar Refuse Truck Chassis Tests
A review of the data indicates that there is good agreement between the GEM simulations
and testing data obtained over the wide range of vehicles and conditions. In general, the
accuracy of the model simulations against the testing data is very well controlled with an error of
less than ±5 percent, although there are a few outliers of the transient simulation cases for Class
8 trucks due to the nature of the high variability of chassis dynamometer tests. The range of
vehicles tested and simulated included vehicles that varied in terms of all of the regulatory
inputs. Thus, the agencies believe that the accuracy of GEM is sufficient to simulate the benefits
of the range technologies that form the basis of the standards.
Figure 4-7 shows the overall comparison between the simulation and test results when
combining all of the testing and simulation into one figure. Overall, the simulation and test
result correlate well.

-------
14
	1:1
LU
1:1+5%
— 1:1-5%
~ Results
2
4
6
8
10
12
14
16
Experimental Tests (MPG)
Figure 4-7 Comparison of Model Simulations and Chassis Test Results for the 130 Vehicle Test
Configurations
While it is encouraging that GEM accurately simulates overall vehicle performance in an
absolute sense, it is actually more important that GEM is accurate in relative comparisons. This
is because the agencies used the same version of GEM to calculate the stringency of the
standards as was used to evaluate baseline performance for this rulemaking. The ultimate
purpose of this new version of GEM will be to evaluate changes or additions in technology, and
compliance is demonstrated on a relative basis to the numerical standards that were also derived
from GEM. The importance of relative comparisons can be further explained with the following
simplified example.
Assume you have two simulation models: one that says a baseline vehicle with Bin 3
aerodynamics and a conventional automatic transmission would have CO2 emissions of 90 g/ton-
mile, and another that said the same vehicle would have emissions of 95 g/ton-mile. Assume
also that there was a similar vehicle that was the basis of the new standards that had Bin 4
aerodynamics and a dual-clutch transmission. If both models simulated the second vehicle as
being 10 g/ton-mile better than the baseline vehicle, then the models would work equally well for
compliance as long as they were also used to set the standards. With the first model, we would
set the standard at 80 g/ton-mile. And with the second, we would set the standard at 85 g/ton-
mile. In both cases, manufacturers adding Bin 4 aerodynamics and dual-clutch transmissions
would meet the standard. In other words, the two models would be equivalent in terms of
measuring the effect of the change in technology on emissions, even though the absolute values
differ.

-------
As is shown below, GEM indeed performs better in this relative sense. The results from
the T700 and T270 trucks, and powertrain tests for the F-650 tow truck shown in Figure 4-3
through Figure 4-5, can also be presented in a format to evaluate GEM's ability to measure the
relative impact of a technology. Table 4-2 shows an example of relative comparisons to
illustrate how the relative comparison is done with the T700 truck. For simplicity, only the
results from the Class 8 T700 tractor on the 65mph cycle are shown in this table. The column
labeled as Chassis Test Fuel Economy Result (MPG) shows the testing results, while the column
with GEM Fuel Economy Result (MPG) shows the GEM simulation results. Each row
represents a single change to the vehicle configuration, relative to the baseline case. The "Delta"
in the last column is the difference between the impact of the vehicle configuration change as
measured on the chassis dynamometer and simulated in GEM (which sometimes differs from the
apparent delta due to rounding). For example, the row with the "+15 percent Crr" variable
compares GEM results to chassis test results for a vehicle that is the same as the baseline vehicle
except that it has tires with a coefficient of rolling resistance 15 percent higher than the baseline
vehicle. For this example, chassis testing indicates the change in rolling resistance increases fuel
consumption for this cycle by 3.9 percent, while GEM predicts it would increase by 4.9 percent,
but the delta difference is only 1.0 percent as shown in the last column.
Table 4-2 Sample of Relative Comparisons for T700 Truck
Drive
Cycle
Vehicle
Attribute
Variables
Chassis
Test Fuel
Economy
Result
(MPG)
GEM Fuel
Economy
Result
(MPG)
Impact of
Variable on
Chassis
Test Result
Impact of
Variable on
GEM
Simulation
Result
Delta
65 mph
Baseline
6.84
6.61
0.0%
0.0%
0.0%
65 mph
+907 kg
6.86
6.55
-0.3%
0.9%
-1.2%
65 mph
+15% Crr
6.57
6.28
3.9%
4.9%
-1.0%
65 mph
-15% Crr
7.27
6.96
-6.3%
-5.3%
-1.0%
65 mph
+15% Cd
6.31
6.05
7.7%
8.4%
-0.7%
65 mph
-15% Cd
7.63
7.25
-11.5%
-9.8%
-1.8%
65 mph
Optimized
Package
8.08
7.65
-18.1%
-15.8%
-2.3%
The same methodology was applied to all other cases including three different trucks, and
six driving cycles and six vehicle variables. The differences between the chassis test and GEM
results from all of these comparisons are plotted in Figure 4-8.

-------
to
|2
T3
C
rc
c


'5
_ro

-------
Therefore, considering the favorable comparison between the powertrain tests and GEM
simulation results, it is fair to say that the overall accuracy of the GEM to represent the relative
changes in fuel economy of a real world vehicle should be in the range of ±2-3 percent.
Figure 4-7 and Figure 4-8, respectively, show the GEM accuracy against over 130
vehicle variants on an absolute and a relative basis. All are done in a total vehicle configuration,
which includes all vehicle components, such as engine, transmission, and driveline. Since
certification would be done in a total vehicle form for CO2 emissions and fuel efficiency, these
types of comparisons are the most important because they demonstrate that GEM is capable of
capturing the impact on the total vehicle CO2 emissions and fuel consumption due to technology
improvement of individual components. In order to show the fidelity of GEM in modeling
individual components in a more detailed level, the comparisons for the key components must be
demonstrated as well. Displayed in Figure 4-9 through Figure 4-11 are the comparisons of
engine speed, fuel rate, and transmission gear numbers as function of time over the CARB
HHDDT cycle for Class 8 T700 truck.
2200 r	r	1	r		,	?	-
2000
1800
g- 1600
CL
£
-o 1400

-------
25
GEM
Test Data
0	100 200 300 400 500 600 700
Time (s)
Figure 4-10 Engine Fuel Rate Comparisons Over the WHVC for a Class 8 T700 Truck
	GEM
Test Data




1 1




,n








!
i




J 1
I

!






J

h
















i

|
!





I

]







i



i
i
i
0	100 200 300 400 500 600 700
Time (s)
Figure 4-11 Transmission Gear Comparisons Over the WHVC for a Class 8 T700 Truck

-------
As shown in Figure 4-9 through Figure 4-11, reasonably good comparisons between
GEM simulations and tests are obtained. GEM basically can capture detailed behaviors of the
engine and transmission. To further provide a more complete picture of the GEM validations,
Figure 4-12 and Figure 4-13 show another set of examples for an F-650 tow truck. Shown in
these two figures are the comparisons of engine speed and transmission output shaft torque over
the World Harmonized Vehicle Cycle (WHVC) between powertrain dynamometer tests and
GEM results. As can be seen from these two figures, reasonable comparisons are again obtained
between GEM and actual test results.
2500
2000
2000
Time (s)
Figure 4-12 Engine Speed Comparisons Over the WHVC for an F-650 Tow Truck

-------



GEM
^ 1 jUU
CD
=3
I

1 1
. I I ll 111.
II. „ .

iT'P i '| lHi m*
liiiiiliilihjUil
•i
i
2000
1 JUU
1 nnn
¦HI' 11 II
VI ' 1 1

XVJVJVJ
Time (s)
Figure 4-13 Transmission Output Torque Comparisons Over the WHVC for an F-650 Tow Truck
4.3.2.2 Validations Using Pre-default Shifting Strategy for All Transmissions
The second part of validations is to use the agency pre-defined shifting strategy for all
transmissions15 to conduct all simulations. Since the difference between the first level and
second level validations is only in transmission shifting, it is recognized that the absolute
comparisons would not be too good for certain conditions, specifically for transient cycles. For
the cruise speed cycles, such as 55 and 65mph, the gear would stay on the top one or two gears.
There is little or no shifting involved. As such, the comparisons between the shifting tables
obtained from manufacturers and case with the agency's default shifting strategy should be about
the same. Because of this reason, the second part of validations only focuses on the new data
obtained from powertrain tests. No attempt was made to validate the model against chassis
dynamometers. Display in Figure 4-14 is the comparisons between GEM with auto-shift and
powertrain tests for ISB engine with Allison transmissions.

-------
8%








6%
• 55mph

•
• 65mph
•

4%
_ Ann

•
	a	1
ARB


w
Error
M
SO
ov



•


• • • •
0%
•

•
~
•
w
•


(
-2%
) 10 £0, • 30 40

•
•


-4%





Test Cases

Figure 4-14 Comparisons between GEM with Auto-shift and Powertrain Dyno Tests
In 55 and 65 cruise speed cases, a road grade was included, and therefore, shifting in
cruise speed cycle can happen depending on the combination of axle ratio, vehicle loads and
many other conditions. As can be seen from this figure, cruise speed cycles are generally in
good agreement with testing data in the range of 0-3 percent difference. However, the
comparisons in the transient ARB cycle are not as good as cruise speed cycles. The main
reasons can be attributed to the thermal management as well as to the different shifting strategy
used in GEM. This happens for many of the ARB tests points. Since there is no mechanism in
GEM with steady state fuel map to account for such transient behaviors, it is no surprise that
GEM would not be able to predict the cycle accurately. That is one of the key reasons why the
agencies introduce a new cycle average approach to replace the GEM simulation with steady
state map, thus much more accurately accounting for the transient behavior. 24 More on this level
of validations can be seen from the SwRI final report.22
4.3.2.3 Validations against Real-world Driving Route
The third part of validations is extremely challenging when GEM is used to simulate real-
world driving routes. The most challenging parts of this validation are difficulty in obtaining the
exact the same engine, transmission and vehicle input information for GEM, and measurement of
the engine torque, the road grade, transmission gear number, and fueling over the entire road.
Furthermore, the auto-shifting or the agency default's shifting is used for validations. Because of
these challenges, it is impossible to compare GEM with absolute measurements of fueling. Only
relative comparisons are meaningful. Detailed description of this part of the validations,
including detailed routes, road grade, and complete vehicle specifications, can be seen from
SwRI final report22, and therefore only a brief summary is described here.
¦ The test procedure follows SAE J1321 "Type II." One truck is used as the controlled
baseline, while the second one contains advanced technology packages, such as trailer
aero treatment and lower rolling resistance tires. Difference of these two trucks in

-------
fueling by driving a pre-defined route is used to GEM validation. The route used for this
purpose is part of Texas State Highway 130 from Seguin to Austin. The total distance
used for GEM modeling is about 180 miles. Three repeat runs were carried out. The
instruments used for these tests are listed below: 5 kN-m driveshaft torque meters
¦	High resolution GPS (10 meter resolution)
¦	CAN data from vehicle
-	Engine torque
-	Fuel flow
-	Engine RPM
-	Gear number
-	Vehicle speed
-	Regeneration status
-	Thermal management status
-	Accelerator pedal position
-	Brake pedal position
-	Air conditioning compressor on/off
-	Cruise control on/off
Table 4-3 shows one of three runs of the comparisons between GEM and testing results.
The column "T393" is the control vehicle defined as the baseline following SAE J1321 "Type
II," while Column "T394" is the truck with advanced aero and tire on the second truck. The last
column is the difference between these two trucks. Comparisons should focus on two rows of
this table, which are GEM overall MPG and Actual overall MPG. Just as expected, the absolute
comparisons between 7.41 and 7.63 MPG for the control vehicle and 8.08 and 7.85 MPG for the
test vehicle are not too impressive, which is 7.7 percent difference shown in the row of Overall
MPG% Difference. The reason is that it is virtually impossible to model the vehicle exactly
because of the alignment of gear shifting, road grade and environmental conditions. In addition,
during the simulations, it was found that the engine torque measured from the drive shaft torque
meter is up to 200 NM higher than the peak torque curve of the engine fuel map used for GEM,
suggesting that the engines in truck and the engine used in GEM are not exactly same.
Furthermore, GEM didn't use exact shifting table. Rather, the default shifting is used. However,
the relative comparisons is in a good agreement between GEM and simulation (9.7 versus 10.2).

-------
Table 4-3 GEM vs On-road Fuel Consumption Results

Control
Truck
Test
Truck
with
Aero
Percent Change from
Control due to Aero
Package
GEM
Predictions
Segment 1 Miles
19.89
19.85

Segment 1 Gal
2.74
2.49

Segment 1 MPG
7.27
7.97
9.7%
Segment 2 Miles
56.65
56.56

Segment 2 Gal
7.36
6.69

Segment 2 MPG
7.69
8.45
9.9%
Segment 3 Miles
29.81
29.77

Segment 3 Gal
4.05
3.70

Segment 3 MPG
7.35
8.05
9.5%
Total Gals
14.15
12.88

Total Miles
106.35
106.18

GEM Overall MPG
7.51
8.25
9.7%
GEM Total Fuel lbs
99.05
90.11
-9.0%
On-Road
Results
(Run 3)
Actual Total Fuel lbs
97.60
88.40
-9.4%
Actual Gals
13.95
12.63

Actual Distance (miles)
106.37
106.21

Actual Overall MPG
7.63
8.41
10.2%
Again, more comprehensive discussions on this part can be seen the final report.22
4.4 EPA and NHTSA HD Vehicle Compliance Model
As described earlier, GEM is a computer model that simulates vehicle operation to
predict CO2 emissions and fuel consumption for a wide variety of heavy-duty vehicles. This
section describes how that computer model is used as a compliance tool to evaluate vehicle
performance relative to the applicable standards. The engineering evaluation of GEM discussed
in Chapter 4.3 was not limited by computing time and presumes all inputs to be accurate.
However, using GEM as a compliance tool requires some simplification of the model. It also
requires the elimination of user inputs that cannot be verified by the agencies.
The Phase 2 GEM is EPA and NHTSA's vehicle compliance simulation model and is
similar to Phase 1 GEM in many respects. However, it differs from the Phase 1 version in two
major aspects. The first involves the significant improvements described in Chapter of 4.2.1.
Second, Phase 2 GEM provides users the opportunity to enter additional vehicle and engine

-------
parameters for the actual vehicle being simulated. As noted above, Phase 1 GEM only allows a
maximum of five user defined inputs for tractors. These are the aerodynamic drag coefficient,
tire rolling resistance, vehicle speed limiter, weight reduction, and idle reduction. For vocational
vehicles there is only one user defined input: tire rolling resistance. In contrast, the Phase 2
GEM allows the user to input many more engine and vehicle parameters, including most of those
that have the greatest impact on emissions. In particular, it allows vehicle manufacturers to input
their own engine fuel maps. Key driveline parameters, such as transmission gear number versus
gear ratio, axle ratio, and tire rolling radius, are also part of the manufacturer inputs.
There are still some GEM input parameters that are pre-defined by the agencies. For
some, such as shifting strategy, this is due to the fact that the parameters are hard to measure and
quantify due to the lack of well-defined test procedures. For others, the manufacturers consider
the parameter values to be proprietary and are reluctant to share the information with other
parties. An example is the transmission gear shifting strategy table. The modeling parameters
associated with torque converters for automatic transmission are also pre-defined by the
agencies. The inertias of all rotational parts, vehicle weights and accessory power losses are also
default parameters defined by the agencies. Finally, in order to have a consistent basis for the
standards, the vehicle weights and payloads are predefined by vehicle class and duty cycle. In
order to understand natures of those inputs, Table 4-4 lists three types of inputs used by GEM.

-------
Table 4-4 GEM Inputs and Technology Improvement
GEM INPUTS
OEM INPUT
EPA DEFAULT
REQUIRED
OEM INPUT
No test required
with optional OEM
input based on a
test
Based on a test
procedure
Engine Fuel map


Yes
Transmission loss map

Yes (40 CFR
1037.565)

Axle power loss map

Yes (40 CFR
1037.560)

Drive axle configuration
6x2, 6x4, 6x4D, 4x2


Axle ratio
Input value


CdA
Vocational: Input
value based on
directions in 40 CFR
1037.520

Tractors: Yes (40 CFR
1037.525)
Crr


Yes (40 CFR
1037.520)
Tire size (revs/mile)


Yes (40 CFR
1037.520)
Transmission Type
MT, AMT, AT,


Weight adjustment
Input value based on
directions in 40 CFR
1037.520


Vehicle Speed Limiter
(Tractor)
Input value based on
directions in 40 CFR
1037.520


Predictive Cruise Control (%)
Input value based on
directions in 40 CFR
1037.520


Accessory Load (%)
Input value based on
directions in 40 CFR
1037.520


Extended Idle Reduction (%)
Input value based on
directions in 40 CFR
1037.520


Tire Pressure System (%)
Input value based on
directions in 40 CFR
1037.520


Neutral Coast
Input value based on
directions in 40 CFR
1037.520


Neutral-Idle


Yes (40 CFR
1036.535(d))
Delta PTO (Vocational)

Yes, default is zero
(40 CFR 1037.540)


-------
Automatic Engine Shutdown
(Vocational)


Yes (40 CFR
1036.535(d))
Start-Stop (Vocational)


Yes (40 CFR
1036.535(d))
Table 4-8 and Table 4-9 list all of the GEM input parameters for tractors and Table 4-13
through Table 4-15 list the predefined parameters for vocational vehicles. These tables also
include weighting factors for each driving cycle for the determination of composite CO2 in
g/ton-mile.
It is important to note that, for many of these parameters, publicly available information
on the values for current and future vehicles is limited. Manufacturers have provided values to
the agencies, but have generally identified them as confidential business information.
Nevertheless, we have used this information to inform our estimation of appropriate default
values.
4.4.1 GEM Input Values
4.4.1.1 Transmissions
One of the major changes in Phase 2 GEM is to allow manufacturers to enter their
transmission gear ratio versus gear number. When entering this information, manufacturers also
have an option to select the type of transmission, which is either manual, automated manual or
automatic with a torque converter. Power loss associated with pumping and spin loss can either
use the default values or use the one created by manufacturers.
One of the areas that required significant development work was the transmission shift
strategy for use in the compliance tool. This was required because (as just noted) transmission
suppliers have been reluctant to provide their shifting strategies to vehicle manufacturers for
vehicle certification due to their concern over protecting intellectual property. The shifting
strategy in the Phase 2 GEM includes the agencies' internally developed automatic shift
algorithm.15 The impact of the use of the agencies' default transmission shifting as opposed to
using manufacturers' shifting strategies has been evaluated and the results are presented in
Figure 4-14, shown above in Chapter 4.3.2.2. Phase 2 GEM includes three types of transmissions
as discussed in Chapter 4.2.2.3.3. They are manual transmissions (MT), automated manual
transmissions (AMT) and automatic transmissions (AT). Due to lack of test data for other types
of transmissions, GEM was not able to be validated against these three cases:
1.	Dual clutch transmission (DCT)
2.	Dual clutch transmission with a torque converter
3.	Allison TC-10 automatic transmission
The agencies will allow use of AMT to model case 1; use of AT to model case 2, and use
of AT to model case 3. The manufacturers will still have the option to either use powertrain
dynamometer tests to quantify the benefits of these or any other special transmissions, rather
than use the GEM values. The detailed test procedure of the powertrain dynamometer tests are
described in 40 CFR 1037.550. Alternatively, the manufacturers can conduct their own

-------
transmission tests on individual gears to get power loss table to replace the default ones used in
GEM. The detailed test procedure on power loss measurement can be seen in 40 CFR 1037.565.
4.4.1.2	Axles
Axle ratios for all model sub-categories will be user defined. If users do not provide a
power loss table, a power loss table generated in GEM will be used. 40 CFR 1037 covers the
specific procedures to generate data for axle input file, including the axle power loss input. Four
types of axles are uniquely modeled in GEM. They are 6x2, 6x4, 4x2 and 6x4D. 6x4D stands
for an axle that can disengage one of the drive axles when certain conditions are met. If 6x4D is
selected, GEM will model the axle as a 6x2 for the 55 and 65mph cruise cycles, and model the
axle as a 6x4 for the transient cycle. However, only one drive axle ratio can be selected. The
user must select the drive axle ratio that is expected to be used for the greatest driving distance.
All other axles, such as 8x4, 8x6, 10x4, 10x6, 12x4, 14x4 or any other "non-conventional" axle
configurations will not be modeled by GEM, rather they will be modeled as 6x4 axles. In
addition to allowing manufactures to input axle losses into GEM, the default losses were also
updated based on CBI from two major axle suppliers. Instead of the default efficiency taking the
form of a fixed efficiency, the losses are modeled as power losses as function of wheel speed and
wheel torque.
4.4.1.3	Weights
It is assumed that the vehicle unloaded weight will vary by vehicle subcategory. Taking
tractors as an example, the total weight ranges from 65,500 to 70,500 lbs, while for Class 7
tractors weight ranges from 46,500 to 50,000 lbs. The payload capacity varies as shown in Table
4-7 through Table 4-15. The development of these weights is discussed in Chapter 3 of the RIA.
4.4.1.4	Inertia
All of the inertias for rotational parts, including engine, transmission and axle, are pre-
defined based on a combination of the agencies' engineering judgment and confidential business
information from OEMs. The default inertia values were used during GEM validation against
respective trucks and they will be used as the default values for all of the vehicles certified using
GEM. Thus, the vehicle OEM will not have flexibility to enter their own inertias.
4.4.1.5	Accessory Load
Based on additional data from manufacturers the agencies are finalizing different
accessory loads from what was proposed. The breakdown in electrical and mechanical load are
shown in Table 4-7 through Table 4-15, and are the default values used in vehicle certification.
The change increased accessory power in all sectors, which increased CO2 mass per ton-mile as
shown in Table 4-5 and Table 4-6.

-------
Table 4-5 Change in CO2 Mass per Ton-Mile from Change in Accessory Load - Tractors
REGULATORY CLASS
% Change in
CO2 (g/ton-mi)
CLASS 8 COMBINATION
Sleeper Cab - High Roof
1.2%
Sleeper Cab - Mid Roof
1.2%
Sleeper Cab - Low Roof
1.2%
Day Cab - High Roof
1.4%
Day Cab - Mid Roof
1.4%
Day Cab - Low Roof
1.5%
CLASS 7 COMBINATION
Day Cab - High Roof
1.7%
Day Cab - Mid Roof
1.7%
Day Cab - Low Roof
1.8%
HEAVY-HAUL
COMBINATION
All Cabs - All Roofs
1.2%
Table 4-6 Change in CO2 Mass per Ton-Mile from Change in Accessory Load - Vocational
REGULATORY CLASS
% Change in CO2
(g/ton-mi)
HHD
Regional Duty Cycle
2.1%
Multi-Purpose Duty Cycle
3.6%
Urban Duty Cycle
4.0%
MHD
Regional Duty Cycle
1.7%
Multi-Purpose Duty Cycle
3.1%
Urban Duty Cycle
3.8%
LHD
Regional Duty Cycle
0.4%
Multi-Purpose Duty Cycle
0.7%
Urban Duty Cycle
0.8%
MHD-SI
Regional Duty Cycle
1.5%
Multi-Purpose Duty Cycle
2.6%
Urban Duty Cycle
3.0%
LHD-SI
Regional Duty Cycle
0.4%
Multi-Purpose Duty Cycle
0.6%
Urban Duty Cycle
0.7%
4.4.1.6 Tires
The tire revolutions per mile value is a user defined input; however, the agencies do
provide default values for custom vocational sub-categories. Static loaded tire radius is used in

-------
GEM for all simulations for every combination tractor and the default value can be overridden
by the vehicle OEM.
Steer and drive axle tire coefficient of rolling resistance (Crr) values are provided by the
user. On tractors the trailer tire Crr assumes a constant value for all trailer tires. This value was
developed through tire testing performed by the SmartWay Transport Partnership.23
4.4.1.7	Idle Cycles and Modeling
GEM will model two additional idle-only cycles to determine both fuel consumption for
use in the CO2 emission calculation in 40 CFR 1037.510(b) when a vocational vehicle is idling,
and to recognize technologies that either reduce the fuel consumption rate or shut the engine off
(and restart) during short-term idle events or parked during the workday based on user inputs of
which idle technologies are selected. GEM will determine a parked idle fuel rate and a driving
idle fuel rate.
The parked idle fuel rate will use a 4-point idle fuel map to calculate CO2 emissions and
fuel consumption at parked idle conditions. If automatic engine shutdown is selected the
calculated parked idle fuel rate will be reduced by 80 percent. The parked idle cycle is
applicable for all weight classes of vocational vehicles (HHD, MHD and LHD, custom chassis)
using the Regional, Multi-Purpose or Urban composite duty cycles.
For idle fueling during driving, the fueling is determine through the steady state engine
fuel map at minimum and maximum idle speed and 0 Nm and 100N. , and then GEM will
calculate reduced CO2 and fueling for automatic transmission that feature neutral idle
technology. Drive idle fuel consumption will be further reduced on vehicles with stop-start
systems, based on an assumption that the effectiveness would represent a 90 percent reduction
over this cycle. The drive idle cycle is applicable for all weight classes of vocational vehicles
(HHD, MHD and LHD and custom chassis) using the Multi-Purpose or Urban composite duty
cycles. The composite weighting factor of the drive idle cycle is zero for Regional vehicles.
More information can be seen in Chapter 3.4.2.3 about the idle cycle. Chapter 4.5 discusses how
these idle technologies are modeled as part of technology improvements that are recognized in
GEM.
4.4.1.8	Cycle Average Test Procedure for Transient ARB
As described in Chapter 2, fuel consumption during transient engine operation typically
is higher than during steady-state operation. The difference can vary significantly, but the trend
is generally consistent. If the GEM simulation relies on steady-state fuel maps to predict
emissions for all the cycles, including the transient cycle, the large error can be expected. The
most significant difference between steady state and transient behavior is the smoke control
during acceleration. Diesel engines must limit the fueling in order to prevent smoke during rapid
acceleration. In contrast, there is no such issue during steady state mapping. Furthermore, all
modern diesel engines use a fueling cut-off technique to shut off the fueling during deceleration
based on a manufacturer-defined set of conditions. Thermal management is another major factor
to that may create a difference between steady state and transient fueling. The aftertreatment
system is very sensitive to exhaust temperature in order to maintain optimal performance of

-------
selective catalytic reduction devices. Post fueling must be injected into the exhaust stream once
the exhaust temperature is below certain criteria, typically in the range of 200 degrees Celsius.
In steady state fueling mapping, the engine always follows the defined testing procedure; thus,
the engine runs hotter even at light loads than in a transient condition at the same speed and
torque; thus, thermal management may not even kick in. Because of these differences, engine
manufacturers typically include at least two distinguished engine calibrations into the engine
control unit - one for typically on-highway operation, and the other for transient operation, such
as urban and mountainous areas.
In view of these technical challenges, and in response to public comments (including
those from a leading manufacturer of diesel engines), the agencies have adopted a test procedure
called cycle average test procedure to account for this transient behavior (40 CFR 1036.540).
The detailed analyses can be also seen in references 24 and 25 as well. Since these two notable
publications, significant progress has been made, which covers a large number of confirmatory
engine dynamometer tests. A wide range of industrial supporting activities, and significant
refinement of numerical schemes for interpreting cycle average engine fuel map are also
conducted. The engine dynamometer tests include Cummins' medium duty ISB engine,
Navistar's heavy duty N13 engine, Volvo's heavy duty D13 engine, and Cummins' heavy-duty
ISX engine. All testing results indicate that the new test procedure would work well for the
transient ARB cycle. As for the cruise-cycles the procedure does generally work well especially
with some recent improvements to the generic vehicle definitions. Therefore, we will optionally
allow certification to be done with cycle average test procedure for these cruise speed cycles,
primarily based on the following reasons. The first reason is that it will allow engine
manufactures to provide engine fuel maps that don't reveal CBI. The second reason is that
allowing the cycle average procedure for cruise cycles opens up the possibility for hybrids that
are not integrated with a transmission to use the cycle average procedure instead of just the
powertrain procedure. By allowing these hybrids to use the cycle average procedure we reduce
the testing burned/cost for these systems. In order to ensure that manufacture using the cycle
average procedure for cruise speed cycle can produce representative results, the test procedures
are written in such a way that it allows EPA to use the steady-state mapping procedure during a
confirmatory test for the cruise cycles.
For all 55 and 65mph cruise speed cycles, a simplified engine fuel map will be used in
GEM by following the test procedure 40 CFR 1036.535, where only 80-90 testing points are
required for the engine fuel mapping.
4.4.1.9 Tractor Tables
Table 4-7 through Table 4-12 display the predefined GEM parameters for the Phase 2
tractor compliance model. The predefined parameters were developed using the same
methodology used in Phase 1.

-------
Table 4-7 Class 8 Combination Tractor Sleeper Cab Predefined Modeling Parameters
REGULATORY CLASS
CLASS 8
COMBINATION
CLASS 8
COMBINATION
CLASS 8
COMBINATION
Sleeper Cab - High
Roof
Sleeper Cab - Mid
Roof
Sleeper Cab - Low Roof
Total weight (kg)
31978
30277
30390
Number of Axles
5
5
5
Default Axle Configuration
6x4
6x4
6x4
Electrical Accessory Power (W)
1200
1200
1200
Mechanical Accessory Power (W)
2300
2300
2300
Environmental Air Temperature (°C)
25
25
25
Payload (tons)
19
19
19
Weight Reduction (lbs)
Add 1/3* weight
reduction to Payload
tons
Add 1/3* weight
reduction to Payload
tons
Add 1/3* weight
reduction to Payload
tons
Tire Crr
=0.425 *Trailer
Crr+0.425*Drive
Crr+0.15 * Steer Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
Drive Cycles & Weightings:



CARB HHDDT
0.05
0.05
0.05
GEM 55 mph
0.09
0.09
0.09
GEM 65 mph
0.86
0.86
0.86
CdA value modeled in GEM is equal to the OEM input minus 0.3 m2 to account for improved trailer aerodynamics
beginning in 2027 MY for high roof tractors. See Section III.E.2.a of the Preamble.
Table 4-8 Class 8 Combination Tractor Day Cab Predefined Modeling Parameters
REGULATORY SUBCATEGORY
CLASS 8
COMBINATION
CLASS 8
COMBINATION
CLASS 8
COMBINATION
Day Cab - High Roof
Day Cab - Mid Roof
Day Cab - Low Roof
Total weight (kg)
31297
29529
29710
Number of Axles
5
5
5
Default Axle Configuration
6x4
6x4
6x4
Electrical Accessory Power (W)
1200
1200
1200
Mechanical Accessory Power (W)
2300
2300
2300
Environmental air temperature (°C)
25
25
25
Payload (tons)
19
19
19
Weight Reduction (lbs)
Add 1/3* weight
reduction to Payload
tons
Add 1/3* weight
reduction to Payload
tons
Add 1/3* weight
reduction to Payload
tons
Tire Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
Drive Cycles & Weightings:



CARB HHDDT
0.19
0.19
0.19
GEM 55 mph
0.17
0.17
0.17
GEM 65 mph
0.64
0.64
0.64
CdA value modeled in GEM is equal to the OEM input minus 0.3 m2 to account for improved trailer aerodynamics
beginning in 2027 MY for high roof tractors. See Section III.E.2.a of the Preamble.

-------
Table 4-9 Class 7 Combination Tractor Predefined Modeling Parameters
REGULATORY SUBCATEGORY
CLASS 7
COMBINATION
CLASS 7
COMBINATION
CLASS 7
COMBINATION
Day Cab - High Roof
Day Cab - Mid Roof
Day Cab - Low Roof
Total weight (kg)
22679
20910
21091
Axle Base
4
4
4
Default Axle Configuration
4x2
4x2
4x2
Electrical Accessory Power (W)
1200
1200
1200
Mechanical Accessory Power (W)
2300
2300
2300
Environmental air temperature (°C)
25
25
25
Payload (tons)
12.5
12.5
12.5
Weight Reduction (lbs)
Add 1/3* weight
reduction to Payload
tons
Add 1/3* weight
reduction to Payload
tons
Add 1/3* weight
reduction to Payload
tons
Tire Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
Drive Cycles & Weightings:



CARB HHDDT
0.19
0.19
0.19
GEM 55 mph
0.17
0.17
0.17
GEM 65 mph
0.64
0.64
0.64
CdA value modeled in GEM is equal to the OEM input minus 0.3 m2 to account for improved trailer aerodynamics
beginning in 2027 MY for high roof tractors. See Section III.E.2.a of the Preamble.
Table 4-10 Heavy-Haul Tractor Predefined Modeling Parameters3
REGULATORY SUBCATEGORY
HEAVY-HAUL COMBINATION
All Cabs - All Roofs
Total weight (kg)
53750
Number of Axles
5
Default Axle Configuration
6x4
Electrical Accessory Power (W)
1200
Mechanical Accessory Power (W)
2300
Environmental air temperature (°C)
25
Payload (tons)
43
Weight Reduction (lbs)
Add 1/3*weight reduction to
Payload tons
Tire Crr
=0.425 *Trailer Crr+0.425 *Drive
Crr+0.15 *Steer Crr
Drive Cycles & Weightings:

CARB HHDDT
0.19
GEM 55 mph
0.17
GEM 65 mph
0.64
Note:
a See 40 CFR 1037.106

-------
Table 4-11 Optional Heavy Class 8 Combination Tractor Sleeper Cab Predefined Modeling Parameters3
REGULATORY CLASS
OPTIONAL HEAVY
CLASS 8
COMBINATION
OPTIONAL HEAVY
CLASS 8
COMBINATION
OPTIONAL HEAVY
CLASS 8
COMBINATION
Sleeper Cab - High
Roof
Sleeper Cab - Mid
Roof
Sleeper Cab - Low Roof
Total weight (kg)
53750
52049
52162
Number of Axles
5
5
5
Default Axle Configuration
6x4
6x4
6x4
Electrical Accessory Power (W)
1200
1200
1200
Mechanical Accessory Power (W)
2300
2300
2300
Environmental Air Temperature (°C)
25
25
25
Payload (tons)
43
43
43
Weight Reduction (lbs)
Add 1/3* weight
reduction to Payload
tons
Add 1/3* weight
reduction to Payload
tons
Add 1/3* weight
reduction to Payload
tons
Tire Crr
=0.425 *Trailer
Crr+0.425*Drive
Crr+0.15 * Steer Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
Drive Cycles & Weightings:



CARB HHDDT
0.05
0.05
0.05
GEM 55 mph
0.09
0.09
0.09
GEM 65 mph
0.86
0.86
0.86
CdA value modeled in GEM is equal to the OEM input minus 0.3 m2 to account for improved trailer aerodynamics.
See Section III.E.2.a of the Preamble.
Note:
a See 40 CFR 1037.670

-------
Table 4-12 Optional Heavy Class 8 Combination Tractor Day Cab Predefined Modeling Parameters3
REGULATORY SUBCATEGORY
OPTIONAL HEAVY
CLASS 8
COMBINATION
OPTIONAL HEAVY
CLASS 8
COMBINATION
OPTIONAL HEAVY
CLASS 8
COMBINATION
Day Cab - High Roof
Day Cab - Mid Roof
Day Cab - Low Roof
Total weight (kg)
53069
51301
51482
Number of Axles
5
5
5
Default Axle Configuration
6x4
6x4
6x4
Electrical Accessory Power (W)
1200
1200
1200
Mechanical Accessory Power (W)
2300
2300
2300
Environmental air temperature (°C)
25
25
25
Payload (tons)
43
43
43
Weight Reduction (lbs)
Add 1/3* weight
reduction to Payload
tons
Add 1/3* weight
reduction to Payload
tons
Add 1/3* weight
reduction to Payload
tons
Tire Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
=0.425*Trailer
Crr+0.425*Drive
Crr+0.15 *Steer Crr
Drive Cycles & Weightings:



CARB HHDDT
0.19
0.19
0.19
GEM 55 mph
0.17
0.17
0.17
GEM 65 mph
0.64
0.64
0.64
CdA value modeled in GEM is equal to the OEM input minus 0.3 m2 to account for improved trailer aerodynamics
beginning in 2027 MY for high roof tractors. See Section III.E.2.a of the Preamble.
Note:
a See 40 CFR 1037.670
4.4.1.10 Vocational Tables
Table 4-13 through Table 4-15 display the predefined GEM parameters for use for the
vocational vehicle compliance model. The optional custom chassis configurations use these
parameters as well. Many of the parameters are based on the vehicles EPA selected to test at
SwRI and are considered to reasonably represent the fleet in their respective categories. For
example, the Kenworth T270 truck and Ford F-650 tow truck are used as vehicles to represent
the MHD and LHD vocational vehicle fleet, while the Kenworth T700 and Autocar refuse trucks
are used to represent the fleet of HHD vocational vehicles. With those vehicles as reference, it
helps to determine the type of transmission and its gear ratio, tire diameters, and all accessory
losses used for all vocational vehicles shown in these three tables. Tire radius and axle ratios
were selected, using good engineering judgment and stakeholder input, to reflect reasonable final
drive ratios to match with our modeled transmissions. With the exception of the HHD vehicles,
the engine power rating is the same as in Phase 1. For the HHD subcategories, the agencies
selected a mix of 15L 455-hp and 11L-350 hp engines because this is a more typical power
rating for vehicles that are not long haul. The baseline engines are described in the RIA Chapter
2.9.1. Other parameters, such as the vehicle weight, payload, weight reduction, tire rolling
resistance, frontal area, and axle base, etc. are defined in the RIA Chapter 2.9.2. The gear
mechanical efficiency as well as axle mechanical efficiency is selected based on the inputs from
stakeholders. The weighting of steer tire Crr and drive tire Crr is different than in Phase 1 to
better reflect the weight distribution over the steer and drive axles. The assignment of 50 percent

-------
of reduced weight back to payload is not the same as for tractors. See the RIA Chapter 2.9 for
details. Chapter 4.4.3 to 4.4.9 explain how these parameters are used in GEM.
The agencies are expanding the number of vocational subcategories from three (in Phase
1) to nine (in Phase 2). It can be seen from Table 4-13 through Table 4-15, the agencies will also
add two idle cycles for vocational vehicles to the duty cycles used in Phase 1 certification.
Table 4-13 Vocational HHD Vehicle Predefined Modeling Parameters
REGULATORY SUBCATEGORY
HHD
HHD
HHD
Regional Duty Cycle
Multi-Purpose Duty
Cycle
Urban Duty Cycle
Total weight (kg)
19051
19051
19051
Number of Axles
3
3
3
Electrical Accessory Power (W)
1200
1200
1200
Mechanical Accessory Power (W)
2300
2300
2300
Environmental Air Temperature (°C)
25
25
25
CdA (m2)
6.86
6.86
6.86
Tire Crr
=0.7*Drive Crr +
0.3*SteerCrr
0.7*Drive Crr +
0.3*SteerCrr
=0.7*Drive Crr +
0.3*SteerCrr
Payload (tons)
7.50
7.50
7.50
Weight Reduction (lbs)
Add 0.5* weight
reduction to Payload
tons
Add 0.5* weight
reduction to Payload
tons
Add 0.5* weight
reduction to Payload
tons
Drive Cycles & Weightings:



CARB HHDDT
0.20
0.54
0.90
GEM 55 mph
0.24
0.23
0.10
GEM 65 mph
0.56
0.23
0.00
Drive Idle cycle
0.00
0.17
0.15
Parked Idle Cycle
0.25
0.25
0.25

-------
Table 4-14 Vocational MHD Vehicle Predefined Modeling Parameters
REGULATORY SUBCATEGORY
MHD
MHD
MHD
Regional Duty Cycle
Multi-Purpose Duty
Cycle
Urban Duty Cycle
Total weight (kg)
11408
11408
11408
Number of Axles
2
2
2
Electrical Accessory Power (W)
900
900
900
Mechanical Accessory Power (W)
1600
1600
1600
Environmental Air Temperature (°C)
25
25
25
CdA (m2)
5.40
5.40
5.40
Tire Crr
=0.7*Drive Crr +
0.3*SteerCrr
0.7*Drive Crr+
0.3*SteerCrr
=0.7*Drive Crr +
0.3*SteerCrr
Payload (tons)
5.60
5.60
5.60
Weight Reduction (lbs)
Add 0.5* weight
reduction to Payload
tons
Add 0.5* weight
reduction to Payload
tons
Add 0.5* weight
reduction to Payload
tons
Drive Cycles & Weightings:



CARB HHDDT
0.20
0.54
0.92
GEM 55 mph
0.24
0.29
0.08
GEM 65 mph
0.56
0.17
0.00
Drive Idle cycle
0.00
0.17
0.15
Parked Idle cycle
0.25
0.25
0.25
Table 4-15 Vocational LHD Vehicle Predefined Modeling Parameters
REGULATORY SUBCATEGORY
LHD
LHD
LHD
Regional Duty Cycle
Multi-Purpose Duty
Cycle
Urban Duty Cycle
Total weight (kg)
7257
7257
7257
Number of Axles
2
2
2
Electrical Accessory Power (W)
500
500
500
Mechanical Accessory Power (W)
1000
1000
1000
Environmental Air Temperature (°C)
25
25
25
CdA (m2)
3.40
3.40
3.40
Tire Crr
=0.7*Drive Crr +
0.3*SteerCrr
0.7*Drive Crr+
0.3*SteerCrr
=0.7*Drive Crr +
0.3*SteerCrr
Payload (tons)
2.85
2.85
2.85
Weight Reduction (lbs)
Add 0.5* weight
reduction to Payload
tons
Add 0.5* weight
reduction to Payload
tons
Add 0.5* weight
reduction to Payload
tons
Drive Cycles & Weightings:



CARB HHDDT
0.20
0.54
0.92
GEM 55 mph
0.24
0.29
0.08
GEM 65 mph
0.56
0.17
0.00
Drive Idle cycle
0.00
0.17
0.15
Parked Idle cycle
0.25
0.25
0.25

-------
4.4.1.11 Trailer Tables
The agencies are adopting an equation-based compliance approach for box van
manufacturers and they are not required to certify their trailers using GEM. However, the
equations for each box van subcategory are based on the simulated trailers described in this
section. The same four input parameters that would be applied in GEM for trailers are also
applied in the GEM-based compliance equations. The following description of the GEM trailer
model as it applies to box vans is included for informational purposes only. Note that non-box
trailers do not use GEM or the GEM-based equation for compliance and a discussion of non-box
trailers is not included here.
Trailers are simulated using the same GEM models as the tractor program. There are
only minor differences between the trailer and tractor modeling parameters and inputs. Table
4-16 lists all of the predefined vehicle parameters of trailer baseline models. The predefined
modeling parameters for the long box dry van subcategory are identical to the Class 8 high-roof
sleeper cab tractor subcategory. The other trailer subcategories differ in tractor cab type, total
weight, aerodynamic characteristics, number of axles, payload, and drive cycle. For example,
the refrigerated vans include a refrigeration unit which adds weight. Short box vans are half the
length, have a single axle, and are pulled by a day cab tractor which reduces total weight and the
total payload carrying capacity. The drive cycle weightings are consistent with the tractor
program. Long box vans are simulated as being pulled by sleeper cabs, and therefore have the
long-haul drive cycle weightings. The short box trailers are pulled by Class 7 day cabs and have
the short-haul weightings.
Similar to the tractor program, trailer manufacturers can provide aerodynamic drag, tire
rolling resistance and weight reduction inputs to the model. The key differences between the
trailer and tractor options are that aerodynamic drag is submitted as a change in drag (delta CdA)
for trailers, which is compared to the baseline CdA values shown in Table 4-16 within GEM, and
only adjustments to trailer tire rolling resistance are allowed. A list of weight reduction options
is available in 40 CFR 1037.520 and manufacturers have the option to indicate that their trailers
use Automatic Tire Inflation Systems (ATIS) or tire pressure monitoring systems (TPMS) for a
predefined additional performance improvement. Additional information about each trailer
subcategory is found in Chapter 2.10 of this RIA. A description of the GEM-based equation
development is provided in Chapter 2.10.5.

-------
Table 4-16 Predefined Modeling Parameters for Box Trailers
REGULATORY
SUBCATEGORY
LONG BOX
DRY VAN
LONG BOX
REFRIGERATED
VAN
SHORT BOX
DRY VAN
SHORT BOX
REFRIGERATED
VAN
Tractor Type
C8 Sleeper Cab - High Roof
C7 Day Cab-High Roof
Engine Fuel Map
MY 2018 15L-455 HP
MY 2018 11L — 350 HP
Total weight (kg)
31978
33778
18306
20106
Baseline CdA Values (m2)
6.0
6.0
5.6
5.6
Total Number of Axles
5
3
Payload (tons)
19
10
Tractor Axle Configuration
6x4
4x2
Electrical Accessory Power (W)
300
Mechanical Accessory Power (W)
1000
Steer Tire RR
6.54
Drive Tire RR
6.92
Tire Radius (m)
0.5
Axle Drive Ratio
3.7
Tire Crr
=0.425*Trailer Crr+0.425*Drive Crr+0.15*Steer Crr
Weight Reduction (lbs)
Add l/3*weight reduction to Payload tons
Drive Cycles & Weightings:

CARB HHDDT
0.05
0.19
GEM 55 mph
0.09
0.17
GEM 65 mph
0.86
0.64
4.5 Technology Improvements that Are Recognized in GEM without
Simulation
The development of GEM as a compliance tool has required the agencies to balance the
need for simplicity against the rigor of the model. As part of that process, the agencies have
identified several technologies and technological improvements that would be difficult to
accurately simulate, but that should be recognized during certification. These are recognized in
the Phase 2 GEM through post-simulation adjustments to the results. This is similar to what was
done in Phase 1, where the GEM interface included pull-down menus for manufacturers to select
these adjustments. For this reason, these adjustments have come to be known as pull-down
technologies.
Phase 2 GEM will continue to recognize those technologies that would be difficult to
model accurately. In addition to those recognized in Phase 1, the technology list is expanded to a
much wider range as discussed in the next few paragraphs of this chapter. In contrast to Phase 1,
Phase 2 GEM uses a different approach in recognizing these technologies. Predefined
improvement values for each of these technologies, developed by the agencies after consulting
various stakeholders and searching for literature values, are defined in 40 CFR 1037.520. The
user is required to enter the predefined improvement value into the GEM input file in the
corresponding technology column.

-------
For some of these technologies, such as predictive cruise control and Automatic Tire
Inflation Systems (ATIS), the actual benefit is dependent on how operators behave in the real
world. For example, ATIS will be of very little benefit where a driver made sure on a daily basis
that the tires were properly inflated, but will have large benefits where a driver never checked the
tires. For other technologies, the benefits of the technology are small relative to the difficulty of
rigorously simulating it. The agencies believe the technology improvement approach is an
appropriate compromise that will achieve the regulatory goal of incentivizing the use of the
technology.
In this approach, the GEM software will adjust the simulation results to decrease the
g/ton-mile results that are output by the model. For example, with a technology that is assigned
a 1 percent benefit, the official result for a vehicle that was simulated as having 500 g/ton-mile
CO2 emissions will be reported as having an emission rate of 495 g/ton-mile.
The technology improvement values used for tractors are shown in Table 4-17. These
values represent the agencies' best judgment about the appropriate value for each of these
technologies and are discussed in more detail in RIA Chapter 2.4. We are generally assigning
minimum values to be conservative and not overestimate the actual in-use benefits. These values
were developed based on all available information, including information from stakeholders.
Table 4-17 Tractor Technology Improvement Values
TECHNOLOGY
IMPROVEMENT
CLASS 8
SLEEPER CABS
CLASS 8
DAY CABS
CLASS 7 DAY
CABS
CLASS 8 HEAVY
HAUL TRACTORS
Automated Manual, Automatic,
and Dual Clutch Transmissions
2%
2%
2%
2%
Predictive Cruise control
2%
2%
2%
2%
High Efficiency Air
Conditioning Compressor
0.5%
0.5%
0.5%
0.5%
Electric Accessories
1%
1%
1%
1%
Extended Idle Reduction
Values range
between 1 - 6 %
N/A
N/A
N/A
Automatic Tire Inflation
System (ATIS)
1.20%
1.20%
1.20%
1.20%
Tire Pressure Monitoring
System
1%
1%
1%
1%
Neutral Coast
1%
1%
1%
1%





Neutral Idle
Emissions during idle cycle calculated using torque and speed values from
fuel map with the transmission in drive and neutral, 10% and 90% of the
cycle time, respectively3
Note:
a See idle fuel consumption test procedure at 40 CFR 1036.535(d).

-------
For vocational vehicles, the technology improvement values in Table 4-18 are being
adopted.
Table 4-18 Vocational Vehicle Technology Improvement Values
TECHNOLOGY
IMPROVEMENT
REGIONAL DUTY
CYCLE
MULTI-PURPOSE
DUTY CYCLE
URBAN DUTY
CYCLE
PTO Delta Fuel (g/ton-mile)
Range 0 to 30; value obtained using separate test
Automatic Tire Inflation System
(ATIS)
1.2%
1.1%
1.1%
Tire Pressure Monitoring System
1.0%
0.9%
0.9%
Electric or high efficiency A/C
compressor3
0.5% for HHD, 1.0% for MHD and LHD
Electric Power Steering
0.5%
1.0%
1.0%
7-speed transmission for Custom
Chassis School & Coach buses
1.7%
N/A
0.9%
Neutral Idle for Custom Chassis
Range depending on the default engine. Input is Yes or No.
Stop-Start Idle Reduction for
Custom Chassis
Automatic Engine Shutdown for
Custom Chassis
Note:
a See instructions at 40 CFR 1037.520
For trailers, the following technologies in Table 4-19 will be considered.
Table 4-19 Trailer Technology Improvement Values
Technology Improvement
Effectiveness
Automatic Tire Inflation System
(ATIS)
1.2%
Tire Pressure Monitoring System
1.0%
If a manufacturer believes that the CO2 reduction benefits assigned by the agencies are an
underestimate, they have the option to perform powertrain testing or request (and demonstrate)
credit in the off-cycle technology process.

-------
References
1	National Academies of Science. "Technologies and Approaches to Reducing the Fuel Consumption of Medium-
and Heavy-Duty Vehicles." 2010. Recommendation 8-4.
2	Daimler Trucks North America Meeting Memo on GEM and Road grade, March 8, 2016, Greenhouse Gas
Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles, Phase 2, Proposed
Rule, Docket ID No: EPA-HQ-OAR-2014-0827 and NHTSA-2014-0132.
3	Allison Meeting Memo on GEM, June, 13 2016, Greenhouse Gas Emissions and Fuel Efficiency Standards for
Medium- and Heavy-Duty Engines and Vehicles, Phase 2, Proposed Rule, Docket ID No: EPA-HQ-OAR-2014-
0827 and NHTSA-2014-0132.
4	EMA Meeting Memo on GEM, February 2, 16 2016, Greenhouse Gas Emissions and Fuel Efficiency Standards for
Medium- and Heavy-Duty Engines and Vehicles, Phase 2, Proposed Rule, Docket ID No: EPA-HQ-OAR-2014-
0827 and NHTSA-2014-0132.
5	Ford Meeting Memo on GEM, June, 2 2016, Greenhouse Gas Emissions and Fuel Efficiency Standards for
Medium- and Heavy-Duty Engines and Vehicles, Phase 2, Proposed Rule, Docket ID No: EPA-HQ-OAR-2014-
0827 and NHTSA-2014-013280; Fed. Reg. 40137 (July 13, 2015).
6	Dana Meeting Memo on GEM, June, 16 2016, Greenhouse Gas Emissions and Fuel Efficiency Standards for
Medium- and Heavy-Duty Engines and Vehicles, Phase 2, Proposed Rule, Docket ID No: EPA-HQ-OAR-2014-
0827 and NHTSA-2014-0132; 80 Fed. Reg. 40137 (July 13, 2015).
7	See EPA's GEM web page at http://www3.epa.gov/otaq/climate/gem.htnx
8	Navistar, Greenhouse Gas Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and
Vehicles, Phase 2, Proposed Rule, Dockets ID No: EPA-HQ-OAR-2014-0827 and NHTSA-2014-0132;80 Fed.
Reg. 40137 (July 13, 2015).
9	Volvo Group, Greenhouse Gas Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines
and Vehicles, Phase 2, Proposed Rule, Dockets ID No: EPA-HQ-OAR-2014-0827 and NHTSA-2014-0132;80
Fed. Reg. 40137 (July 13, 2015).
10	Paccar, Inc., Greenhouse Gas Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines
and Vehicles; Phase 2; Proposed Rule, 80 Fed. Reg. 40138 (July 13, 2015); Docket I.D. No.: EPA-HQ-OAR-2014-
0827 and NHTSA-2014-0132; Fed. Reg. 40137 (July 13, 2015).
11	Allison, Inc., Greenhouse Gas Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines
and Vehicles; Phase 2; Proposed Rule, 80 Fed. Reg. 40138 (July 13, 2015); Docket I.D. No.: EPA-HQ-OAR-2014-
0827 and NHTSA-2014-0132; Fed. Reg. 40137 (July 13, 2015).
12	Eaton Corporation., Greenhouse Gas Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty
Engines and Vehicles; Phase 2; Proposed Rule, 80 Fed. Reg. 40138 (July 13, 2015); Docket I.D. No.: EPA-HQ-
OAR-2014-0827 and NHTSA-2014-0132; Fed. Reg. 40137 (July 13, 2015).
13	Daimler Trucks North America LLC, Detroit Diesel Corporation, And Mercedes-Benz USA, Greenhouse Gas
Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles, Phase 2, Proposed
Rule, Docket ID No: EPA-HQ-OAR-2014-0827 and NHTSA-2014-0132; Fed. Reg. 40137 (July 13, 2015).
14	"Peer Review of the Greenhouse gas Emissions Model (GEM) and EPA's Response to Comments," found in the
docket of this rulemaking, EPA-420-R-15-009.
15	K. Newman, J. Kargul, and D. Barba, "Development and Testing of an Automatic Transmission Shift Schedule
Algorithm for Vehicle Simulation, "SAEInt. J. Engines 8(3):2015, doi: 10.4271/2015-01-1142.
16Reinhart, T.E. (2015, June). Commercial Medium- and Heavy-Duty Truck Fuel Efficiency Technology Study -
Report #1. (Report No. DOT HS 812 146). Washington, DC: National Highway Traffic Safety Administration.
17	Matlab information (© 1994-2010 The MathWorks, Inc.) can be found at
http://www.mathworks.com/products/matlab. Simulink information (© 1994-2010 The MathWorks, Inc.) can be
found at http://www.mathworks.com/products/simulink. StateFlow information (© 1994-2010 The MathWorks,
Inc.) can be found at http://www.mathworks.com/products/stateflow..
18	US EPA, "Technical Research Workshop supporting EPA and NHTSA Phase 2 Standards for MD/HD
Greenhouse Gas and Fuel Efficiency — December 10 and 11, 2014," http://www3.epa.gov/otaq/climate/regs-heavy-
duty.htnx

-------
19	J. W. Anthony, J.V. Sarlashkar, A. Phan, C. E. Roberts, Jr., H. Zhang, J. Sanchez, M. Spears, A. Cullen,
Powertrain test cell and its test procedure development for greenhouse gas emission measurements, ASME Int. J.,
Vol. 4, Issue 3, 2015 (DOI: 10.1504/LJPT.2015.071729), 2015.
20	SwRI, "GEM Validation," Technical Research Workshop supporting EPA and NHTSA Phase 2 Standards for
MD/HD Greenhouse Gas and Fuel Efficiency — December 10 and 11, 2014,
http://www3.epa.gov/otaq/climate/regs-heavy-duty.htm.
21	J. Anthony, J. Sarlashkar, H. Zhang, J. Sanchez, M. Spears, A. Cullen, Chassis testing versus powertrain testing
for heavy-duty vehicle greenhouse gas emission measurements, SAE 2014 Commercial Vehicle Engineering
Congress, Oct. 7-9, 2014.
22	Michael Ross, Validation Testing for Phase 2 Greenhouse Gas Test Procedures and the Greenhouse Gas Emission
Model (GEM) for Medium and Heavy-Duty Engines and Powertrains, Final Report to EPA, Southwest Research
Institute, June, 2016 , found in docket of this rulemaking, Phase 2 (Docket ID No. EPA-HQ-OAR-2014-0827.
23	United States Environmental Protection Agency. SmartWay Transport Partnership July 2010 e-update accessed
July 16, 2010, from http://www3.epa.gov/smartwaylogistics/newsroom/documents/e-update-july-10.pdf.
24	H. Zhang, J, Sanchez, M, Spears, "Alternative Heavy-duty Engine Test Procedure for Full Vehicle Certification,",
SAE Int. J. Commer. Veh. 8(2): 2015, doi: 10.4271/2015-01-2768.
25	G. Salemme, E.D., D. Kieffer, M. Howenstein, M. Hunkler, and M. Narula, An Engine and Powertrain Mapping
Approach for Simulation of Vehicle C02 Emissions. SAE Int. J. Commer. Veh., October 2015. 8:: p. 440-450.

-------
Chapter 5: Impacts on Emissions and Fuel Consumption
5.1 Executive Summary
Climate change is widely viewed as the most significant long-term threat to the global
environment. According to the IPCC, it is extremely likely (>95 percent probability) that human
influence was the dominant cause of the observed warming since the mid-20th century. The
primary GHGs of concern are carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O),
hydrofluorocarbons (HFCs), perfluorocarbons, and sulfur hexafluoride.1 Mobile sources emitted
27 percent of all U.S. GHGs in 2013 when considering all upstream and downstream emissions,
and the transportation-related GHGs alone have grown 16 percent between 1990 and 2013.2
Mobile sources addressed in the recent endangerment finding under CAA section 202(a) -
highway vehicles including passenger cars, light-duty trucks, heavy-duty trucks, buses, and
motorcycles - accounted for almost 23 percent of all U.S. GHGs in 2013.3 Heavy-duty vehicles
emit CO2, methane, nitrous oxide, and hydrofluorocarbons and are responsible for almost 24
percent of all mobile source GHGs (over 6 percent of all U.S. GHGs) and about 28 percent of
CAA section 202(a) mobile source GHGs. For heavy-duty vehicles in 2013, CO2 emissions
represented roughly 96 percent of all GHG emissions (including HFCs).
This chapter provides the anticipated emissions impacts from the final standards. The
reductions in emissions are expected for carbon dioxide (CO2), methane (CH4), nitrous oxide
(N2O) and hydrofluorocarbons (HFCs). In addition to reducing the emissions of greenhouse
gases, this program will also affect the emissions of "criteria" air pollutants and their precursors,
including carbon monoxide (CO), fine particulate matter (PM2.5), oxides of sulfur (SOx), volatile
organic compounds (VOC) and oxides of nitrogen (NOx), and several air toxics, such as
benzene, 1,3-butadiene, formaldehyde, acetaldehyde, and acrolein.
The final standards will affect both diesel- and gasoline-fueled heavy-duty vehicles, as
well as those running on natural gases. The analyses account for both vehicle emissions
("downstream" emissions) and emissions from fuel production and distribution ("upstream"
emissions). The agencies conducted two analyses by employing DOT's CAFE model and EPA's
Motor Vehicle Emission Simulator (MOVES2014a)4, relative to different reference cases (i.e.,
different baselines). The agencies used EPA's MOVES model to estimate fuel consumption and
emissions impacts for tractor-trailers (including the engines which power the vehicle), and
vocational vehicles (including the engine which powers the vehicle). For heavy-duty pickups
and vans, the agencies performed separate analyses, which we refer to as "Method A" and
"Method B," to estimate fuel consumption and emissions from these vehicles. See Section 5.3
for additional details. The changes in upstream emissions result from decreased fuel
consumption. The emission factors from GREET5 were used to estimate the changes in
upstream emissions. In some cases, the GREET values were modified or updated by the
agencies to be consistent with the EPA's National Emission Inventory (NEI) and emission
factors from MOVES.
Table 5-1 through Table 5-3 summarize the impact of the program on GHG emissions
from the heavy-duty sector in calendar years 2025, 2040 and 2050, using Method A and B,
relative to two reference cases - flat (Alternative la) and dynamic (Alternative lb). Table 5-4

-------
through Table 5-6 summarize the projected fuel savings from the program in calendar years
2025, 2040 and 2050, using Method A and B, relative to the two reference cases.
Table 5-1 Annual Total GHG Emissions Impacts in Calendar Years 2025,2040 and 2050 - Final Program vs.
Alt lb using Analysis Method Aa

CY2025
CY2040
CY2050
MMT
CCheq
% Change
MMT C02eq
% Change
MMT C02eq
% Change
Downstream
-26.6
-4.9%
-103.3
-17.0%
-123.8
-18.0%
Upstream
-9.0
-4.9%
-35.5
-17.0%
-42.5
-19.0%
HFC
-0.1
-15.0%
-0.3
-13.0%
-0.3
-13.0%
Total
-35.7
-4.9%
-139.1
-17.0%
-166.6
-19.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 5-2 Annual Total GHG Emissions Impacts in Calendar Years 2025,2040 and 2050 - Final Program vs.
Alt la using Analysis Method Aa

CY2025
CY2040
CY2050
MMT
CC^eq
% Change
MMT
CCheq
% Change
MMT C02eq
% Change
Downstream
-28.9
-5.3%
-114.1
-19.0%
-136.9
-20.0%
Upstream
-9.8
-5.3%
-39.3
-19.0%
-47.2
-20.0%
HFC
-0.1
-15.0%
-0.3
-13.0%
-0.3
-13.0%
Total
-38.8
-5.3%
-153.7
-19.0%
-184.4
-20.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 5-3 Annual Total GHG Emissions Impacts in Calendar Years 2025,2040 and 2050 - Final Program vs.
Alt la using Analysis Method Ba

CY2025
CY2040
CY2050
MMT
CC^eq
% Change
MMT C02eq
% Change
MMT C02eq
% Change
Downstream
-27.8
-4.6%
-124.3
-18.4%
-148.4
-20.0%
Upstream
-9.5
-4.7%
-42.2
-18.7%
-50.5
-20.3%
HFCb
-0.1
-15.0%
-0.3
-13.0%
-0.3
-13.0%
Total
-37.4
-4.7%
-166.8
-18.5%
-199.2
-20.1%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b HFC represents HFC emission reductions and percent change from the vocational vehicle category only.

-------
Table 5-4 Annual Fuel Savings in Calendar Years 2025,2040 and 2050 - Final Program vs. Alt lb using
Analysis Method Aa

CY2025
CY2040
CY2050
Billion
Gallons
% Savings
Billion
Gallons
% Savings
Billion
Gallons
% Savings
Diesel
2.3
4.9%
9.2
17.8%
11.1
19.3%
Gasoline
0.4
5.0%
1.0
12.2%
1.2
12.8%
Total
2.7
4.9%
10.2
17.0%
12.3
18.5%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 5-5 Annual Fuel Savings in Calendar Years 2025,2040 and 2050 - Final Program vs. Alt la using
Analysis Method Aa

CY2025
CY2040
CY2050
Billion
Gallons
% Savings
Billion
Gallons
% Savings
Billion
Gallons
% Savings
Diesel
2.4
5.2%
10.2
19.0%
12.3
21.0%
Gasoline
0.5
5.6%
1.2
13.0%
1.3
14.0%
Total
2.9
5.2%
11.4
18.0%
13.6
20.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 5-6 Annual Fuel Savings in Calendar Years 2025,2040 and 2050 - Final Program vs. Alt la using
Analysis Method Ba

CY2025
CY2040
CY2050
Billion
Gallons
% Savings
Billion
Gallons
% Savings
Billion
Gallons
% Savings
Diesel
2.5
5.0%
10.8
19.4%
13.0
21.0%
Gasoline
0.3
2.8%
1.7
13.3%
1.9
14.4%
Total
2.8
4.6%
12.5
18.3%
14.9
19.9%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
The non-GHG impacts of the final program are largely driven by three factors. The
largest contributor is from the projected increased use of auxiliary power units (APUs), which
provide power, heat and cooling for trucks during extended engine idling. Note that since the
proposal, the assumptions of APU usage were changed in the final rulemaking (see Section
III.D.l.a of the Preamble) and EPA is adopting Phase 1 and Phase 2 requirements to control
PM2.5 emissions from APUs installed in new tractors (see Section III.C.3 of the Preamble).
Reduced emissions from upstream fuel production and distribution also contribute significantly
to the emissions benefits. Emissions of certain pollutants, such as NOx and PM2.5 are further
reduced through improved engine efficiency, aerodynamics and tire rolling resistance and

-------
absolute changes in average total running weight of the vehicles. To a smaller extent, a rebound
of vehicle miles traveled (VMT) will increase the emissions of all pollutants proportional to the
VMT rebound amount. The emissions impacts of non-GHGs on both downstream and upstream
from the heavy-duty sector in calendar years 2025, 2040 and 2050 are summarized in Table 5-7
through Table 5-9, using Method A and B, relative to the two reference cases.
Table 5-7 Annual Total Impacts (Upstream and Downstream) of Criteria Pollutants and Air Toxics from
Heavy-Duty Sector in Calendar Years 2025,2040 and 2050 - Final Program vs. Alt lb using Analysis Method
Aa
POLLUTANT
CY2025
CY2040
CY2050
US Short
Tons
% Change
US Short
Tons
% Change
US Short
Tons
% Change
1,3-Butadiene
0.3
0.1%
0.1
0.1%
-0.4
-0.3%
Acetaldehyde
-4
-0.1%
-30
-1.3%
-35
-1.4%
Acrolein
-0.2
0%
-2
-0.7%
-3
-0.9%
Benzene
-25
-1.2%
-101
-6.3%
-118
-6.7%
CO
-12,830
-0.9%
-49,416
-3.7%
-59,724
-4.0%
Formaldehyde
-39
-0.5%
-167
-2.7%
-205
-2.9%
NOx
-21,337
-2.0%
-89,218
-11.0%
-108,157
-12.0%
pm25
-1,033
-2.0%
-4,213
-10.0%
-5,071
-11.0%
SOx
-6,005
-4.9%
-23,401
-17.0%
-28,047
-19.0%
voc
-5,188
-2.7%
-18,293
-11.0%
-21,513
-12.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 5-8 Annual Total Impacts (Upstream and Downstream) of Criteria Pollutants and Air Toxics from
Heavy-Duty Sector in Calendar Years 2025,2040 and 2050 - Final Program vs. Alt la using Analysis Method
Aa
POLLUTANT
CY2025
CY2040
CY2050
US Short
Tons
% Change
US Short
Tons
% Change
US Short
Tons
% Change
1,3-Butadiene
0.2
0.1%
-0.2
-0.1%
-1
-0.5%
Acetaldehyde
-5
-0.2%
-29
-1.3%
-35
-1.4%
Acrolein
-0.2
0%
-2
-0.7%
-3
-1.0%
Benzene
-27
-1.4%
-110
-6.8%
-129
-7.2%
CO
-13,086
-0.9%
-50,800
-3.8%
-61,438
-4.1%
Formaldehyde
-40
-0.5%
-170
-2.7%
-207
-2.9%
NOx
-23,492
-2.2%
-100,407
-12.0%
-121,985
-14.0%
pm25
-1,143
-2.2%
-4,731
-12.0%
-5,707
-13.0%
SOx
-6,568
-5.3%
-25,902
-19.0%
-31,096
-20.0%
voc
-5,641
-3.0%
-19,954
-12.0%
-23,502
-13.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
Table 5-9 Annual Total Impacts (Upstream and Downstream) of Criteria Pollutants and Air Toxics from
Heavy-Duty Sector in Calendar Years 2025,2040 and 2050 - Final Program vs. Alt la using Analysis Method
Ba
POLLUTANT
CY2025
CY2040
CY2050
US Short
Tons
% Change
US Short
Tons
% Change
US Short
Tons
% Change
1,3-Butadiene
-2
-0.5%
-8
-3.7%
-9
-4.1%
Acetaldehyde
-10
-0.3%
-53
-2.0%
-61
-2.1%
Acrolein
-1
-0.1%
-4
-1.3%
-5
-1.3%
Benzene
-35
-1.1%
-165
-6.8%
-192
-7.5%
CO
-13,254
-0.6%
-52,594
-3.3%
-63,869
-3.8%
Formaldehyde
-40
-0.5%
-187
-2.7%
-227
-2.9%
NOx
-22,710
-1.9%
-101,961
-12.1%
-123,824
-13.3%
SOx
-1,110
-1.9%
-5,081
-11.1%
-6,100
-12.1%
PM2.5
-6,080
-4.8%
-26,933
-18.9%
-32,282
-20.5%
voc
-5,305
-2.2%
-25,070
-11.9%
-29,253
-13.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
5.2 Introduction
5.2.1 Downstream (Tailpipe) Emissions
As described in more detail in this chapter, the downstream reductions in emissions due
to the final program will be achieved through improvements in engine efficiency, road load
reduction, and projected increase in idle reduction technologies (for additional details, see
Chapter 5.3.2.3.1). Absolute reductions in tailpipe emissions are projected to grow over time as
the fleet turns over to vehicles affected by the final standards, meaning that the emissions
benefits of the program will continue to grow as older vehicles in the fleet are replaced by newer
vehicles that emit less CO2.
The effect of the regulations on the timing of fleet turnover and total VMT can have an
impact on downstream GHG and other emissions, as discussed in Section IX of the Preamble. If
the regulations spur firms to increase their purchase of new vehicles before efficiency standards
are in place ("pre-buy") or to delay their purchases once the standards are in place then there will
be a delay in achieving the full GHG and other emission reductions from improved fuel economy
across the fleet. If the lower per-mile costs associated with higher fuel economy lead to an
increase in VMT (the "rebound effect"), then the total emission reductions will also be reduced.
Chapter 8 of the RIA provides more detail on how the rebound effect was calculated in the
agencies' analysis. The analysis discussed in this chapter incorporates the rebound effect into
the estimates. However, the impacts of any delayed fleet turnover are not estimated.

-------
5.2.2 Upstream Emissions
In addition to downstream emission reductions, reductions are expected in the emissions
associated with the processes involved in getting fuel to the pump, including the extraction and
transportation of crude oil, the production and distribution of finished gasoline and diesel, and
the production and transportation of renewable fuels. Changes are anticipated in upstream
emissions due to the expected reduction in the overall volume of gasoline and diesel consumed.
Less fuel consumed means less fuel transported, less fuel refined, and less crude oil extracted
and transported to refineries. Thus, there will be reductions in the emissions associated with
each of these steps in the fuel production and distribution processes. In addition, any changes in
downstream reductions associated with changes in fleet turnover, and VMT are reflected in a
corresponding change in upstream emissions associated with fuel processing and distribution.
The agencies recognize that the standards could lower the world price of oil (the
"monopsony" effect, further discussed in Chapter 8 of the RIA). Lowering oil prices could lead
to an uptick in oil consumption globally, resulting in a corresponding increase in GHG emissions
in other countries. This global increase in emissions will slightly offset some of the emission
reductions achieved domestically as a result of the regulation. EPA does not provide quantitative
estimates of the impact of the final program on global petroleum consumption and GHG
emissions in this RIA.
5.2.3 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 5-10). When expressed in CCheq terms, each gas is
weighted by its heat trapping ability relative to that of CO2. The GWPs used in this analysis are
consistent with the 2007 Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment
Report (AR4) on a 100-year timescale.6
Table 5-10 Global Warming Potentials of GHGs
GAS
GLOBAL WARMING POTENTIAL (C02EQ)
O
O
1
ch4
25
n20
298
HFC134a
1,430
5.3 Program Analysis and Modeling Methods
5.3.1 Models Used
Different tools exist for estimating potential fuel consumption and emissions impacts
associated with fuel efficiency and GHG emissions standards. One such tool is EPA's official
mobile source emissions inventory model named Motor Vehicle Emissions Simulator
(MOVES).4 The agencies used a revised version of the official public model, MOVES2014a, to

-------
quantify the impacts of these standards on GHG emissions, fuel consumption, as well as criteria
pollutants and air toxics emissions.
Since the notice of proposed rulemaking, MOVES has undergone a series of updates in
response to the public comments on the proposal: (1) the projections of vehicle sales,
populations, and activity in the version used for the final rulemaking were updated to incorporate
the latest projections from the U.S. Department of Energy's Annual Energy Outlook 2015
report7; (2) the extended idle and APU emission rates in MOVES were updated based on the
analyses of latest test programs that reflect the current prevalence of clean idle certified engines;
and (3) the baseline adoption rates of idle reduction technology were reassessed and projected to
be lower than what was assumed in the proposal, as described in Section III.D.l.a of the
Preamble. In addition, changes to APU emissions rates for PM2.5 were implemented in MOVES
reflecting the fact that EPA is adopting requirements to control PM2.5 emissions from APUs, as
discussed in Section III.C.3 of the Preamble. Finally, methodological improvements were made
in classifying vehicle types and in forecasting vehicle populations and activity. The
aforementioned updates above, along with other changes, are documented in the memorandum to
the docket.8
The agencies ran MOVES with user input databases that reflected the projected
technological improvements resulting from the final rules, such as the improvements in engine
and vehicle efficiency, aerodynamic drag, and tire rolling resistance. The changes made to the
default MOVES database are described below in Chapter 5.3.2. All the input data, MOVES
runspec files, and the scripts used for the analysis, as well as the version of MOVES used to
generate the emissions inventories, can be found in the docket.9
Another such tool is DOT's CAFE model. For this analysis, the model was reconfigured
to use the work based attribute metric of "work factor" established in the Phase 1 rule for heavy-
duty pickups and vans, instead of the light-duty "footprint" attribute metric. The CAFE model
takes user-specified inputs on, among other things, vehicles that will be produced in a given
model year, technologies available to improve fuel efficiency on those vehicles, potential
regulatory standards that will drive improvements in fuel efficiency, and economic assumptions.
The CAFE model takes every vehicle in each manufacturer's fleet and decides what technologies
to add to those vehicles in order to allow each manufacturer to comply with the standards in the
most cost-effective way and uses a representation of the HD pickup and van fleet that captures
heterogeneity at the manufacturer, model year, and powertrain (and other technology) level.
Based on the resulting improved vehicle fleet, the CAFE model then calculates total fuel
consumption and GHG, criteria, and toxics emissions impacts based on those inputs, along with
economic costs and benefits. The CAFE model is discussed in greater detail in Chapter 10 of the
RIA.
For this rulemaking, the agencies conducted two analyses by employing DOT's CAFE
model and EPA's MOVES model. These models were used to project the impacts resulting from
the standards on fuel consumption, GHG emissions, as well as criteria pollutants and air toxics
emissions. As described in Chapter 5.3.2, the agencies used EPA's MOVES model to estimate
fuel consumption and emissions impacts for tractor-trailers (including the engines which power
the vehicle), and vocational vehicles (including the engine which powers the vehicle). For
heavy-duty pickups and vans, the agencies performed separate analyses, referred to as "Method

-------
A" and "Method B," to estimate fuel consumption and emissions from these vehicles. For these
methods, the agencies analyzed the impact of the final rules, relative to two different reference
cases - flat and dynamic. The flat baseline projects very little improvement in new vehicles in
the absence of new Phase 2 standards. In contrast, the dynamic baseline projects more
significant improvements in vehicle fuel efficiency. The agencies considered both reference
cases. The results for all of the regulatory alternatives relative to both reference cases, derived
via the same methodologies discussed in this Chapter, are presented in Chapter 11 of the RIA.
For brevity, a subset of these analyses are presented in this section, and the reader is
referred to both Chapter 11 of the RIA and NHTSA's FEIS Chapters 3, 4 and 5 for complete sets
of these analyses. In this Chapter, Method A is presented for the final standards, relative to both
the dynamic baseline (Alternative lb) and the flat baseline (Alternative la). Method B is
presented for the final standards, relative only to the flat baseline.
Because reducing fuel consumption also affects emissions that occur as a result of fuel
production and distribution (including renewable fuels), the agencies also calculated those
"upstream" changes using the "downstream" fuel consumption reductions predicted by the
MOVES model for vocational vehicles and tractor-trailers. As described earlier, for HD pickups
and vans, separate analyses of estimating the emissions from upstream processes were conducted
using the fuel consumption estimates from DOT's CAFE model (Method A) and EPA's MOVES
model (Method B), relative to the two reference cases. Method A used a modified version of the
CAFE model to estimate vehicular fuel consumption and emissions impacts for HD pickups and
vans and to calculate upstream impacts. For vocational vehicles and combination tractor-trailers,
both Method A and Method B estimated the projected corresponding changes in upstream
emissions using the same tools originally created for the Renewable Fuel Standard 2 (RFS2)
rulemaking analysis,10 used in the LD GHG rulemakings,11 HD GHG Phase 1,12 and updated for
the current analysis. The estimate of emissions associated with production and distribution of
gasoline and diesel from crude oil is based on emission factors in the "Greenhouse Gases,
Regulated Emissions, and Energy Use in Transportation" model (GREET) developed by DOE's
Argonne National Lab. In some cases, the GREET values were modified or updated by the
agencies to be consistent with the National Emission Inventory (NEI) and emission factors from
MOVES. Method B used the same tool described above to estimate the upstream impacts for
HD pickups and vans.
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 accounting for impacts of fuel requirements on vapor emissions from storage
and transport. 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 2005 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 rulemaking
(MSAT) inventory for benzene; these pollutant tons were divided by refinery energy use or
gasoline distribution quantities published by the DOE Energy Information Administration (EIA)
to get emission factors in terms of grams per million BTU of finished gasoline and diesel. These
updates are consistent with those used for the upstream analysis included in the LD GHG
rulemaking and HD GHG Phase 1. The actual calculation of the emission inventory impacts of

-------
the decreased gasoline and diesel production is done in EPA's tool for upstream emission
impacts.13
5.3.2 Calculation of Downstream Emissions
5.3.2.1	Model inputs and Assumptions for the Flat Reference Case
The flat reference case (identified as Alternative la in Section X of the Preamble and
Chapter 11 of the RIA), a "no action" alternative, functions as one the baselines against which
the impacts of the standards can be evaluated. The MOVES2014a default road load parameters
and energy rates were used for the vocational vehicles and HD pickups and vans for this
alternative because we assumed no market-driven improvements in fuel efficiency. The tractor-
trailer road load parameters were changed from the MOVES2014a default values to account for
projected improvements in the efficiency of the box trailers pulled by combination tractors due to
increased penetration of aerodynamic technologies and low rolling resistance tires attributed to
both EPA's SmartWay Transport Partnership and California Air Resources Board's Tractor-
Trailer Greenhouse Gas regulation, as described in Section IV of the Preamble. We maintained
the same road load inputs for tractor-trailers for 2018 and beyond.
The flat reference case assumed the growth in vehicle populations and miles traveled
based on the relative annual VMT growth from AEO2015 Final Release for model years 2014
and later.7 In the proposal, the agencies assumed the baseline APU adoption rate of 30 percent.
However, based on the comments received from the proposed rulemaking, the flat reference case
assumes that 9 percent of all combination long-haul tractors model year 2010 and later use an
APU during extended idling (see Section III.D.l.a of the Preamble).
5.3.2.2	Model inputs and Assumptions for the Dynamic Reference Case
The dynamic reference case (identified as Alternative lb in Section X of the Preamble
and Chapter 11 of the RIA), also includes the impact of Phase 1 and generally assumes that fuel
efficiency and GHG emission standards are not improved beyond the required 2018 model year
levels. However, for this case, the agencies assume market forces will lead to additional fuel
efficiency improvements for tractors and trailers. These additional assumed improvements are
described in Section X of the Preamble. No additional fuel efficiency improvements due to
market forces were assumed for vocational vehicles. For HD pickups and vans, the agencies
applied the CAFE model using the input assumption that manufacturers having achieved
compliance with Phase 1 standards will continue to apply technologies for which increased
purchase costs will be "paid back" through corresponding fuel savings within the first six months
of vehicle operation. The agencies conducted the MOVES analysis of this case in the same
manner as for the flat reference case.
5.3.2.3	Model Inputs and Assumptions for the Control Case
The control case (identified as Alternative 3 in Chapter 11 of the RIA) represents the
agencies' fuel efficiency and GHG standards for HD engines, HD pickup trucks and vans, Class
2b through Class 8 vocational vehicles, Class 7 and 8 combination tractors, and trailers. To
account for improvements of engine and vehicle efficiency in vocational vehicles and

-------
combination tractor-trailers, EPA developed additional user input data for MOVES runs in
estimating the control case inventories.
The agencies used the percent reduction in aerodynamic drag and tire rolling resistance
coefficients and absolute changes in average total running weight (gross combined weight)
expected from the final rules to develop the road load inputs for the control case. For running
emissions, the key concept underlying the definition of operating mode in MOVES is scaled
tractive power (STP), vehicle speed and vehicle acceleration.14 STP represents the vehicle's
tractive power scaled by a constant factor. It is calculated using mass of the vehicle and road
load factors that include tire rolling resistance, aerodynamic drag, and friction losses in the
drivetrain. STP is estimated using the equation below:
Av, +Bv,2 +CV,3 +mvtat
STPt = —'
Where:
•fscale
Equation 5-1
A = the rolling resistance coefficient [kW-sec/m],
B = the rotational resistance coefficient [kW-sec2/m2],
C = the aerodynamic drag coefficient [kW-sec3/m3],
m = mass of individual vehicle [metric ton],
/scale = fixed mass factor,
vt = instantaneous vehicle velocity at time t [m/s],
at = instantaneous vehicle acceleration [m/s2]
The improvements in road load factors will reduce the tractive power exerted by a vehicle
to move itself and its cargo. The emissions from heavy-duty trucks are a function of STP as
determined from a variety of data sources. Thus, a reduction in road load factors are expected to
result in reduced GHG and non-GHG emissions. The improvements in tire rolling resistance,
aerodynamic drag, and absolute changes in average vehicle weight expected from the
technologies which could be used to meet the standards were modified in the
"sourceusetypephysics" table.A
For vocational vehicles and tractor-trailers, the agencies developed energy inputs for the
control case runs using the percent reduction in CO2 emissions expected from the powertrain and
other vehicle technologies not accounted for in the aerodynamic drag and tire rolling resistance
improvements in the final rules. In contrast, for HD pickup trucks and vans, the standards were
evaluated only in terms of the total vehicle reductions in fuel use and CO2 emissions, since
nearly all of these vehicles would be certified on a chassis dynamometer. Finally, EPA assumed
increased penetration of idle reduction technology during extended idling, based on the
expectation that manufacturers will use APUs and other idle reduction technologies to meet the
A Class 2b and 3 trucks do not use the STP metric and are regulated based on chassis testing (gram per mile basis)
rather than engine testing (gram per brake horsepower-hour basis), therefore road load reductions are not expected to
result in reduced non-GHG emissions.

-------
vehicle GHG standard for combination long-haul tractors, as discussed in Section III.D of the
Preamble.
5.3.2.3.1 Emission Rate and Road Load Inputs
Both the stringency and the form of the fuel consumption and CO2 emission standards
vary by vehicle category. Accordingly, the modeling of the standards in MOVES varies by the
vehicle category. For the vocational vehicles and combination tractor-trailers, EPA has analyzed
the impacts of the standards by evaluating the technologies applied to the energy rates as well as
to the road load inputs. However, the impacts on the HD pickup trucks and vans were estimated
only in terms of reduction in energy rates.
5.3.2.3.1.1 Tractor-Trailers
Similar to the approach used in the HD Phase 1 analysis, EPA aggregated the nine tractor
subcategories into the two MOVES combination tractor-trailer categories - short-haul and long-
haul. The agencies used sales distribution data from the HD Phase 1 analysis and determined the
long-haul reductions in energy rates and road load factors, based on a sales mix assumption of 80
percent high roof, 15 percent mid roof, and 5 percent low roof sleeper cabs. The short-haul
combination tractors were evaluated using a day cab sales distribution assumption of 7 percent
Class 7 low roof, 10 percent Class 7 high roof, 40 percent Class 8 low roof, 35 percent Class 8
high roof, and 8 percent vocational tractors, based on the information used in the HD Phase 1
analysis. The details of the analyses aggregating the tractor subcategories into MOVES
categories using the sales mix assumption described above can be found in the docket.15
The trailer category encompasses many types of trailers. As with the tractor category,
EPA aggregated the trailer subcategories into two MOVES combination tractor-trailer
categories. EPA used a combination of ACT Research's 2013 factory shipment data16 for trailer
distribution by type and "primary trip length" information from the U.S. Census' 2002 Vehicle
Inventory and Use Survey17 to distribute each trailer type into long- and short-haul categories.
EPA applied the trailer market percentages as shown in Table 5-11 to determine the trailer
impact on the MOVES long- and short-haul combination tractor-trailer categories.
Table 5-11 Aggregation of Trailer Types into MOVES Combination Tractor-Trailer Categories
TRAILER TYPE
Combination Long-Haul
Combination Short-Haul Tractor-

Tractor-Trailers
Trailers
Long Dry Van
51.6%
15.6%
Short Dry Van
20.6%
27.9%
Long Refrigerated Van
21.2%
2.5%
Short Refrigerated Van
6.6%
3.9%
Container Chassis
0.0%
8.4%
Flatbed
0.0%
8.4%
Tank
0.0%
8.3%
Excluded Trailers
0.0%
25.0%
Table 5-12 describes the improvements in the energy rate expected from the heavy-duty
engine, transmission, and driveline technologies which will be applied to meet the tractor

-------
standards. The percentage reductions from the reference case were applied to the default
MOVES energy rates in the appropriate source bins by modifying MOVES
"emissionrateadjustment" table.
Table 5-12 Estimated Reductions in Energy Rates for the Final Standards for Tractor-Trailers
VEHICLE TYPE
FUEL
MODEL YEARS
REDUCTION FROM
FLAT BASELINE
Long-haul Tractor-
Trailers
Diesel
2018-2020
1.0%
2021-2023
7.9%
2024-2026
12.4%
2027+
16.3%
Short-haul Tractor-
TrailersB
Diesel
2018-2020
0.6%
2021-2023
7.4%
2024-2026
11.9%
2027+
15.0%
Table 5-13 contains the improvements in tire rolling resistance, coefficient of drag, and
weight reductions expected from the technologies which could be used to meet the Phase 2
standards for combination tractor-trailers. The percentage reductions in tire rolling resistance
and drag coefficients and the absolute changes in average vehicle weight were modified in the
"sourceusetypephysics" table.
Table 5-13 Estimated Reductions in Road Load Factors for the Final Standards for Tractor-Trailers
VEHICLE TYPE
MODEL
REDUCTION IN
REDUCTION IN
WEIGHT

YEARS
TIRE ROLLING
AERODYNAMIC
REDUCTION


RESISTANCE
DRAG COEFFICIENT
(LB)a


COEFFICIENT


Combination Long-
haul Tractor-
Trailers
2018-2020
6.1%
5.6%
-140
2021-2023
13.3%
12.5%
-199
2024-2026
16.3%
19.3%
-294
2027+
18.0%
28.2%
-360
Combination Short-
haul Tractor-
Trailers0
2018-2020
5.2%
0.9%
-23
2021-2023
11.9%
4.0%
-43
2024-2026
14.1%
6.2%
-43
2027+
15.9%
8.8%
-43
Note:
a Negative weight reductions reflect an expected weight increase as a byproduct of aerodynamic improvements and
other improvements to the vehicle.
B Vocational and heavy-haul tractors are included in the short-haul tractor segment.
c Vocational and heavy-haul tractors are included in the short-haul tractor segment.

-------
In addition, the projected use of auxiliary power units (APUs) during extended idling,
shown below in Table 5-14, was included in the modeling for the long-haul combination tractor-
trailers by modifying the "hotellingactivitydistribution" table in MOVES.
Table 5-14 Assumed APU Use during Extended Idling for Combination Long-haul Tractor-Trailers a
VEHICLE TYPE
MODEL
YEARS
DIESEL APU
PENETRATION
BATTERY APU
PENETRATION
Combination
Long-Haul
Trucks
2010-2020
9%
0%
2021-2023
30%
10%
2024-2026
40%
10%
2027+
40%
15%
Note:
a Other idle reduction technologies (such as automatic engine shutdown, fuel operated heaters, and
stop-start systems) were modeled as part of the energy rates.
5.3.2.3.1.2 Vocational Vehicles
Similar to the approach for tractor-trailers, EPA aggregated the nine vocational vehicle
subcategories into each of the seven MOVES vehicle types.0 The energy rate inputs were
derived by applying the anticipated levels of engine, axle, transmission, and idle reduction
technologies across the weight classes and vehicle types. Each of these technology packages is
described in Chapter 2 of the RIA. The differences between gasoline and diesel vocational
vehicles in energy rate reduction from the reference cases, shown in Table 5-15, are due to the
differences in anticipated engine-level technology packages, as described in Chapter 2 of the
RIA.
The percentage reductions from the reference case were applied to the default MOVES
energy rates in the appropriate source bins by modifying MOVES "emissionrateadjustment"
table.
D Seven MOVES vehicle types for vocational vehicles are intercity bus, transit bus, school bus, refuse truck, single-
unit short-haul truck, single-unit long-haul truck, and motor home.

-------
Table 5-15 Estimated Reductions in Energy Rates for the Final Standards for Vocational Vehicles
VEHICLE TYPE
FUEL
MODEL
YEARS
REDUCTION FROM
FLAT BASELINE
Single-Frame
Vocational
Diesel & CNG
2021-2023
7.8%
2024-2026
12.3%
2027+
16.0%
Gasoline
2021-2023
6.9%
2024-2026
9.8%
2027+
13.3%
Urban Buses
Diesel & CNG
2021-2023
7.0%
2024-2026
11.8%
2027+
14.4%
The agencies used MOVES population data for new vehicles expected to be sold in 2018
for each weight class, as well as assumptions about their distribution among the three new
vocational vehicle duty cycles. This population allocation is shown in Table 5-16.
Table 5-16 Vocational Vehicle Types and Population Allocation
VEHICLE TYPE
REGIONAL
MULTI-PURPOSE
URBAN
Short Haul Straight Truck
20%
28%
52%
Long Haul Straight Truck, Motor
Home, Intercity Bus
100%
0%
0%
School Bus
0%
10%
90%
Transit Bus
0%
0%
100%
Refuse
0%
10%
90%
All Class 4-5
15%
10%
19%
All Class 6-7
11%
7%
19%
All Class 8
5%
4%
10%
Using these population distribution estimates and the technology application rates
described in Chapter 2 of the RIA, EPA derived the levels of improvements in tire rolling
resistance and weight reduction.
Table 5-17 contains the improvements in tire rolling resistance, and weight reductions
expected from the technologies which will be used to meet the standards for vocational vehicles.
No reduction in aerodynamic drag coefficient was modeled for vocational vehicles because the
final standards for vocational vehicles do not assume any aerodynamic improvements (see
E Vocational vehicles modeled in MOVES include heavy heavy-duty, medium heavy-duty, and light heavy-duty
vehicles. However, for light heavy-duty vocational vehicles, class 2b and 3 vehicles are not included in the
inventories for the vocational sector. Instead, all vehicles with GVWR less than 14,000 lbs were modeled using the
energy rate reductions described below for HD pickup trucks and vans. In practice, many manufacturers of these
vehicles choose to average the lightest vocational vehicles into chassis-certified families (i.e., heavy-duty pickups
and vans).

-------
Section V.C. 1 .c.i of the Preamble). The percentage reductions in tire rolling resistance and the
absolute changes in average vehicle weight were modified in the "sourceusetypephysics" table in
MOVES. The analyses used to develop the MOVES inputs for vocational vehicles, described
above, can be found in the docket.18
Table 5-17 Estimated Reductions in Road Load Factors for the Final Standards for Vocational Vehicles
VEHICLE
MODEL
REDUCTION IN TIRE
WEIGHT
TYPE
YEARS
ROLLING
RESISTANCE
COEFFICIENT
REDUCTION
(LB)
Intercity
2021-2023
18.2%
0
Buses
2024-2026
20.8%
0

2027+
24.7%
0
Transit Buses
2021-2023
0%
0

2024-2026
0%
0

2027+
12.1%
0
School Buses
2021-2023
10.1%
0

2024-2026
14.9%
0

2027+
19.7%
0
Refuse Trucks
2021-2023
0%
0

2024-2026
0%
0

2027+
12.1%
0
Single Unit
2021-2023
6.4%
4.4
Short-haul
2024-2026
6.4%
10.4
Trucks
2027+
10.2%
16.5
Single Unit
2021-2023
8.4%
7.9
Long-haul
2024-2026
13.3%
23.6
Trucks
2027+
13.3%
39.4
Motor Homes
2021-2023
20.8%
0

2024-2026
20.8%
0

2027+
24.7%
0
5.3.2.3.1.3 Heavy-Duty Pickup Trucks and Vans
As explained above, the agencies used both DOT's CAFE model and EPA's MOVES
model, using analytical Method A and B, respectively, to project fuel consumption, GHG and
non-GHG emissions impacts resulting from the standards for HD pickups and vans, including
downstream vehicular emissions as well as emissions from upstream processes related to fuel
production, distribution, and delivery.
5.3.2.3.1.3.1 Method A for HD Pickups and Vans
For Method A, NHTSA used the CAFE model which applies fuel properties (density and
carbon content) to estimated fuel consumption in order to calculate vehicular CO2 emissions,
applies per-mile emission factors from MOVES to estimated VMT (for each regulatory
alternative, adjusted to account for the rebound effect) in order to calculate vehicular CH4 and
N2O emissions (as well, as discussed below, of non-GHG pollutants), and applies per-gallon

-------
upstream emission factors from GREET in order to calculate upstream GHG (and non-GHG)
emissions.
As discussed above in Section VI, the standards for HD pickups and vans increase in
stringency by 2.5 percent annually during model years 2021-2027. The standards define targets
specific to each vehicle model, but no vehicle is required to meet its target; instead, the
production-weighted averages of the vehicle-specific targets define average fuel consumption
and CO2 emission rates that a given manufacturer's overall fleet of produced vehicles is required
to achieve. The standards are specified separately for gasoline and diesel vehicles, and vary with
work factor. Both the NPRM and today's analysis assume that some application of mass
reduction could enable increased work factor in cases where manufacturers increase a vehicle's
rated payload and/or towing capacity, but there are other ways manufacturers may change work
factor which the analysis does not capture. Average required levels will depend on the future
mix of vehicles and the work factors of the vehicles produced for sale in the U.S. Since these can
only be estimated at this time, average required and achieved fuel consumption and CO2
emission rates are subject to uncertainty. Between the NPRM and the issuance of today's final
rules, NHTSA updated the market forecast (and other inputs) used to analyze HD pickup and van
standards, and doing so leads to different estimates of required and achieved fuel consumption
and CO2 emission rates (as well as different estimates of impacts, costs, and benefits).
The following four tables present stringency increases and estimated required and
achieved fuel consumption and CO2 emission rates for the two No Action Alternatives
(Alternative la and lb) and the standards defining the Preferred Alternative. Stringency
increases are shown relative to standards applicable in model year 2018 (and through model year
2020). As mathematical functions, the standards themselves are not subject to uncertainty. By
2027, they are 16.2 percent more stringent (i.e., lower) than those applicable during 2018-2020.
NHTSA estimates that, by model 2027, these standards could reduce average required fuel
consumption and CO2 emission rates to about 4.88 gallons/100 miles and about 4 grams/mile,
respectively. NHTSA further estimates that average achieved fuel consumption and CO2
emission rates could correspondingly be reduced to about the same levels. If, as represented by
Alternative lb, manufacturers will, even absent today's standards, voluntarily make
improvements that pay back within six months, these model year 2027 levels are about 12
percent lower than the agencies' estimate could be achieved under the Phase 1 standards defining
the No Action Alternative. If, as represented by Alternative la, manufacturers will, absent
today's standards, only apply technology as required to achieve compliance, these model year
2027 levels are about 13 percent lower than the agencies' estimate could be achieved under the
Phase 1 standards. As indicated below, NHTSA's estimate that these improvements in fuel
consumption and CO2 emission rates will build from model year to model year, beginning as
soon as model year 2017 (insofar as manufacturers may make anticipatory improvements if
warranted given planned product cadence).
The NPRM analysis suggested that both the achieved and required fuel consumption and
CO2 reductions would be larger than the current Method A analysis suggests. The NPRM
suggested that achieved reductions would be 13.5 and 15 percent, for the dynamic and flat
baselines, respectively. The change in the standards and fuel consumption reductions can be
attributed to the projected increased work factor of the 2015 fleet relative to the 2014 fleet.

-------
Section VI discusses in more detail the changes in the distribution of work factor for key market
players from the MY2014 to the MY2015 fleet.
Table 5-18 Stringency of HD Pickup and Van Standards, Estimated Average Required and Achieved Fuel
Consumption Rates for Method A, Relative to Alternative lb a
MODEL
YEAR
STRINGENCY
(VS. 2018)
AVE. REQUIRED FUEL CONS.
(GAL./100 MI.)
AVE. ACHIEVED FUEL CONS.
(GAL./100 MI.)
No
Final
Reduction
No
Final
Reduction
2016
MYs 2016-
2020 Subject to
Phase 1
Standards
6.32
6.32
0.0%
6.14
6.14
0.0%
2017
6.16
6.16
0.0%
6.02
5.89
2.2%
2018
5.83
5.83
0.0%
5.97
5.78
3.2%
2019
5.81
5.81
0.0%
5.77
5.47
5.3%
2020
5.80
5.80
0.0%
5.75
5.46
5.1%
2021
2.5%
5.79
5.65
2.4%
5.68
5.28
7.2%
2022
4.9%
5.80
5.52
4.8%
5.64
5.22
7.5%
2023
7.3%
5.80
5.38
7.2%
5.64
5.21
7.6%
2024
9.6%
5.80
5.25
9.5%
5.65
5.22
7.6%
2025
11.9%
5.81
5.12
11.8%
5.65
5.14
9.1%
2026
14.1%
5.81
5.01
13.7%
5.65
5.02
11.1%
2027
16.2%
5.80
4.88
15.8%
5.57
4.92
11.7%
2028*
16.2%
5.81
4.91
15.5%
5.57
4.89
12.2%
2029*
16.2%
5.81
4.91
15.6%
5.57
4.88
12.4%
2030*
16.2%
5.81
4.91
15.6%
5.57
4.88
12.4%
Notes:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline,
la, and dynamic baseline, lb, please see Section X. A. 1.
* Absent further action, standards assumed to continue unchanged after model year 2027.

-------
Table 5-19 Stringency of HD Pickup and Van Standards, Estimated Average Required and Achieved CO2
Emission Rates for Method A, Relative to Alternative lb a
MODEL
YEAR
STRINGENCY
(VS. 2018)
AVE. REQUIRED C02 RATE
(G./MI.)
AVE. ACHIEVED C02 RATE
(G./MI.)
No
Final
Reduction
No Action
Final
Reduction
2016
MYs 2016-
2020 Subject to
Phase 1
Standards
597
597
0.0%
578
578
0.0%
2017
582
582
0.0%
567
554
2.2%
2018
550
550
0.0%
562
544
3.2%
2019
548
548
0.0%
543
514
5.3%
2020
547
547
0.0%
541
513
5.1%
2021
2.5%
545
532
2.4%
534
496
7.1%
2022
4.9%
546
519
4.9%
530
491
7.4%
2023
7.3%
545
506
7.2%
529
490
7.5%
2024
9.6%
547
494
9.5%
531
491
7.5%
2025
11.9%
547
483
11.7%
530
483
9.0%
2026
14.1%
547
472
13.7%
530
472
11.0%
2027
16.2%
546
460
15.8%
523
462
11.5%
2028*
16.2%
547
462
15.5%
523
460
12.0%
2029*
16.2%
547
462
15.5%
524
460
12.2%
2030*
16.2%
547
462
15.5%
524
460
12.2%
Notes:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline,
la, and dynamic baseline, lb, please see Section X. A. 1.
* Absent further action, standards assumed to continue unchanged after model year 2027.

-------
Table 5-20 Stringency of HD Pickup and Van Standards, Estimated Average Required and Achieved Fuel
Consumption Rates for Method A, Relative to Alternative laa
MODEL
YEAR
STRINGENCY
(VS. 2018)
AVE. REQUIRED FUEL CONS.
(GAL./100 MI.)
AVE. ACHIEVED FUEL CONS.
(GAL./100 MI.)
No Action
Final
Reduction
No Action
Final
Reduction
2016
MYs 2016-
2020 Subject to
Phase 1
Standards
6.32
6.32
0.0%
6.14
6.14
0.0%
2017
6.16
6.16
0.0%
6.00
5.85
2.4%
2018
5.83
5.83
0.0%
5.94
5.75
3.2%
2019
5.81
5.81
0.0%
5.74
5.43
5.4%
2020
5.80
5.80
0.0%
5.73
5.43
5.2%
2021
2.5%
5.79
5.65
2.4%
5.70
5.27
7.5%
2022
4.9%
5.80
5.52
4.8%
5.69
5.23
8.2%
2023
7.3%
5.80
5.38
7.2%
5.69
5.22
8.3%
2024
9.6%
5.80
5.25
9.5%
5.70
5.22
8.3%
2025
11.9%
5.81
5.13
11.8%
5.70
5.13
10.0%
2026
14.1%
5.81
5.02
13.6%
5.70
5.03
11.9%
2027
16.2%
5.80
4.89
15.8%
5.64
4.92
12.8%
2028*
16.2%
5.81
4.91
15.4%
5.64
4.89
13.3%
2029*
16.2%
5.81
4.91
15.5%
5.64
4.89
13.4%
2030*
16.2%
5.81
4.91
15.5%
5.64
4.89
13.4%
Notes:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline,
la, and dynamic baseline, lb, please see Section X. A. 1.
* Absent further action, standards assumed to continue unchanged after model year 2027.
** Increased work factor for some vehicles produces a slight increase in average required fuel consumption.

-------
Table 5-21 Stringency of HD Pickup and Van Standards, Estimated Average Required and Achieved CO2
Emission Rates for Method A, Relative to Alternative laa
MODEL
YEAR
STRINGENCY
(VS. 2018)
AVE. REQUIRED C02 RATE
(G./MI.)
AVE. ACHIEVED C02 RATE
(G./MI.)
No Action
Final
Reduction
No Action
Final
Reduction
2016
MYs 2016-
2020 Subject to
Phase 1
Standards
597
597
0.0%
578
578
0.0%
2017
582
582
0.0%
564
551
2.3%
2018
550
550
0.0%
559
541
3.2%
2019
548
548
0.0%
540
511
5.4%
2020
547
547
0.0%
538
510
5.2%
2021
2.5%
545
532
2.4%
535
495
7.4%
2022
4.9%
546
519
4.8%
534
491
8.0%
2023
7.3%
545
506
7.2%
533
490
8.2%
2024
9.6%
547
494
9.5%
535
491
8.2%
2025
11.9%
547
483
11.7%
535
483
9.8%
2026
14.1%
547
472
13.6%
535
473
11.7%
F 2027
16.2%
546
460
15.8%
529
462
12.6%
2028*
16.2%
547
462
15.5%
530
460
13.1%
2029*
16.2%
547
462
15.5%
530
460
13.2%
2030*
16.2%
547
462
15.5%
530
460
13.2%
Notes:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline,
la, and dynamic baseline, lb, please see Section X. A. 1.
* Absent further action, standards assumed to continue unchanged after model year 2027.
** Increased work factor for some vehicles produces a slight increase in the average required CO2 emission rate.
While the above tables show NHTSA's estimates of average fuel consumption and CO2
emission rates manufacturers of pickups and vans might achieve under today's standards, total
U.S. fuel consumption and GHG emissions from HD pickups and vans will also depend on how
many of these vehicles are produced, and how they are operated over their useful lives. Relevant
to estimating these outcomes, the CAFE model applies vintage-specific estimates of vehicle
survival and mileage accumulation, and adjusts the latter to account for the rebound effect. This
impact of the rebound effect is specific to each model year (and, underlying, to each vehicle
model in each model year), varying with changes in achieved fuel consumption rates. These
details of the model are further discussed in Chapter 10 of this RIA and Section VI of the
Preamble.
For Method B, MOVES model was used to estimate fuel consumption and GHG
emissions for HD pickups and vans. MOVES evaluated the standards for HD pickup trucks and
vans in terms of grams of CO2 per mile or gallons of fuel per 100 miles. Since nearly all HD
pickup trucks and vans are certified on a chassis dynamometer, the CO2 reductions for these
vehicles were not represented as engine and road load reduction components, but rather as total
vehicle CO2 reductions. The stringency increases relative to the Phase 1 standards for HD
pickup trucks and vans (Table 5-23) were modified in the "emissionrateadjustment" table in
MOVES.

-------
Table 5-22 Estimated Total Vehicle CO2 Reductions for the Final Standards and In-Use Emissions for HD
Pickup Trucks and Vans in Method B a
VEHICLE TYPE
FUEL
MODEL YEAR
C02 REDUCTION
FROM FLAT
BASELINE
HD Pickup Trucks
and Vans
Gasoline
and Diesel
2021
2.50%
2022
4.94%
2023
7.31%
2024
9.63%
2025
11.89%
2026
14.09%
2027+
16.24%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an
explanation of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1
5.3.2.3.2 VMT Inputs
The "HPMSVtype" table in MOVES was modified to reflect the VMT rebound (VMT
rebound is described in more detail in Chapter 8.3 of the RIA). This table includes VMT values
for all calendar years. For the control case, the absolute VMT for vocational vehicle and
combination tractor-trailer were increased from the reference cases by 0.30 percent, and 0.75
percent, respectively, to reflect the VMT rebound.19 Since VMT is applied by calendar year and
not by model year, post-processing of the results were performed to ensure that only the model
years affected by the program experienced VMT rebound - the results from the reference cases
were used in the control case inventories for model years not affected by the final rules.
For HD pickups and vans, Method A used the CAFE model, which simulates VMT in a
dynamic fashion that responds to changes in vehicle fuel economy and fuel prices and adjusts the
marginal VMT of each vehicle model, at every age (so in each calendar year). In general, the
more stringent alternatives considered lead to larger improvements in fuel economy and, thus, a
greater number of vehicle miles traveled as a result of the rebound effect. In the CAFE model,
the rebound effect represents a symmetric driver of changes to VMT; if the per-mile price of
driving declines relative to today (either from improvements in vehicle fuel economy or declines
in fuel prices), VMT increases by the amount of the rebound effect, conversely, if the per-mile
price of driving increases relative to today (due to increases in the price of fuel), VMT will
decline by the amount of the rebound effect. In Method B, the VMT rebound effect was
modeled using the MOVES model, which assumed an increase in VMT from the reference levels
of 1.08 percent.19
5.3.3 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; this

-------
includes all stages prior to the final filling of vehicle fuel tanks at retail service stations.
Additionally, it includes the production of renewable fuels and transportation of such fuel, either
separately or mixed with conventional fuels.
As described in Chapter 5.3.1, the decreased volumes of the crude based fuels and the
various crude production and transport emission factors from GREET were used to estimate the
net impact of fuel use changes on upstream emissions. The analysis for this final rulemaking
assumes that all changes in volumes of fuel used affect only gasoline and diesel, with no effects
on use of ethanol, biodiesel or other renewable fuels. The production and transport of these
renewable fuels, although unchanged in volume for this analysis, are still accounted for in the
total inventory in this rulemaking. Although impacts to agriculture related to renewable fuels
and the associated transport of these feedstocks were originally included in the RFS2
rulemaking, the effects to these sectors from the regulations will be minimal and they have
therefore been excluded from this analysis.
The agencies recognize the unique GHG emission characteristics associated with
biofuels, and specifically that in the context of biofuels, "upstream emissions" include not only
GHG emissions, but also any net biological sequestration that takes place. When considered on
a lifecycle basis (including both tailpipe and upstream emissions), the net GHG emission impact
of individual biofuels can vary significantly from both petroleum-based fuels and from one
biofuel to another. EPA's Renewable Fuel Standard (RFS) program, as modified by EISA,
examined these differences in lifecycle emissions in detail. For example, EPA found that with
respect to aggregate lifecycle emissions including non-tailpipe GHG emissions (such as
feedstock growth, transportation, fuel production, and land use), lifecycle GHG emissions in
2022 for biodiesel from soy, using certain advanced production technologies, are about 50
percent less than diesel from petroleum.
Non-GHG fuel production and distribution emission impacts of the program were
estimated in conjunction with the development of lifecycle 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 sub-process. It relies
partially on the GREET model, but takes advantage of additional information and models to
significantly strengthen and expand on the GREET analysis, as discussed in Chapter 5.3.1. The
details of the assumptions, data sources, and calculations that were used to estimate the emission
impacts presented here can be found in the docket memo, "Calculation of Upstream Emissions
for the GHG Vehicle Rule," initially created for use in the LD GHG rulemaking.20 The agencies
note that to the extent future policy decisions involve upstream emissions, the agencies will need
to consider the unique emission characteristics associated with biofuels. More broadly, the
agencies recognize that biofuels, including biodiesel, will play an important role in reducing the
nation's dependence on foreign oil, thereby increasing domestic energy security. The volumes
of renewable fuels are defined by the RFS2 standards as well as the Annual RFS rulemakings,
and are projected using AEO2015. The volumes of renewable fuel for these standards remain in
place regardless of overall volume of fuel affected by this rulemaking. Therefore, we have
assumed that the effect of the Phase 2 standards on biofuels agriculture and transportation of raw
agricultural goods will be minimal and excluded it from this analysis.

-------
As described earlier, the agencies estimated the impact of the final rules on upstream
using the downstream fuel consumption reductions predicted by MOVES for vocational vehicles
and tractor-trailers. For HD pickups and vans, separate analyses of estimating the emissions
from upstream processes were conducted using the fuel consumption estimates from DOT's
CAFE model and EPA's MOVES model, using Method A and B, respectively.
5.3.4 Calculation of HFC EmissionsF
EPA is adopting new air conditioning (A/C) leakage standards for vocational vehicles to
reduce HFC emissions. The Vintaging Model, developed by EPA Office of Atmospheric
programs, produces HFC inventories for several categories of stationary and mobiles sources.
However, it does not include air conditioning systems in medium and heavy-duty trucks within
its inventory calculations. For this final rulemaking, we conducted an analysis based on the
inputs to the Vintaging Model and the inputs to the MOVES analysis discussed in Chapter 5.3.2
above.
The general equation for calculating HFC emissions follows:
HFC emissionsYearx = A/C Systemsyearx Average Charge Size x HFC loss rate
We determined the number of functioning A/C systems in each year based on the
projected sales of vehicles, the fraction of vehicles with A/C systems, and the average lifetime of
an A/C system. Sales were drawn from the MOVES analysis and we assumed that every vehicle
had a functioning A/C system when sold. The Vintaging Model assumes that all light-duty
passenger vehicle A/C systems (in the U.S.) last exactly 12 years.0 In the absence of other
information, we assumed that heavy-duty vehicles A/C systems last for the same period of time
as light-duty vehicles. Light, medium and heavy-duty vehicles use largely the same components
in their air conditioning systems (sometimes from the same suppliers), which will indicate
similar periods of durability.
The charge size was determined using the Minnesota refrigerant leakage database.21
EPA sorted the data based on A/C charge size and evaluated only the largest 25 percent of A/C
systems to be more representative of HD systems. The average charge size is 1,025 grams of
refrigerant.
Due to the similarity in system design, we assumed that the light-duty vehicle emission
rate in the Vintaging Model was applicable to the current analysis, as shown in Table 5-23. The
Vintaging Model assumes that losses occur from three events: leak, service, and disposal.
Although vehicle A/C systems are serviced during discrete events and not usually every year,
emissions from those events are averaged over the lifetime of the A/C system in the Vintaging
model. Leak and service emissions are considered "annual losses" and are applied every year;
F The U.S. has submitted a proposal to the Montreal Protocol which, if adopted, would phase-down production and
consumption of HFCs.
G This is in agreement with the IPCC report IPCC/TEAP 2005 Safeguarding the Ozone Layer and the Global
Climate System - Issues Related to Hydrofluorocarbons and Perfluorocarbons, which indicates lifetimes
(worldwide) of 9 to 12 years.

-------
disposal is considered an "end of life loss" and is applied only once for each vintage of
vehicles.H
Table 5-23 Annual In-use Vehicle HFC134a Emission Rate from Vintaging Model
KIND OF LOSS
LOSS FRACTION
Leakage
8%
Maintenance /Servicing
10%
End of Life
43%
The Vintaging Model assumes that charge loss is replaced every year; i.e., assuming an
18 percent rate of charge loss, a vehicle with a charge of 1,000 grams will lose a constant rate of
180 grams per year. While this loss rate is not representative of any single given vehicle, it is
assumed accurate for the fleet as a whole. Other emissions, such as fugitive emissions at a
production facility, leaks from cylinders in storage, etc., are not explicitly modeled, but such
emissions are accounted for within the average annual loss rate.
EPA's analysis of the Minnesota database of MY 2010 vehicles suggests that many of the
modeled vehicles likely contain some of the technology required to meet the leakage standard,
and as a consequence are leaking less. We assume that these improvements are independent of
EPA regulation, rather than a preemptive response to regulation. Consequently, this rulemaking
does not take credit for these emission reductions. EPA also reviewed a study conducted by the
Eastern Research Group (ERG) of R134a leaks in heavy-duty vehicles to California Air
Resources Board.22 The study included a total of 70 medium- and heavy-duty vehicles and off-
road equipment, of which 18 of the samples were HD tractors ranging between 1990 and 2008
model years. The mobile air conditioning capacity in the tractors ranged between 1,080 grams to
1,950 grams. The study measured HFC leakage during sample times which ranged between 0.3
and 0.6 years. ERG then calculated an annualized in-use leakage rate with an assumed linear
projection of measured leak rates to annual leak rates, which may be an over-estimate. The
annualize leakage rate for tractors ranged between nearly 0 to nearly 1.5 grams leakage per gram
of MAC capacity. These leakage rates did not include other leakage sources such as
maintenance or end of life recovery. ERG found that the average of all MD and HD trucks and
equipment which were 2006 MY or newer had an average leakage of 103 grams of R134a per
year. Based on these results, the agency believes that our estimates for HFC reductions may
understate the benefits of the program. The agency will continue to analyze this and other
studies that may be conducted in the future.
Based on the Minnesota database, we determined that it is possible to reduce the HFC
emissions from these vehicles on average by 13 percent. EPA calculated this based on the
assumption that vehicles currently in the fleet which meet the MY 2021 standard will not make
any additional improvements to reduce leakage. We also assumed that the systems which
currently have leakage rates above the standard will reduce their leakage to the level of the
standard. We then applied the 13 percent reduction to the baseline 18 percent leakage rate to
H The U.S. EPA has reclamation requirements for refrigerants in place under Title VI of the Clean Air Act.

-------
develop a 15.6 percent leakage rate for MY 2021 and later vehicles to determine the reduction in
emission rate which should be credited to this rulemaking.1
We calculated our emission reductions based on the difference between the baseline case
of 2010 vehicle technology (discussed above) and the control scenario where the loss prevention
technology has been applied to 100 percent of the new vocational vehicles starting in 2021
model year, as will be required by the standards.
Total HFC reductions are 179 metric tons over the MY 2021 baseline AJC system in
2040 and 220 metric tons in 2050. This is equivalent to a reduction of 256,061 metric tons of
CCheq emissions in 2040; and 314,930 metric tons CCheq in 2050.J
5.3.5 Development of Onroad Emission Inventories for Air Quality Modeling
This section summarizes the onroad emission inventories that were used to create
emissions inputs to the air quality modeling described in Chapter 6.2 of the RIA. Details on the
development of emission inventories for sectors other than onroad, as well as additional
information on the methodologies for producing onroad inventories for air quality modeling, are
provided in the Emission Inventories for Air Quality Modeling Technical Support Document,
which can be found in the docket for this rulemaking.23
The emission inventories for air quality modeling requires estimating the inventories for
the entire U.S. by 12 km grid cell and hour of the day for each day of the year, involving a
methodology with much greater detail than the national emission inventories discussed above. In
addition to the methodological differences, due to the long lead time needed to do the air quality
runs, differences exist in the modeling tools and inputs used for the national inventories and air
quality modeling, and in essence, they are separate analyses.
Because using this modeling methodology with added precision is time-consuming and
resource-intensive, the inventories for air quality modeling were developed using an earlier
version of MOVESK than what was used for the national inventories. The series of updates in
MOVES that were implemented since the NPRM, described in Chapter 5.3.1 of the RIA, were
not included in the air quality modeling version of MOVES. Additional details on the
differences between the two versions are documented in the memorandum to the docket.8 The
MOVES model used to generate the inventories for air quality modeling can also be found in the
docket.9
Furthermore, the model inputs used to generate the inventories for air quality modeling
differ from the ones for national inventories. Because the development of air quality inventories
had to be started prior to receiving the comments from the proposal, the flat baseline (Alternative
1 Using 18 percent as the base emission rate may overstate the net emission reductions. However, numbers from the
ERG Report to CARB studying the leakage rate of heavy-duty vehicles are actually much larger (range of near 0 to
150 percent annually), and this places an 18 percent annual loss rate well within the literature. However, (a) the net
impact is very small, (b) these numbers have significant uncertainty, and (c) it is unclear what the appropriate
modification would be.
1 Using a Global Warming Potential of 1,430 for HFC-134a.
K A revised version of MOVES2014 was used to develop the inventories for air quality modeling
(MOVES20150507 code andMOVESDB20150515).

-------
la) for the air quality inventories assumed the APU adoption rate of 30 percent, instead of the 9
percent assumed in the national inventories based on public comments. For modeling of the
Phase 2 standards, we used the projected technological improvements, such as the improvements
in engine and vehicle efficiency, aerodynamic drag, and tire rolling resistance, from the proposal
(Alternative 3). Also, the inventories for air quality modeling assumed higher projected use of
APUs to meet the Phase 2 standards than the national inventories (Figure 5-4). Lastly, the
additional PM2.5 control on APUs being required in the final rules was not modeled in the air
quality inventories. Chapter 5.5.2.3 of this RIA presents the differences between the air quality
and final national inventories.
The onroad mobile source emission inventories were generated for two calendar years,
2011 and 2040, using Method B.L The emission inventories for 2011 were developed to provide
a base year for forecasting future air quality. Calendar year 2040 was run for both the flat
baseline (Alternative la)M and the preferred alternative (Alternative 3) from the proposal. The
meteorological data used to develop and temporally allocate emissions for both 2011 and 2040
were consistent with the 2011 data used for the air quality modeling. In addition, the inventories
for air quality modeling accounted for the county-specific information on vehicle populations,
VMT, age distributions, and inspection-maintenance programs, as well as the anti-idling
mandates, such as the one in California.
5.4 Greenhouse Gas Emission and Fuel Consumption Impacts
The following subsections summarize two slightly different analyses of the annual GHG
emissions and fuel consumption reductions expected from the final standards, as well as the
reductions in GHG emissions and fuel consumption expected over the lifetime of each heavy-
duty vehicle category. Chapter 5.4.1 shows the impacts of the final rules on fuel consumption
and GHG emissions using the MOVES model for tractor-trailers and vocational vehicles, and the
DOT's CAFE model for HD pickups and vans (Method A), relative to two different reference
cases - flat and dynamic. Chapter 5.4.2 shows the impacts of the final standards, relative to the
flat reference case only, using the MOVES model for all heavy-duty vehicle categories.
5.4.1 Impacts of the Final Rules using Analysis Method A
5.4.1.1 Calendar Year Analysis
5.4.1.1.1 Downstream Impacts
As described in Section VILA of the FRM Preamble, for the analysis using Method A,
NHTSA used MOVES to estimate downstream GHG inventories from the final rules for
vocational vehicles and tractor-trailers. For HD pickups and vans, DOT's CAFE model was
used.
The following two tables summarize NHTSA's estimates of HD pickup and van fuel
consumption and GHG emissions under the current standards defining the No-Action and final
L For an explanation of analytical Methods A and B, please see Section I.D of the Preamble.
M For an explanation of the flat baseline, la, and dynamic baseline, lb, please see Section X.A. 1 of the Preamble.

-------
program, respectively, using Method A. Table 5-24 shows results assuming manufacturers will
voluntarily make improvements that pay back within six months (i.e., Alternative lb). Table
5-25 shows results assuming manufacturers will only make improvements as needed to achieve
compliance with standards (i.e., Alternative la). While underlying calculations are all performed
for each calendar year during each vehicle's useful life, presentation of outcomes on a model
year basis aligns more clearly with consideration of cost impacts in each model year, and with
consideration of standards specified on a model year basis. In addition, Method A analyzes
manufacturers' potential responses to HD pickup and van standards on a model year basis
through 2030, and any longer-term costs presented in today's notice represent extrapolation of
these results absent any underlying analysis of longer-term technology prospects and
manufacturers' longer-term product offerings.
Table 5-24 Estimated Fuel Consumption and GHG Emissions over Useful Life of HD Pickups and Vans
Produced in Each Model Year for Method A, Relative to Alternative lb a
MODEL
YEAR
FUEL CONSUMPTION (B. GAL.)
OVER FLEET'S USEFUL LIFE
GHG EMISSIONS (MMT C02EQ)
OVER FLEET'S USEFUL LIFE
No Action
Final
Reduction
No Action
Final
Reduction
2016
10.4
10.4
0.0%
127
127
0.0%
2017
10.4
10.2
2.0%
127
124
2.0%
2018
10.5
10.2
2.9%
127
124
2.9%
2019
10.1
9.60
4.8%
123
117
4.8%
2020
10.1
9.60
4.6%
123
117
4.6%
2021
9.82
9.17
6.6%
120
112
6.5%
2022
9.67
9.01
6.9%
118
110
6.8%
2023
9.64
8.97
7.0%
117
109
6.9%
2024
9.67
9.00
7.0%
118
110
6.9%
2025
9.79
8.98
8.3%
119
109
8.2%
2026
9.91
8.90
10.2%
121
109
10.1%
2027
9.89
8.84
10.7%
120
108
10.5%
2028
10.0
8.89
11.1%
122
108
10.9%
2029
10.1
8.97
11.2%
123
109
11.1%
2030
10.1
8.94
11.2%
123
109
11.1%
Note:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline,
la, and dynamic baseline, lb, please see Section X. A. 1.

-------
Table 5-25 Estimated Fuel Consumption and GHG Emissions over Useful Life of HD Pickups and Vans
Produced in Each Model Year for Method A, Relative to Alternative laa
MODEL
YEAR
FUEL CONSUMPTION (B. GAL.)
OVER FLEET'S USEFUL LIFE
GHG EMISSIONS (MMT C02EQ)
OVER FLEET'S USEFUL LIFE
No Action
Final
Reduction
No Action
Final
Reduction
2016
10.43
10.43
0.0%
122
122
0.0%
2017
10.37
10.15
2.2%
122
119
2.2%
2018
10.41
10.10
3.0%
122
118
3.1%
2019
10.04
9.55
4.9%
118
112
5.1%
2020
10.03
9.56
4.7%
118
112
4.9%
2021
9.84
9.16
6.9%
115
107
7.1%
2022
9.74
9.01
7.5%
114
105
7.7%
2023
9.71
8.97
7.6%
114
105
7.8%
2024
9.75
9.00
7.6%
114
105
7.8%
2025
9.88
8.97
9.1%
116
105
9.3%
2026
10.00
8.92
10.8%
117
104
11.1%
2027
10.01
8.84
11.7%
117
103
11.9%
2028
10.12
8.89
12.1%
119
104
12.4%
2029
10.22
8.98
12.1%
120
105
12.4%
2030
10.18
8.95
12.2%
119
105
12.4%
Note:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline,
la, and dynamic baseline, lb, please see Section X. A. 1.
To more clearly communicate these trends visually, the following two charts present the
above results graphically for Method A, relative to Alternative lb. As shown, fuel consumption
and GHG emissions follow parallel though not precisely identical paths. Though not presented,
the charts for Alternative la will appear sufficiently similar that differences between Alternative
la and Alternative lb remain best communicated by comparing values in the above tables.

-------
11
10
9
8
7
6
5
4
3
2
1
0
5-:
Fuel (No-Action)
Fuel (Standards)
ID
r-N
00

O
rH
CM
ro

LD
KD
r*s
00
G)
O
rH
rH
rH
rH
CM
(N
(N
CM
CM
(N
fM
fM
fM
CM
ro
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
CM
CM
CM
CM
CM
CM
CM
CM
fM
fM

-------
140
130
120
110
100
90
80
70
60
50
40
30
20
10
0
e 5-2
GHG (No-Action)
GHG (Standards)

r*s
00

O
H

\D
rv
00
G)
O
rH
rH
rH
rH

-------
Table 5-26 Annual Downstream GHG Emissions Impacts in Calendar Years 2025,2040 and 2050 by Heavy-
Duty Vehicle Category - Final Program vs. Alt lb using Analysis Method Aa
CY
VEHICLE
C02
ch4
N20
TOTAL DOWNSTREAM

CATEGORY
(MMT)
(MMT
(MMT





C02EQ)
CO2EQ)
MMT CO2EQ
% CHANGE
2025
HD Pickups and Vans
-4.3
0.0005
0.001
-4.3
-4.8%

Vocational
-4.3
0.0001
0
-4.3
-4.1%

Tractor-Trailers
-17.9
-0.005
0.0006
-17.9
-5.1%

Total
-26.5
-0.004
0.002
-26.6
-4.9%
2040
HD Pickups and Vans
-9.7
0.002
0.005
-9.7
-10.0%

Vocational
-18.1
0
0.0003
-18.1
-15.0%

Tractor-Trailers
-75.5
-0.02
0.001
-75.5
-19.0%

Total
-103.3
-0.02
0.006
-103.3
-17.0%
2050
HD Pickups and Vans
-10.7
0.002
0.006
-10.7
-11.0%

Vocational
-21.2
0
0.0003
-21.2
-16.0%

Tractor-Trailers
-91.9
-0.03
0.001
-91.9
-21.0%

Total
-123.8
-0.03
0.007
-123.8
-18.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 5-27 Annual Downstream GHG Emissions Impacts in Calendar Years 2025,2040 and 2050 by Heavy-
Duty Vehicle Category - Final Program vs. Alt la using Analysis Method Aa
CY
VEHICLE
O
O
ch4
N20
TOTAL DOWNSTREAM

CATEGORY
(MMT)
(MMT
(MMT





CO2EQ)
CO2EQ)
MMT CO2EQ
% CHANGE
2025
HD Pickups and Vans
-4.7
0.0005
0.002
-4.7
-5.2%

Vocational
-4.3
0.0001
0.0001
-4.3
-4.1%

Tractor-Trailers
-19.9
-0.006
0.0006
-19.9
-5.7%

Total
-28.9
-0.005
0.003
-28.9
-5.3%
2040
HD Pickups and Vans
-10.6
0.002
0.005
-10.6
-11.2%

Vocational
-18.1
0
0.0003
-18.1
-14.9%

Tractor-Trailers
-85.4
-0.02
0.001
-85.4
-21.3%

Total
-114.1
-0.02
0.006
-114.1
-18.5%
2050
HD Pickups and Vans
-11.7
0.002
0.006
-11.7
-11.7%

Vocational
-21.2
-0.0001
0.0003
-21.2
-16.1%

Tractor-Trailers
-104.0
-0.03
0.001
-104.0
-23.0%

Total
-136.9
-0.03
0.007
-136.9
-20.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
Table 5-28 Annual Fuel Savings in Calendar Years 2025,2040 and 2050 by Heavy-Duty Vehicle Category -
Final Program vs. Alt lb using Analysis Method Aa
CY
VEHICLE CATEGORY
DIESEL
GASOLINE
BILLION
GALLONS
% SAVINGS
BILLION
GALLONS
% SAVINGS
2025
HD Pickups and Vans
0.2
4.0%
0.3
5.5%
Vocational
0.3
4.1%
0.1
3.8%
Tractor-Trailers
1.8
5.4%
0
0%
Total
2.3
4.9%
0.4
5.0%
2040
HD Pickups and Vans
0.3
8.3%
0.7
12.0%
Vocational
1.5
15.0%
0.3
13.0%
Tractor-Trailers
7.4
19.0%
0
0%
Total
9.2
17.8%
1.0
12.2%
2050
HD Pickups and Vans
0.4
8.7%
0.8
13.0%
Vocational
1.7
17.0%
0.4
13.0%
Tractor-Trailers
9.0
21.0%
0
0%
Total
11.1
19.3%
1.2
12.8%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 5-29 Annual Fuel Savings in Calendar Years 2025,2040 and 2050 by Heavy-Duty Vehicle Category-
Final Program vs. Alt la using Analysis Method Aa
CY
VEHICLE CATEGORY
DIESEL
GASOLINE
BILLION
GALLONS
% SAVINGS
BILLION
GALLONS
% SAVINGS
2025
HD Pickups and Vans
0.2
3.8%
0.4
6.2%
Vocational
0.3
4.1%
0.1
3.8%
Tractor-Trailers
1.9
5.7%
0
0%
Total
2.4
5.2%
0.5
5.5%
2040
HD Pickups and Vans
0.3
8.6%
0.8
13.0%
Vocational
1.5
15.5%
0.4
12.8%
Tractor-Trailers
8.4
21.3%
0
0%
Total
10.2
19.0%
1.2
13.0%
2050
HD Pickups and Vans
0.4
9.0%
0.9
14.0%
Vocational
1.7
16.7%
0.4
13.5%
Tractor-Trailers
10.2
23.0%
0
0%
Total
12.3
21.0%
1.3
14.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
5.4.1.1.1
Upstream Impacts
Table 5-30 Annual Upstream GHG Emissions Impacts in Calendar Years 2025,2040 and 2050 by Heavy-
Duty Vehicle Category - Final Program vs. Alt lb using Analysis Method Aa
CY
VEHICLE
CATEGORY
C02
(MMT)
ch4
(MMT
C02EQ)
N20
(MMT
CO2EQ)
TOTAL UPSTREAM
MMT CO2EQ
% CHANGE
2025
HD Pickups and
Vans
-1.1
-0.2
-0.04
-1.3
-4.8%
Vocational
-1.3
-0.1
-0.006
-1.4
-4.1%
Tractor-Trailers
-5.7
-0.6
-0.03
-6.3
-5.1%
Total
-8.1
-0.9
-0.08
-9.0
-4.9%
2040
HD Pickups and
Vans
-2.4
-0.4
-0.1
-2.9
-10.0%
Vocational
-5.4
-0.6
-0.03
-6.0
-15.0%
Tractor-Trailers
-24.0
-2.4
-0.1
-26.5
-19.0%
Total
-31.8
-3.4
-0.2
-35.5
-17.0%
2050
HD Pickups and
Vans
-2.6
-0.5
-0.1
-3.2
-11.0%
Vocational
-6.3
-0.7
-0.03
-7.0
-16.0%
Tractor-Trailers
-29.2
-3.0
-0.1
-32.3
-21.0%
Total
-38.1
-4.2
-0.2
-42.5
-19.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
Table 5-31 Annual Upstream GHG Emissions Impacts in Calendar Years 2025,2040 and 2050 by Heavy-
Duty Vehicle Category - Final Program vs. Alt la using Analysis Method Aa
CY
VEHICLE CATEGORY
O
O
ch4
N20
TOTAL UPSTREAM


(MMT)
(MMT
(MMT





C02EQ)
CO2EQ)
MMT CO2EQ
% CHANGE
2025
HD Pickups and Vans
-1.1
-0.2
-0.05
-1.4
-5.2%

Vocational
-1.3
-0.1
-0.01
-1.4
-4.1%

Tractor-Trailers
-6.3
-0.6
-0.03
-7.0
-5.7%

Total
-8.7
-0.9
-0.09
-9.8
-5.3%
2040
HD Pickups and Vans
-2.6
-0.5
-0.1
-3.2
-11.0%

Vocational
-5.4
-0.6
-0.03
-6.0
-15.1%

Tractor-Trailers
-27.2
-2.8
-0.1
-30.1
-21.3%

Total
-35.2
-3.9
-0.2
-39.3
-19.0%
2050
HD Pickups and Vans
-2.8
-0.5
-0.1
-3.5
-12.0%

Vocational
-6.3
-0.7
-0.03
-7.0
-16.3%

Tractor-Trailers
-33.1
-3.4
-0.2
-36.7
-23.0%

Total
-42.2
-4.6
-0.3
-47.2
-20.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
5.4.1.1.1	HFC Impacts
The projected HFC emission reductions due to the AJC leakage standards are estimated to
be 86,735 metric tons of CCheq in 2025, 256,061 metric tons of CCheq in 2040, and 314,930
metric tons CCheq in 2050.
5.4.1.1.2	Total (Downstream + Upstream + HFC) Impacts
Table 5-32 Annual Total GHG Emissions Impacts in Calendar Years 2025,2040 and 2050 - Final Program
vs. Alt lb using Analysis Method Aa

CY2025
CY204
0
CY2050
MMT
C02eq
% Change
MMT C02eq
%
Change
MMT C02eq
% Change
Downstream
-26.6
-4.9%
-103.3
-17.0%
-123.8
-18.0%
Upstream
-9.0
-4.9%
-35.5
-17.0%
-42.5
-19.0%
HFC
-0.1
-15.0%
-0.3
-13.0%
-0.3
-13.0%
Total
-35.7
-4.9%
-139.1
-17.0%
-166.6
-19.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
Table 5-33 Annual Total GHG Emissions Impacts in Calendar Years 2025,2040 and 2050 - Final Program
vs. Alt la using Analysis Method Aa

CY2025
CY204
0
CY2050
MMT
CC^eq
% Change
MMT C02eq
%
Change
MMT C02eq
% Change
Downstream
-28.9
-5.3%
-114.1
-19.0%
-136.9
-20.0%
Upstream
-9.8
-5.3%
-39.3
-19.0%
-47.2
-20.0%
HFC
-0.1
-15.0%
-0.3
-13.0%
-0.3
-13.0%
Total
-38.8
-5.3%
-153.7
-19.0%
-184.4
-20.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
5.4.1.1 Model Year Lifetime Analysis
Table 5-34 Lifetime GHG Reductions and Fuel Savings by Heavy-Duty Vehicle Category - Summary for
Model Years 2018-2029 using Analysis Method Aa

FINAL PROGRAM

(ALTERNATIVE 3)
NO-ACTION ALTERNATIVE
(BASELINE)
lb (Dynamic)
la (Flat)
Fuel Savings (Billion Gallons)
71.1
77.7
HD Pickups and Vans
9.0
9.8
Vocational
12.4
12.3
Tractor/Trailers
49.7
55.6
Total GHG Reductions (MMT C02eq)
958
1,049
HD Pickups and Vans
111
120
Vocational
162
162
Tractor/Trailers
685
767
Note:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline,
la, and dynamic baseline, lb, please see Section X. A. 1.
5.4.2 Impacts of the Final Rules using Analysis Method B
5.4.2.1 Calendar Year Analysis
5.4.2.1.1 Downstream Impacts
After all the MOVES runs and post-processing were completed, the flat reference
(Alternative la) and control case (Alternative 3) inventories were totaled for all heavy-duty
vehicle types and emission processes to estimate total downstream GHG and fuel consumption
impacts of the program.

-------
The fuel savings from the final rules were calculated from the estimates of total energy
consumption from MOVES using the fuel heating values assumed in the Renewable Fuels
Standard rulemakingN and in MOVES.0
Table 5-35 summarizes these downstream GHG impacts in calendar years 2025, 2040,
and 2050, relative to Alternative la, for the final program. Table 5-36 shows the estimated fuel
savings from the final program in 2025, 2040, and 2050, relative to Alternative la. The
reductions in CO2 emissions result from all heavy-duty vehicle categories (including the engines
associated with tractor-trailer combinations and vocational vehicles) due to engine and vehicle
improvements. N2O emissions show a very slight increase because of a rebound in vehicle miles
traveled (VMT). However, since N2O is produced as a byproduct of fuel combustion, the
increase in N2O emissions is expected to be more than offset by the improvements in fuel
efficiency from the final rules.p The methane emissions decrease primarily due to differences in
hydrocarbon emission characteristics between on-road diesel engines and APUs. The amount of
methane emitted as a fraction of total hydrocarbons is expected to be less for APUs than for
diesel engines. Overall, downstream GHG emissions will be reduced significantly. In addition,
substantial fuel savings will be achieved from improved fuel efficiency. All emissions impacts
reflect the heavy-duty sector only, and do not include emissions from light-duty vehicles or any
other vehicle sector.
N Renewable Fuels Standards assumptions of 115,000 BTU/gallon gasoline (E0) and 76,330 BTU/gallon ethanol
(E100) were weighted 90 percent and 10 percent, respectively, forElO and 85 percent and 15 percent, respectively,
for E15 and converted to kJ at 1.055 kJ/BTU. The conversion factors are 117,245 kJ/gallon for gasoline blended
with ten percent ethanol (E10) and 115,205 kJ/gallon for gasoline blended with fifteen percent ethanol (E15).
0 The conversion factor for diesel is 138,451 kJ/gallon. See MOVES2004 Energy and Emission Inputs. EPA420-P-
05-003, March 2005. http://www3.epa.gov/otaq/models/ngm/420p05003.pdf.
p MOVES is not capable of modeling the changes in exhaust N20 emissions from the improvements in fuel
efficiency. Due to this limitation, a conservative approach was taken to only model the VMT rebounds in estimating
the emissions impact on N2O from the final rules, resulting in a slight increase in downstream N20 inventory.

-------
Table 5-35 Annual Downstream GHG Emissions Impacts in Calendar Years 2025,2040 and 2050 by Heavy-
Duty Vehicle Category - Final Program vs. Alt la using Analysis Method B a
CY
VEHICLE
CATEGORY
C02
(MMT)
ch4
(MMT
C02EQ)
N20
(MMT
CO2EQ)
TOTAL DOWNSTREAM
MMT CO2EQ
% CHANGE
2025
HD Pickups and
Vans
-3.6
0.0004
0.001
-3.6
-2.5%
Vocational
-4.3
0.0001
0.0001
-4.3
-4.1%
Tractor-Trailers
-19.9
-0.006
0.0006
-19.9
-5.7%
Total
-27.8
-0.005
0.002
-27.8
-4.6%
2040
HD Pickups and
Vans
-20.9
0.001
0.002
-20.8
-13.6%
Vocational
-18.1
0
0.0003
-18.1
-14.9%
Tractor-Trailers
-85.4
-0.02
0.001
-85.4
-21.3%
Total
-124.3
-0.02
0.004
-124.3
-18.4%
2050
HD Pickups and
Vans
-23.2
0.001
0.003
-23.2
-14.8%
Vocational
-21.2
-0.0001
0.0003
-21.2
-16.0%
Tractor-Trailers
-104.0
-0.03
0.001
-104.0
-23.0%
Total
-148.4
-0.03
0.004
-148.4
-20.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
Table 5-36 Annual Fuel Savings in Calendar Years 2025,2040 and 2050 by Heavy-Duty Vehicle Category -
Final Program vs. Alt la using Analysis Method B a
CY
VEHICLE
CATEGORY
DIESEL
GASOLINE


BILLION
GALLONS
% SAVINGS
BILLION
GALLONS
% SAVINGS
2025
HD Pickups and
Vans
0.2
2.6%
0.2
2.5%

Vocational
0.3
4.1%
0.1
3.8%

Tractor-Trailers
1.9
5.7%
0
0%

Total
2.5
5.0%
0.3
2.8%
2040
HD Pickups and
Vans
0.9
13.9%
1.3
13.5%

Vocational
1.5
15.5%
0.4
12.8%

Tractor-Trailers
8.4
21.3%
0
0%

Total
10.8
19.4%
1.7
13.3%
2050
HD Pickups and
Vans
1.1
15.0%
1.5
14.7%

Vocational
1.7
16.7%
0.4
13.5%

Tractor-Trailers
10.2
23.0%
0
0%

Total
13.0
21.0%
1.9
14.4%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
5.4.2.1.2 Upstream Impacts
The upstream GHG impacts of final program associated with the production and
distribution of gasoline and diesel from crude oil, relative to Alternative la, are summarized in
Table 5-37, for calendar years 2025, 2040, and 2050. These estimates show impacts for
domestic emission reductions only. Additionally, since this rulemaking is not expected to impact
biofuel volumes mandated by the Annual Renewable Fuel Standards (RFS) regulations, the
impacts on upstream emissions from changes in biofuel feedstock (i.e., agricultural sources such
as fertilizer, fugitive dust, and livestock) are not included. In other words, we attribute decreased
fuel consumption from this program to petroleum-based fuels only, while assuming no net effect
on volumes of renewable fuels. We used this approach because annual renewable fuel volumes
are mandated independently from this rulemaking under RFS. As a consequence, it is not
possible to conclude whether the decreasing petroleum consumption projected here would
increase the fraction of the U.S. fuel supply that is made up by renewable fuels (if RFS volumes
remained constant), or whether future renewable fuel volume mandates would decrease in
proportion to the decreased petroleum consumption projected here.
As background, EPA sets annual renewable fuel volume mandates through a separate
RFS notice-and-comment rulemaking process, and the final volumes are based on EIA
projections, EPA's own market assessment, and information obtained from the RFS notice and
comment process. Also, RFS standards are nested within each other, which means that a fuel

-------
with a higher GHG reduction threshold can be used to meet the standards for a lower GHG
reduction threshold. This creates additional uncertainty in projecting this rule's net effect on
future annual RFS standards.
In conclusion, the impacts of this rulemaking on annual renewable fuel volume mandates
are difficult to project at the present time. However, since it is not centrally relevant to the
analysis for this rulemaking, we have not included any impacts on renewable fuel volumes in this
analysis. The reductions in upstream GHGs are proportional to the amount of fuel saved.
Table 5-37 Annual Upstream GHG Emissions Impacts in Calendar Years 2025,2040 and 2050 by Heavy-
Duty Vehicle Category - Final Program vs. Alt la using Analysis Method B a
CY
VEHICLE
CATEGORY
C02
(MMT)
ch4
(MMT
C02EQ)
N20
(MMT
CO2EQ)
TOTAL UPSTREAM
MMT CO2EQ
% CHANGE
2025
HD Pickups and
Vans
-1.0
-0.1
-0.01
-1.1
-2.6%
Vocational
-1.3
-0.1
-0.01
-1.4
-4.1%
Tractor-Trailers
-6.3
-0.6
-0.03
-7.0
-5.7%
Total
-8.6
-0.9
-0.04
-9.5
-4.7%
2040
HD Pickups and
Vans
-5.4
-0.7
-0.03
-6.1
-13.7%
Vocational
-5.4
-0.6
-0.03
-6.0
-15.1%
Tractor-Trailers
-27.2
-2.8
-0.1
-30.1
-21.3%
Total
-38.0
-4.0
-0.2
-42.2
-18.7%
2050
HD Pickups and
Vans
-6.1
-0.8
-0.03
-6.8
-14.9%
Vocational
-6.3
-0.7
-0.03
-7.0
-16.3%
Tractor-Trailers
-33.1
-3.4
-0.2
-36.7
-23.0%
Total
-45.5
-4.8
-0.2
-50.5
-20.3%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
5.4.2.1.3	HFC Impacts
Based on projected HFC emission reductions due to the AC leakage standards, EPA
estimates the HFC reductions to be 86,735 metric tons of CCheq in 2025, 256,061 metric tons of
CCheq in 2040, and 314,930 metric tons CCheq in 2050.
5.4.2.1.4	Total (Downstream + Upstream + HFC) Impacts
The combined annual GHG emissions reductions of final program from downstream,
upstream, and HFC, relative to Alternative la, are summarized in Table 5-38 for calendar years
2025, 2040 and 2050.

-------
Table 5-38 Annual Total GHG Emissions Impacts in Calendar Years 2025,2040 and 2050 - Final Program
vs. Alt la using Analysis Method B a

CY2025
CY204
0
CY2050
MMT
CCheq
% Change
MMT C02eq
%
Change
MMT C02eq
% Change
Downstream
-27.8
-4.6%
-124.3
-18.4%
-148.4
-20.0%
Upstream
-9.5
-4.7%
-42.2
-18.7%
-50.5
-20.3%
HFCh
-0.1
-15.0%
-0.3
-13.0%
-0.3
-13.0%
Total
-37.4
-4.7%
-166.8
-18.5%
-199.2
-20.1%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section XA. 1
b HFC represents HFC emission reductions and percent change from the vocational vehicle category only.
Figure 5-3 graphically illustrates the total annual GHG trends for both Phase 1 and Phase
2 rules, using Method B, for calendar years from 2016 to 2050. The flat baseline from Phase 2
rule is assumed to be equivalent to the Phase 1 program.
Pre-Phase 1	Phase 2 - Flat Baseline — Phase 2 - Dynamic Baseline	Phase 2 Final Program
1200
1100
Phase 1
Reduction
1000
u
^ 900
Phase 2
Reduction
800
700
600
ICNCOOOHNmf^lflrNCOfllOrlNmViniONOOOgrlNfO'finiONCpOlO
ooooooooooooooooooooooooooooooooooo
(N!NNfN(N(MNNtMN(NtMN(N(NtNl(N(N(MfNfM(MNfM(N(N(NNfMM(NPJ(NNfM
Calendar Year
Figure 5-3 Total Annual GHG Trends for Phase 1 and Phase 2 Rule, using Analysis Method B

-------
5.4.2.2 Model Year Lifetime Analysis
In addition to the annual GHG emissions and fuel consumption reductions expected from
the final rules, we estimated the combined (downstream and upstream) GHG and fuel
consumption impacts over the model year lifetimes of the impacted vehicles sold in the
regulatory timeframe. In contrast to the calendar year analysis, the model year lifetime analyses
show the impacts of the program on each of these model year fleets over the course of their
lifetimes. Table 5-39 shows the fleet-wide GHG reductions and fuel savings from the final rules
through the lifetimeQ of heavy-duty vehicles, relative to Alternative la.
Table 5-39 Lifetime GHG Reductions and Fuel Savings by Heavy-Duty Vehicle Category - Summary for
Model Years 2018-2029 using Analysis Method Ba

FINAL PROGRAM
(ALTERNATIVE 3)
NO-ACTION ALTERNATIVE (BASELINE)
la (Flat)
Fuel Savings (Billion Gallons)
82.2
HD Pickups and Vans
14.3
Vocational
12.3
Tractor/Trailers
55.6
Total GHG Reductions (MMT C02eq)
1,097.6
HD Pickups and Vans
169.2
Vocational
161.6
Tractor/Trailers
766.7
Note:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline,
la, and dynamic baseline, lb, please see Section X. A. 1.
Furthermore, the combined lifetime GHG reductions and fuel savings of Phase 1 and
Phase 2 programs are presented in Table 5-40. To be consistent with the emissions modeling
done for this program, the lifetime GHG reductions and fuel savings from Phase 1 were
estimated using the same modeling tools used in the Phase 2 final rulemaking.
Q A lifetime of 30 years is assumed in MOVES.

-------
Table 5-40 Combined Lifetime GHG Reductions and Fuel Savings of Phase 1 and Phase 2 Program using
Analysis Method B a

TOTAL GHG REDUCTIONS
(MMT C02EQ)
FUEL SAVINGS
(BILLION GALLONS)
Phase 1


MY 2014-2018
338
26
MY 2019-2029
1,081
84
Phase 2


MY 2018-2029
1,098
82
Combined Total
2,517
192
Note:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline,
la, and dynamic baseline, lb, please see Section X. A. 1.
5.5 Non-Greenhouse Gas Emission Impacts
The medium- and heavy-duty vehicle standards will influence the emissions of criteria air
pollutants and several air toxics. Similar to Chapter 5.4, the following subsections summarize
two slightly different analyses of the annual non-GHG emissions reductions expected from the
standards. Chapter 5.5.1 shows the impacts of the final rules on non-GHG emissions using the
analytical Method A, relative to two different reference cases - flat and dynamic. Chapter 5.5.2
shows the impacts of the standards, relative to the flat reference case only, using the MOVES
model for all heavy-duty vehicle categories.
5.5.1 Impacts of the Final Rules using Analysis Method A
5.5.1.1 Calendar Year Analysis
5.5.1.1.1 Downstream Impacts

-------
Table 5-41 Annual Downstream Impacts of Heavy-Duty Non-GHG Emissions in Calendar Years 2025,2040
and 2050 - Final Program vs. Alt lb using Analysis Method Aa
POLLUTANT
CY2025
CY2040
CY2050
US Short
Tons
% Change
US Short
Tons
% Change
US Short
Tons
% Change
1,3-Butadiene
1
0.5%
4
3.6%
4
3.4%
Acetaldehyde
-1
0%
-16
-0.7%
-19
-0.8%
Acrolein
0.2
0%
-0.3
-0.1%
-1
-0.4%
Benzene
-2
-0.1%
-13
-1.2%
-13
-1.1%
CO
-9,045
-0.6%
-34,702
-2.8%
-42,095
-3.0%
Formaldehyde
-21
-0.3%
-96
-1.6%
-119
-1.8%
NOx
-12,082
-1.3%
-53,254
-9.1%
-65,068
-9.9%
pm25
-58
-0.2%
-363
-2.0%
-453
-2.2%
SOx
-201
-4.1%
-851
-16.0%
-1,028
-17.0%
voc
-769
-0.8%
-3,436
-5.3%
-4,128
-5.8%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 5-42 Annual Downstream Impacts of Heavy-Duty Non-GHG Emissions in Calendar Years 2025,2040
and 2050 - Final Program vs. Alt la using Analysis Method Aa
POLLUTANT
CY2025
CY2040
CY2050
US Short
Tons
% Change
US Short
Tons
% Change
US Short
Tons
% Change
1,3-Butadiene
1
0.5%
4
3.7%
4
3.5%
Acetaldehyde
-1
0%
-14
-0.7%
-18
-0.8%
Acrolein
0.2
0%
-0.3
-0.1%
-1
-0.4%
Benzene
-2
-0.2%
-13
-1.2%
-14
-1.2%
CO
-8,944
-0.6%
-34,502
-2.8%
-41,880
-3.0%
Formaldehyde
-20
-0.3%
-91
-1.6%
-113
-1.7%
NOx
-13,368
-1.5%
-60,594
-10.2%
-74,206
-11.0%
pm25
-78
-0.2%
-473
-2.6%
-591
-2.9%
SOx
-219
-4.5%
-941
-17.0%
-1,138
-19.0%
voc
-831
-0.8%
-3,736
-5.8%
-4,499
-6.3%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
5.5.1.1.2 Upstream Impacts
Table 5-43 Annual Upstream Impacts of Heavy-Duty Non-GHG Emissions in Calendar Years 2025,2040 and
2050 - Final Program vs. Alt lb using Analysis Method Aa
POLLUTANT
CY2025
CY2040
CY2050
US Short
Tons
% Change
US Short
Tons
% Change
US Short
Tons
% Change
1,3-Butadiene
-1
-4.9%
-4
-18.0%
-5
-19.0%
Acetaldehyde
-3
-4.4%
-14
-15.0%
-16
-16.0%
Acrolein
-0.4
-4.6%
-2
-16.0%
-2
-17.0%
Benzene
-23
-4.8%
-88
-16.0%
-105
-18.0%
CO
-3,785
-4.9%
-14,714
-17.0%
-17,629
-19.0%
Formaldehyde
-18
-4.9%
-71
-17.0%
-86
-19.0%
NOx
-9,255
-4.9%
-35,964
-17.0%
-43,089
-19.0%
pm25
-975
-4.9%
-3,850
-18.0%
-4,618
-19.0%
SOx
-5,804
-4.9%
-22,550
-17.0%
-27,019
-19.0%
voc
-4,419
-4.8%
-14,857
-15.0%
-17,385
-16.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 5-44 Annual Upstream Impacts of Heavy-Duty Non-GHG Emissions in Calendar Years 2025,2040 and
2050 - Final Program vs. Alt la using Analysis Method Aa
POLLUTANT
CY2025
CY2040
CY2050
US Short
Tons
% Change
US Short
Tons
% Change
US Short
Tons
% Change
1,3-Butadiene
-1
-5.3%
-4
-20.0%
-5
-21.0%
Acetaldehyde
-4
-4.6%
-15
-16.0%
-17
-17.0%
Acrolein
-0.4
-4.9%
-2
-17.0%
-2
-18.0%
Benzene
-25
-5.1%
-96
-18.0%
-115
-19.0%
CO
-4,142
-5.4%
-16,298
-19.0%
-19,558
-20.0%
Formaldehyde
-20
-5.3%
-79
-19.0%
-95
-20.0%
NOx
-10,124
-5.4%
-39,813
-19.0%
-47,779
-20.0%
pm25
-1,065
-5.3%
-4,258
-19.0%
-5,117
-21.0%
SOx
-6,349
-5.4%
-24,961
-19.0%
-29,958
-20.0%
voc
-4,810
-5.2%
-16,218
-16.0%
-19,004
-17.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
5.5.1.1.3 Total Impacts
Table 5-45 Annual Total Impacts (Upstream and Downstream) of Heavy-Duty Non-GHG Emissions in
Calendar Years 2025,2040 and 2050 - Final Program vs. Alt lb using Analysis Method Aa
POLLUTANT
CY2025
CY2040
CY2050
US Short
Tons
% Change
US Short
Tons
% Change
US Short
Tons
% Change
1,3-Butadiene
0.3
0.1%
0.1
0.1%
-0.4
-0.3%
Acetaldehyde
-4
-0.1%
-30
-1.3%
-35
-1.4%
Acrolein
-0.2
0%
-2
-0.7%
-3
-0.9%
Benzene
-25
-1.2%
-101
-6.3%
-118
-6.7%
CO
-12,830
-0.9%
-49,416
-3.7%
-59,724
-4.0%
Formaldehyde
-39
-0.5%
-167
-2.7%
-205
-2.9%
NOx
-21,337
-2.0%
-89,218
-11.0%
-108,157
-12.0%
pm25
-1,033
-2.0%
-4,213
-10.0%
-5,071
-11.0%
SOx
-6,005
-4.9%
-23,401
-17.0%
-28,047
-19.0%
voc
-5,188
-2.7%
-18,293
-11.0%
-21,513
-12.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 5-46 Annual Total Impacts (Upstream and Downstream) of Heavy-Duty Non-GHG Emissions in
Calendar Years 2025,2040 and 2050 - Final Program vs. Alt la using Analysis Method Aa
POLLUTANT
CY2025
CY2040
CY2050
US Short
Tons
% Change
US Short
Tons
% Change
US Short
Tons
% Change
1,3-Butadiene
0.2
0.1%
-0.2
-0.1%
-1.0
-0.5%
Acetaldehyde
-5
-0.2%
-29
-1.3%
-35
-1.4%
Acrolein
-0.2
0%
-2
-0.7%
-3
-1.0%
Benzene
-27
-1.4%
-109
-6.8%
-129
-7.2%
CO
-13,086
-0.9%
-50,800
-3.8%
-61,438
-4.1%
Formaldehyde
-40
-0.5%
-170
-2.7%
-208
-2.9%
NOx
-23,492
-2.2%
-100,407
-12.0%
-121,985
-14.0%
pm25
-1,143
-2.2%
-4,731
-12.0%
-5,708
-13.0%
SOx
-6,568
-5.3%
-25,902
-19.0%
-31,096
-20.0%
voc
-5,641
-3.0%
-19,954
-12.0%
-23,503
-13.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
5.5.1.2 Model Year Lifetime Analysis
Table 5-47 Lifetime Non-GHG Reductions by Heavy-Duty Vehicle Category - Summary for Model Years
2018-2029 using Analysis Method A (US Short Tons)a

FINAL PROGRAM

(ALTERNATIVE 3)
NO-ACTION ALTERNATIVE
(BASELINE)
lb (Dynamic)
la (Flat)
NOx
492,070
545,780
HD Pickups and Vans
23,702
26,297
Vocational
42,621
42,621
Tractor/Trailers
425,747
477,021
PM2.5
27,605
30,594
HD Pickups and Vans
2,164
2,385
Vocational
4,436
4,436
Tractor/Trailers
21,005
23,773
SOx
157,579
172,952
HD Pickups and Vans
17,477
19,214
Vocational
25,082
25,082
Tractor/Trailers
115,020
128,656
Note:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline,
la, and dynamic baseline, lb, please see Section X. A. 1.
5.5.2 Impacts of the Final Rules using Analysis Method B
5.5.2.1 Calendar Year Analysis
5.5.2.1.1 Downstream Impacts
After all the MOVES runsR and post-processing were completed, the flat reference
(Alternative la) and control case (Alternative 3) inventories were aggregated for all vehicle types
and emission processes to estimate the total downstream non-GHG impacts of the program.
Table 5-48 summarizes these downstream non-GHG impacts of final program for calendar years
2025, 2040 and 2050, relative to Alternative la. The results are shown both in changes in
absolute tons and in percent reductions from the flat reference to alternatives for the heavy-duty
sector.
The agencies expect the Phase 2 program to impact the downstream emissions of non-
GHG pollutants. These pollutants include oxides of nitrogen (NOx), oxides of sulfur (SOx),
volatile organic compounds (VOC), carbon monoxide (CO), fine particulate matter (PM2.5), and
R For non-GHGs, MOVES was run only for January and July and the annual emissions were extrapolated by scaling
up each month by a factor of 5.88 for all pollutants except particulate matter (PM). For PM, to offset the
disproportionate effect of the cold temperature on January results, a scaling factor of 4.3 was applied to January and
7.5 to July; these factors were determined based on analysis of annual PM emissions during modeling for the RFS2
rule. Note that for GHGs, MOVES was run for all months.

-------
air toxics. The agencies expect reductions in downstream emissions of NOx, PM2.5, VOC, SOx,
CO, and air toxics. Much of these estimated net reductions are a result of the agencies'
anticipation of increased use of auxiliary power units (APUs) in combination tractors during
extended idling; APUs emit these pollutants at a lower rate than on-road engines during extended
idle operation, with the exception of PM2.5. As discussed in Section III.C.3, EPA is adopting
Phase 1 and Phase 2 requirements to control PM2.5 emissions from APUs installed in new
tractors and therefore, eliminate the unintended consequence of increases in PM2.5 emissions
from increased APU use.
The downstream emission reductions of non-GHG pollutants estimated in the final
rulemaking are significantly less than what was estimated for the proposal, mainly because of the
changes in projected use of auxiliary power units (APUs) during extended idling. The idle
reduction adoption rates were reassessed and projected to be lower (Table 5-14) than what was
assumed in the proposal, as described in Section III.D.l.a of the Preamble. Lower penetration of
APUs assumed in the final program results in lower downstream reductions of criteria pollutants
and air toxics, compared to the proposal.
Furthermore, in response to the public comments received on the proposal, the MOVES
emission rates for extended idle were lowered significantly for criteria pollutants based on the
analyses of the latest test programs that reflect the current prevalence of clean idle certified
engines.24 For example, the extended idle rate for NOx was changed from 203 g/hr to 42.6 g/hr
for model year 2013 and later. This change resulted in smaller differences between emission
rates for extended idle and APUs for all criteria pollutants. Therefore, the emissions benefits of
using APUs during extended idle, instead of the main engine, are lower for non-GHGs in the
final rulemaking than the proposal.
Additional reductions in tailpipe emissions of NOx and CO and refueling emissions of
VOC will be achieved through improvements in engine efficiency and reduced road load
(improved aerodynamics and tire rolling resistance), which reduces the amount of work required
to travel a given distance and increases fuel economy.
For vehicle types not affected by road load improvements, such as HD pickups and
vanss, non-GHG emissions will increase very slightly due to VMT rebound. In addition, brake
wear and tire wear emissions of PM2.5 will also increase very slightly due to VMT rebound. The
agencies estimate that downstream emissions of SOx will be reduced, because they are roughly
proportional to fuel consumption.
s HD pickups and vans are subject to gram per mile (distance) emissions standards, as opposed to larger heavy-duty
vehicles which are certified to a gram per brake horsepower (work) standard.

-------
Table 5-48 Annual Downstream Impacts of Heavy-Duty Non-GHG Emissions in Calendar Years 2025,2040
and 2050 - Final Program vs. Alt la using Analysis Method B a
POLLUTANT
CY2025
CY2040
CY2050
US Short
Tons
% Change
US Short
Tons
% Change
US Short
Tons
% Change
1,3-Butadiene
-1
-0.2%
-3
-1.5%
-3
-1.8%
Acetaldehyde
-3
-0.1%
-18
-0.8%
-23
-0.9%
Acrolein
-0.1
0%
-1
-0.3%
-1
-0.4%
Benzene
-5
-0.2%
-22
-1.4%
-26
-1.6%
CO
-9,445
-0.4%
-35,710
-2.4%
-43,642
-2.7%
Formaldehyde
-20
-0.2%
-97
-1.5%
-120
-1.7%
NOx
-13,396
-1.4%
-60,681
-9.7%
-74,362
-10.8%
PM25
-73
-0.2%
-462
-2.2%
-580
-2.5%
SOx
-252
-4.7%
-1,122
-18.5%
-1,341
-20.1%
voc
-1,071
-0.8%
-5,060
-5.9%
-6,013
-6.6%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
As noted above, EPA is adopting Phase 1 and Phase 2 requirements to control PM2.5
emissions from APUs installed in new tractors. In the NPRM, an unintended increase in
downstream PM2.5 emissions was projected because engines powering APUs are currently
required to meet less stringent PM standards (40 CFR 1039.101) than on-road engines (40 CFR
86.007-11) and because the increase in emissions from APUs more than offset the reduced
tailpipe emissions from improved engine efficiency and road load. However, with the new
requirements for APUs, the final program is projected to lead to reduced downstream PM2.5
emissions of 462 tons in 2040 and 580 tons in 2050 (Table 5-48). As shown in Table 5-49, the
net reductions in national PM2.5 emissions with further PM control on APUs are 927 tons and
1,114 tons in 2040 and 2050, respectively. For additional details on EPA's PM emission
standards for APUs, see Section III.C.3 of the Preamble. The development of APU emission
rates with PM control is documented in the memorandum to the docket.25
Table 5-49 Projected Impact on PM2.5 Emissions of Further PM2.5 Control on APUs using Analysis Method
Ba
CY
BASELINE
NATIONAL
HEAVY-DUTY
VEHICLE PM2 5
EMISSIONS
(TONS)
FINAL HD PHASE
2 PROGRAM
NATIONAL PM2 5
EMISSIONS
WITHOUT
FURTHER PM
CONTROL (TONS)
FINAL HD PHASE
2 PROGRAM
NATIONAL PM2 5
EMISSIONS WITH
FURTHER PM
CONTROL (TONS)
NET IMPACT ON
NATIONAL PM2 5
EMISSION WITH
FURTHER PM
CONTROL ON
APUS (TONS)
2040
20,939
21,403
20,476
-927
2050
22,995
23,529
22,416
-1,114
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an
explanation of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1

-------
It is worth noting that the emission reductions shown in Table 5-48 are not incremental to
the emissions reductions projected in the Phase 1 rulemaking. This is because the agencies have
revised their assumptions about the adoption rate of APUs. This final rule assumes that without
the Phase 2 program (i.e., in the Phase 2 baselines), the APU adoption rate will be 9 percent for
model years 2010 and later, which is lower than the value used in both the Phase 1 control case
and Phase 2 proposal. This decision was based on the agencies' assessment of how the current
level of automatic engine shutdown and idle reduction technologies are used by the tractor
manufacturers to comply with the 2014 model year CO2 and fuel consumption standards. To
date, the manufacturers are meeting the 2014 model year standards without the use of this
technology. Compared to Phase 1, the final program projects lower and much delayed
penetration of APUs (including both diesel- and battery-powered) and other idle reduction
technologies starting in model year 2021 (Figure 5-4).
—*—Phase 1 Control	Phase 2 NPRM - Reference	Phase 2 NPRM - Control • Phase 2 FRM - Reference ¦ Phase 2 FRM - Control
120
100
80
c
o
ro
QJ
§ 60
Q.
CL
<
40
k	*	*-
20
0
K? qN*	^ qv3	^	<$>	<$> -o*
'V'V'V'V'V'V'V'V'V'V'V'V'V'V'V'V'V'V'V'V JS*
Model Year
Figure 5-4 Comparison of Assumed Diesel and Battery-Powered APU Use during Extended Idle in Phase 1
and Phase 2
Considering the change in assumptions about APU use and the magnitude of impact of
APUs on criteria emissions, and the revised extended idle rates, EPA conducted an analysis
estimating the combined impacts of the Phase 1 and Phase 2 programs on downstream emissions
for NOx, VOC, SOx and PM2.5 in calendar year 2050. The analysis estimated the combined
Phase 1 and Phase 2 emissions impacts by comparing the Phase 2 control case inventories to the
Phase 1 reference case inventories. To be consistent with the emissions modeling done for this
program, the emissions inventories for Phase 1 reference case were estimated using the same


-------
version of MOVES used for the Phase 2 final rulemaking.1 The results are shown in Table 5-50
The differences in downstream reduction estimates between Phase 2 alone (Table 5-48) and
combined Phase 1 and Phase 2 (Table 5-50) reflect the improvements in road loads from Phase
1. For NOx and PM2.5 only, we also estimated the combined Phase 1 and Phase 2 downstream
and upstream emissions impacts for calendar year 2025, and project that the two rules combined
will reduce NOx by up to 55,000 tons and PM2.5 by up to 33,000 tons in that year.
Table 5-50 Combined Phase 1 and Phase 2 Annual Downstream Emissions Impacts in Calendar Year 2050
using Analysis Method B a
CY
NOx
voc
SOx
PM2.5
2050
-100,878
-10,067
-2,249
-1,001
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for
an explanation of the flat baseline, la, and dynamic baseline, lb, please see Preamble
Section X.A.I
5.5.2.1.2 Upstream Impacts
The final program is projected to reduce the upstream emissions associated with fuel
production and distribution because the projected fuel savings of the program will reduce the
demands for gasoline and diesel. Table 5-51 summarizes the annual upstream reductions of the
final program for criteria pollutants and individual air toxic pollutants in calendar years 2025,
2040 and 2050, relative to Alternative la. The results are shown both in changes in absolute tons
and in percent reductions from the flat baseline for the heavy-duty sector.
Table 5-51 Annual Upstream Impacts of Heavy-Duty Non-GHG Emissions in Calendar Years 2025,2040 and
2050 - Final Program vs. Alt la using Analysis Method Ba
POLLUTANT
CY2025
CY2040
CY2050
US Short
Tons
% Change
US Short
Tons
% Change
US Short
Tons
% Change
1,3-Butadiene
-1
-4.8%
-5
-19.0%
-6
-20.6%
Acetaldehyde
-7
-3.2%
-35
-14.5%
-38
-15.9%
Acrolein
-1
-3.5%
-3
-15.2%
-4
-16.7%
Benzene
-30
-3.8%
-143
-16.1%
-166
-17.6%
CO
-3,809
-4.8%
-16,884
-18.9%
-20,227
-20.5%
Formaldehyde
-20
-4.6%
-90
-18.3%
-107
-19.9%
NOx
-9,314
-4.8%
-41,280
-18.9%
-49,462
-20.5%
PM25
-1,037
-4.7%
-4,619
-18.7%
-5,520
-20.3%
SOx
-5,828
-4.8%
-25,811
-18.9%
-30,941
-20.5%
voc
-4,234
-3.7%
-20,010
-15.9%
-23,240
-17.4%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
T The emissions modeling for Phase 1 was performed using MOVES2010a.

-------
5.5.2.1.3 Total Impacts
As shown in Table 5-52, the agencies estimate that this program will result in overall net
reductions of NOx, VOC, SOx, CO, PM2.5, and air toxics emissions. The results are shown both
in changes in absolute tons and in percent reductions from the flat baseline for the heavy-duty
sector.
Table 5-52 Annual Total Impacts (Upstream and Downstream) of Heavy-Duty Non-GHG Emissions in
Calendar Years 2025,2040 and 2050 - Final Program vs. Alt la using Analysis Method B a
POLLUTANT
CY2025
CY2040
CY2050
US Short
Tons
% Change
US Short
Tons
% Change
US Short
Tons
% Change
1,3-Butadiene
-2
-0.5%
-8
-3.7%
-9
-4.1%
Acetaldehyde
-10
-0.3%
-53
-2.0%
-61
-2.1%
Acrolein
-1
-0.1%
-4
-1.3%
-5
-1.3%
Benzene
-35
-1.1%
-165
-6.8%
-192
-7.5%
CO
-13,254
-0.6%
-52,594
-3.3%
-63,869
-3.8%
Formaldehyde
-40
-0.5%
-187
-2.7%
-227
-2.9%
NOx
-22,710
-1.9%
-101,961
-12.1%
-123,824
-13.3%
PM2.5
-1,110
-1.9%
-5,081
-11.1%
-6,100
-12.1%
SOx
-6,080
-4.8%
-26,933
-18.9%
-32,282
-20.5%
VOC
-5,305
-2.2%
-25,070
-11.9%
-29,253
-13.0%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
5.5.2.2 Model Year Lifetime Analysis
In addition to the annual non-GHG emissions reductions expected from the final
program, the combined (downstream and upstream) non-GHG impacts for the lifetime of the
impacted vehicles were estimated by heavy-duty vehicle category. Table 5-53 shows the fleet-
wide reductions of NOx, PM2.5 and SOx from the final program, relative to Alternative la,
through the lifetime11 of heavy-duty vehicles.
u A lifetime of 30 years is assumed in MOVES.

-------
Table 5-53 Lifetime Non-GHG Reductions by Heavy-Duty Vehicle Category - Summary for Model Years
2018-2029 using Analysis Method B (US Short Tons)a

FINAL PROGRAM

(ALTERNATIVE 3)
NO-ACTION ALTERNATIVE
(BASELINE)
la (Flat)
NOx
549,881
HD Pickups and Vans
30,239
Vocational
42,621
Tractor/Trailers
477,021
PM25
32,251
HD Pickups and Vans
4,042
Vocational
4,436
Tractor/Trailers
23,773
sox
175,202
HD Pickups and Vans
21,464
Vocational
25,082
Tractor/Trailers
128,656
Note:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline,
la, and dynamic baseline, lb, please see Section X. A. 1.
5.5.2.3 Comparison between Emission Inventories for Air Quality Modeling
and Final Rule Inventories
Emissions and air quality modeling decisions are made early in the analytical process
because of the time and resources associated with full-scale photochemical air quality modeling.
As a result, it was necessary to use emissions from the proposed program to conduct the air
quality modeling for this action. The air quality inventories and the final inventories are
consistent in many ways but exhibit several important differences, as illustrated by the
comparison presented in Table 5-54. The final program emission reductions shown in the table
reflect updates to underlying assumptions, modeling inputs, and program standards, but the
largest differences between these inventories and the air quality modeling inventories can be
specifically attributed to changes in our assumptions about APU use and additional requirements
to control PM2.5 emissions from APUs. For example, as described in Preamble Section III.C.3,
EPA is adopting Phase 1 and Phase requirements to control PM2.5 emissions from APUs
installed in new tractors, so we do not expect increases in downstream PM2.5 emissions from the
Phase 2 program; however, the air quality inventories do not reflect these requirements for
APUs, and therefore show increases in downstream PM2.5 emissions. Assumptions about the
penetration of APUs also differ between the air quality inventories and the final rule inventories;
as shown in Figure 5-4, the air quality (proposal) inventories assumed more widespread
penetration of APUs than was assumed for the final program (see Chapter 5.3.2.3.1.1 of this RIA
and Preamble Section III.D.l.a for more detail on the APU assumptions).
Furthermore, because of the differences in methodology between the national inventories
and air quality inventories, particularly the treatment of local variables, such as vehicle

-------
populations, VMT, age distributions, vehicle speed distributions, and the handling of the
temperature effects in MOVES, the more detailed approach used for the air quality inventory
produced different emission estimates than those described in the national inventory section
above.
Table 5-54 Emissions Reductions from the AQ Inventory and the Final Program Inventory


AQ INVENTORY
FINAL PROGRAM INVENTORY
NOx
Downstream
-244,904
-60,681
Upstream
-9,871
-41,280
Total
-254,785
-101,961
PM2.5
Downstream
1,674
-462
Upstream
-2,202
-4,619
Total
-528
-5,082
voc
Downstream
-29,207
-5,060
Upstream
-11,297
-20,010
Total
-40,504
-25,071
SOx
Downstream
-891
-1,122
Upstream
-8,972
-25,811
Total
-9,863
-26,933

-------
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 (IPCC).
2	U.S. EPA. 2015. Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2013. EPA 430-R-15-003.
Available at
http://www3.epa.gov/climatechange/Downloads/ghgemissions/US-GHG-Inventory-2014-Main-Text.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
https://www3.epa.gov/climatechange/Downloads/endangerment/Endangerment_TSD.pdf
4	MOVES homepage: https://www3.epa.gov/otaq/models/moves/index.htm.
5	Argonne National Laboratory. The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation
(GREET) Model versions 1.8.c. http://greet.es.anl.gov/files/372dv49w.
6	2007 Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4)
7	Annual Energy Outlook 2015. http://www.eia.gov/forecasts/archive/aeol5/.
8	U.S. EPA. Updates to MOVES for Emissions Analysis of Greenhouse Gas Emissions and Fuel Efficiency
Standards for Medium- and Heavy-Duty Engines and Vehicles - Phase 2 FRM. Docket No. EPA-HQ-OAR-2016
July, 2016.
9	Memorandum to the Docket "Runspecs, Model Inputs, MOVES Code and Database for HD GHG Phase 2 FRM
Emissions Modeling" Docket No. EPA-HQ-OAR-2016. July, 2016.
10	U.S. EPA. Regulatory Impact Analysis: Changes to Renewable Fuel Standard Program. Chapters 2 and 3. May
26, 2009. Docket No. EPA-HQ-OAR-2009-0472-0119.
11	2017 and Later Model Year Light-Duty Vehicle Greenhouse Gas Emissions and Corporate Average Fuel
Economy Standards (77 FR 62623, October 15, 2012).
12	Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and
Vehicles (76 FR 57106, September 15, 2011).
13	Memorandum to the Docket "Upstream Emissions Modeling Files for HDGHG Phase 2 FRM" Docket No. EPA-
HQ-OAR-2016. July, 2016.
14	U.S. EPA. 2015. "Exhaust Emission Rates for Heavy-Duty On-road Vehicles in MOVES2014" EPA-420-R-15-
015a.
15	Memorandum to the Docket "FRM - Tractor-Trailer Inputs to MOVES" Docket No. EPA-HQ-OAR-2016. July,
2016.
16	ACT Research Co., LLC. U.S. Trailers Monthly Market Indicators. Available at www.actresearch.net/reports
Accessed 7/28/2014.
17	U.S. Census Bureau. 2002 Vehicle Inventory and Use Survey. Available at
https://www.census.gov/svsd/www/vius/2002.html Accessed 6/30/2014.
18	Memorandum to the Docket "FRM - Vocational Inputs to MOVES" Docket No. EPA-HQ-OAR-2016. July, 2016.
19	Memorandum to the Docket "VMT Rebound Inputs to MOVES for HD GHG Phase 2 FRM" Docket No. EPA-
HQ-OAR-2016. July, 2016.
20	Craig Harvey, EPA, "Calculation of Upstream Emissions for the GHG Vehicle Rule." 2009. Docket No. EPA-
HQ-OAR-2009-0472-0216.
21	The Minnesota refrigerant leakage data: https://www.pca.state.mn.us/quick-links/climate-change-mobile-air-
conditioners.
22	Eastern Research Group. "A Study of R134a Leaks in Heavy Duty Vehicles." CARB Contract 06-342. Presented
during CARB Seminar on January 6, 2011.
23	Memorandum to the Docket "Emission Inventories for Air Quality Modeling Technical Support Document"
Docket No. EPA-HQ-OAR-2016. July, 2016.

-------
Chapter 6: Health and Environmental Impacts
6.1 Health and Environmental Effects of Non-GHG Pollutants
6.1.1 Health Effects Associated with Exposure to Non-GHG Pollutants
Along with reducing GHGs, the Phase 2 standards also have an impact on non-GHG
(criteria and air toxic pollutant) emissions. As discussed in Chapter 5, the standards will impact
exhaust emissions of these pollutants from vehicles and will also impact emissions that occur
during the refining and distribution of fuel (upstream sources).
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 will not be directly regulated by the standards, but the
standards will affect emissions of these pollutants and precursors.
6.1.1.1 Particulate Matter
6.1.1.1.1 Background on Particulate Matter
Particulate matter (PM) is a highly complex mixture of solid particles and liquid droplets
distributed among numerous atmospheric gases which interact with solid and liquid phases.
Particles range in size from those smaller than 1 nanometer (10~9 meter) to over 100 micrometers
(|im, or 10"6 meter) in diameter (for reference, a typical strand of human hair is 70 |im in
diameter and a grain of salt is about 100 |im). Atmospheric particles can be grouped into several
classes according to their aerodynamic and physical sizes. Generally, the three broad classes of
particles include ultrafine particles (UFPs, generally considered as particulates with a diameter
less than or equal to 0.1 |im [typically based on physical size, thermal diffusivity or electrical
mobility]), "fine" particles (PM2.5; particles with a nominal mean aerodynamic diameter less
than or equal to 2.5 |im), and "thoracic" particles (PM10; particles with a nominal mean
aerodynamic diameter less than or equal to 10 |im). Particles that fall within the size range
between PM2.5 and PM10, are referred to as "thoracic coarse particles" (PM 10-2.5, particles with a
nominal mean aerodynamic diameter less than or equal to 10 |im and greater than 2.5 |im). EPA
currently has standards that regulate PM2.5 and PMio.a
Particles span many sizes and shapes and may consist of hundreds of different chemicals.
Particles are emitted directly 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. Particle concentration and composition varies by time of year and location, and, in
addition to differences in source emissions, is affected by several weather-related factors, such as
temperature, clouds, humidity, and wind. A further layer of complexity comes from particles'
A Regulatory definitions of PM size fractions, and information on reference and equivalent methods for measuring
PM in ambient air, are provided in 40 CFR Parts 50, 53, and 58. With regard to national ambient air quality
standards (NAAQS) which provide protection against health and welfare effects, the 24-hour PM10 standard
provides protection against effects associated with short-term exposure to thoracic coarse particles (i.e., PM10-2.5).

-------
ability to shift between solid/liquid and gaseous phases, which is influenced by concentration and
meteorology, especially temperature.
Fine particles are produced primarily by combustion processes and by transformations of
gaseous emissions (e.g., sulfur oxides (SOx), nitrogen oxides (NOx) and volatile organic
compounds (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 components including sulfates, nitrates, organic compounds,
elemental carbon and metal compounds. These particles can remain in the atmosphere for days
to weeks and travel through the atmosphere hundreds to thousands of kilometers.1
6.1.1.1.2 Health Effects of Particulate Matter
Scientific studies show exposure to ambient PM is associated with a broad range of
health effects. These health effects are discussed in detail in the Integrated Science Assessment
for Particulate Matter (PM ISA), which was finalized in December 2009.2 The PM ISA
summarizes health effects evidence for short- and long-term exposures to PM2.5, PM 10-2.5, and
ultrafine particles.6 The PM ISA concludes that human exposures to ambient PM2.5 are
associated with a number of adverse health effects and characterizes the weight of evidence for
broad health categories (e.g., cardiovascular effects, respiratory effects, etc.).c The discussion
below highlights the PM ISA's conclusions pertaining to health effects associated with both
short- and long-term PM exposures. Further discussion of health effects associated with PM can
also be found in the rulemaking documents for the most recent review of the PM NAAQS
completed in 2012.3'4
EPA has concluded that "a causal relationship exists" between both long- and short-term
exposures to PM2.5 and premature mortality and cardiovascular effects and that "a causal
relationship is likely to exist" between long- and short-term PM2.5 exposures and respiratory
effects. Further, there is evidence "suggestive of a causal relationship" between long-term PM2.5
exposures and other health effects, including developmental and reproductive effects (e.g., low
birth weight, infant mortality) and carcinogenic, mutagenic, and genotoxic effects (e.g., lung
cancer mortality).0
As summarized in the Final PM NAAQS rule, and discussed extensively in the 2009 PM
ISA, the available scientific evidence significantly strengthens the link between long- and short-
B The ISA also evaluated evidence for PM components, but did not reach causal determinations for components.
c The causal framework draws upon the assessment and integration of evidence from across epidemiological,
controlled human exposure, and toxicological studies, and the related uncertainties that ultimately influence our
understanding of the evidence. This framework employs a five-level hierarchy that classifies the overall weight of
evidence and causality using the following categorizations: causal relationship, likely to be causal relationship,
suggestive of a causal relationship, inadequate to infer a causal relationship, and not likely to be a causal relationship
(U.S. EPA. (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-08/139F, Table 1-3).
D These causal inferences are based not only on the more expansive epidemiological evidence available in this
review of the PM NAAQS but also reflect consideration of important progress that has been made to advance
understanding of a number of potential biologic modes of action or pathways for PM-related cardiovascular and
respiratory effects (U.S. EPA. (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R-08/139F, Chapter 5).

-------
term exposure to PM2.5 and premature mortality, while providing indications that the magnitude
of the PM2.5- mortality association with long-term exposures may be larger than previously
estimated.5'6 The strongest evidence comes from recent studies investigating long-term exposure
to PM2.5 and cardiovascular-related mortality. The evidence supporting a causal relationship
between long-term PM2.5 exposure and mortality also includes consideration of studies that
demonstrated an improvement in community health following reductions in ambient fine
particles.7
Several studies evaluated in the 2009 PM ISA have examined the association between
cardiovascular effects and long-term PM2.5 exposures in multi-city studies conducted in the U.S.
and Europe. These studies have provided new evidence linking long-term exposure to PM2.5
with an array of cardiovascular effects such as heart attacks, congestive heart failure, stroke, and
mortality. This evidence is coherent with studies of short-term exposure to PM2.5 that have
observed associations with a continuum of effects ranging from subtle changes in indicators of
cardiovascular health to serious clinical events, such as increased hospitalizations and emergency
department visits due to cardiovascular disease and cardiovascular mortality.8
As detailed in the 2009 PM ISA, extended analyses of seminal epidemiological studies,
as well as more recent epidemiological studies conducted in the U.S. and abroad, provide strong
evidence of respiratory-related morbidity effects associated with long-term PM2.5 exposure. The
strongest evidence for respiratory-related effects is from studies that evaluated decrements in
lung function growth (in children), increased respiratory symptoms, and asthma development.
The strongest evidence from short-term PM2.5 exposure studies has been observed for increased
respiratory-related emergency department visits and hospital admissions for chronic obstructive
pulmonary disease (COPD) and respiratory infections.9
The body of scientific evidence detailed in the 2009 PM ISA is still limited with respect
to associations between long-term PM2.5 exposures and developmental and reproductive effects
as well as cancer, mutagenic, and genotoxic effects. The strongest evidence for an association
between PM2.5 and developmental and reproductive effects comes from epidemiological studies
of low birth weight and infant mortality, especially due to respiratory causes during the post-
neonatal period (i.e., 1 month to 12 months of age). With regard to cancer effects, "[mjultiple
epidemiologic studies have shown a consistent positive association between PM2.5 and lung
cancer mortality, but studies have generally not reported associations between PM2.5 and lung
cancer incidence."10'11
In addition to evaluating the health effects attributed to short- and long-term exposure to
PM2.5, the 2009 PM ISA also evaluated whether specific components or sources of PM2.5 are
more strongly associated with specific health effects. An evaluation of those studies resulted in
the 2009 PM ISA concluding that "many [components] of PM can be linked with differing health
effects and the evidence is not yet sufficient to allow differentiation of those [components] or
sources that are more closely related to specific health outcomes."12
For PM10-2.5, the 2009 PM ISA concluded that available evidence was "suggestive of a
causal relationship" between short-term exposures to PM10-2.5 and cardiovascular effects (e.g.,
hospital admissions and ED visits, changes in cardiovascular function), respiratory effects (e.g.,
ED visits and hospital admissions, increase in markers of pulmonary inflammation), and

-------
premature mortality. The scientific evidence was "inadequate to infer a causal relationship"
between long-term exposure to PM10-2.5 and various health effects. n-14-15
For UFPs, the 2009 PM ISA concluded that the evidence was "suggestive of a causal
relationship" between short-term exposures and cardiovascular effects, including changes in
heart rhythm and vasomotor function (the ability of blood vessels to expand and contract). It
also concluded that there was evidence "suggestive of a causal relationship" between short-term
exposure to UFPs and respiratory effects, including lung function and pulmonary inflammation,
with limited and inconsistent evidence for increases in ED visits and hospital admissions.
Scientific evidence was "inadequate to infer a causal relationship" between short-term exposure
to UFPs and additional health effects including premature mortality as well as long-term
exposure to UFPs and all health outcomes evaluated.16'17
The 2009 PM ISA conducted an evaluation of specific groups within the general
population potentially at increased risk for experiencing adverse health effects related to PM
exposures. 18>19>20>21 The evidence detailed in the 2009 PM ISA expands our understanding of
previously identified at-risk populations and lifestages (i.e., children, older adults, and
individuals with pre-existing heart and lung disease) and supports the identification of additional
at-risk populations (e.g., persons with lower socioeconomic status, genetic differences).
Additionally, there is emerging, though still limited, evidence for additional potentially at-risk
populations and lifestages, such as those with diabetes, people who are obese, pregnant women,
and the developing fetus.22
6.1.1.2 Ozone
6.1.1.2.1 Background on Ozone
Ground-level ozone pollution is typically formed through reactions involving VOCs and
NOx in the lower atmosphere in the presence of sunlight. These pollutants, often referred to as
ozone precursors, are emitted by many types of pollution sources such as highway and nonroad
motor vehicles and engines, power plants, chemical plants, refineries, makers of consumer and
commercial products, industrial facilities, and smaller area sources.
The science of ozone formation, transport, and accumulation is complex. Ground-level
ozone is produced and destroyed in a cyclical set of chemical reactions, many of which are
sensitive to temperature and sunlight. When ambient temperatures and sunlight levels remain
high for several days and the air is relatively stagnant, ozone and its precursors can build up and
result in more ozone than typically occurs on a single high-temperature day. Ozone and its
precursors can be transported hundreds of miles downwind of precursor emissions, resulting in
elevated ozone levels even in areas with low VOC or NOx emissions.
The highest levels of ozone are produced when both VOC and NOx emissions are
present in significant quantities on clear summer days. Relatively small amounts of NOx enable
ozone to form rapidly when VOC levels are relatively high, but ozone production is quickly
limited by removal of the NOx. Under these conditions NOx reductions are highly effective in
reducing ozone while VOC reductions have little effect. Such conditions are called "NOx-
limited." Because the contribution of VOC emissions from biogenic (natural) sources to local

-------
ambient ozone concentrations can be significant, even some areas where man-made VOC
emissions are relatively low can be NOx-limited.
Ozone concentrations in an area also can be lowered by the reaction of nitric oxide (NO)
with ozone, forming nitrogen dioxide (NO2). As the air moves downwind and the cycle
continues, the NO2 forms additional ozone. The importance of this reaction depends, in part, on
the relative concentrations of NOx, VOC, and ozone, all of which change with time and location.
When NOx levels are relatively high and VOC levels relatively low, NOx forms inorganic
nitrates (i.e., particles) but relatively little ozone. Such conditions are called "VOC-limited."
Under these conditions, VOC reductions are effective in reducing ozone, but NOx reductions can
actually increase local ozone under certain circumstances. Even in VOC-limited urban areas,
NOx reductions are not expected to increase ozone levels if the NOx reductions are sufficiently
large. Rural areas are usually NOx-limited, due to the relatively large amounts of biogenic VOC
emissions in such areas. Urban areas can be either VOC- or NOx-limited, or a mixture of both,
in which ozone levels exhibit moderate sensitivity to changes in either pollutant.
6.1.1.2.2 Health Effects of Ozone
This section provides a summary of the health effects associated with exposure to
ambient concentrations of ozone.E The information in this section is based on the information
and conclusions in the February 2013 Integrated Science Assessment for Ozone (Ozone ISA).23
The Ozone ISA concludes that human exposures to ambient concentrations of ozone are
associated with a number of adverse health effects and characterizes the weight of evidence for
these health effects. F The discussion below highlights the Ozone ISA's conclusions pertaining
to health effects associated with both short-term and long-term periods of exposure to ozone.
For short-term exposure to ozone, the Ozone ISA concludes that respiratory effects,
including lung function decrements, pulmonary inflammation, exacerbation of asthma,
respiratory-related hospital admissions, and mortality, are causally associated with ozone
exposure. It also concludes that cardiovascular effects, including decreased cardiac function and
increased vascular disease, and total mortality are likely to be causally associated with short-term
exposure to ozone and that evidence is suggestive of a causal relationship between central
nervous system effects and short-term exposure to ozone.
For long-term exposure to ozone, the Ozone ISA concludes that respiratory effects,
including new onset asthma, pulmonary inflammation and injury, are likely to be causally related
with ozone exposure. The Ozone ISA characterizes the evidence as suggestive of a causal
relationship for associations between long-term ozone exposure and cardiovascular effects,
reproductive and developmental effects, central nervous system effects and total mortality. The
E 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 breathing route and rate.
F The ISA evaluates evidence and draws conclusions on the causal relationship between relevant pollutant exposures
and health effects, assigning one of five "weight of evidence" determinations: 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 more information on these levels of evidence, please refer to Table II in the Preamble of
the ISA.

-------
evidence is inadequate to infer a causal relationship between chronic ozone exposure and
increased risk of lung cancer.
Finally, interindividual variation in human responses to ozone exposure can result in
some groups being at increased risk for detrimental effects in response to exposure. In addition,
some groups are at increased risk of exposure due to their activities, such as outdoor workers and
children. The Ozone ISA identified several groups that are at increased risk for ozone-related
health effects. These groups are people with asthma, children and older adults, individuals with
reduced intake of certain nutrients (i.e., Vitamins C and E), outdoor workers, and individuals
having certain genetic variants related to oxidative metabolism or inflammation. Ozone
exposure during childhood can have lasting effects through adulthood. Such effects include
altered function of the respiratory and immune systems. Children absorb higher doses
(normalized to lung surface area) of ambient ozone, compared to adults, due to their increased
time spent outdoors, higher ventilation rates relative to body size, and a tendency to breathe a
greater fraction of air through the mouth. Children also have a higher asthma prevalence
compared to adults. Additional children's vulnerability and susceptibility factors are listed in
Section XIV of the Preamble.
6.1.1.3 Nitrogen Oxides
6.1.1.3.1	Background on Nitrogen Oxides
Oxides of nitrogen (NOx) refers to nitric oxide (NO) and nitrogen dioxide (NO2). For
the NOx NAAQS, NO2 is the indicator. Most NO2 is formed in the air through the oxidation of
nitric oxide (NO) emitted when fuel is burned at a high temperature. NOx is also a major
contributor to secondary PM2.5 formation. The health effects of ambient PM are discussed in
Chapter 6.1.1.1.2. NOx along with VOCs are the two major precursors of ozone. The health
effects of ozone are covered in Chapter 6.1.1.2.2.
6.1.1.3.2	Health Effects of Nitrogen Oxides
The most recent review of the health effects of oxides of nitrogen completed by EPA can
be found in the 2016 Integrated Science Assessment for Oxides of Nitrogen - Health Criteria
(Oxides of Nitrogen ISA).G The primary source of NO2 is motor vehicle emissions, and ambient
NO2 concentrations tend to be highly correlated with other traffic-related pollutants. Thus, a key
issue in characterizing the causality of N02-health effect relationships was evaluating the extent
to which studies supported an effect of NO2 that is independent of other traffic-related
pollutants. EPA concluded that the findings for asthma exacerbation integrated from
epidemiologic and controlled human exposure studies provided evidence that is sufficient to
infer a causal relationship between respiratory effects and short-term NO2 exposure. The
strongest evidence supporting an independent effect of NO2 exposure comes from controlled
human exposure studies demonstrating increased airway responsiveness in individuals with
asthma following ambient-relevant NO2 exposures. The coherence of this evidence with
epidemiologic findings for asthma hospital admissions and ED visits as well as lung function
G U.S. EPA. Integrated Science Assessment for Oxides of Nitrogen - Health Criteria (2016 Final Report). U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R-15/068, 2016.

-------
decrements and increased pulmonary inflammation in children with asthma describe a plausible
pathway by which NO2 exposure can cause an asthma exacerbation. The 2016 ISA for Oxides
of Nitrogen also concluded that there is likely to be a causal relationship between long-term NO2
exposure and respiratory effects. This conclusion is based on new epidemiologic evidence for
associations of NO2 with asthma development in children combined with biological plausibility
from experimental studies.
In evaluating a broader range of health effects, the 2016 ISA for Oxides of Nitrogen
concluded evidence is "suggestive of, but not sufficient to infer, a causal relationship" between
short-term NO2 exposure and cardiovascular effects and mortality and between long-term NO2
exposure and cardiovascular effects and diabetes, birth outcomes, and cancer. In addition, the
scientific evidence is inadequate (insufficient consistency of epidemiologic and toxicological
evidence) to infer a causal relationship for long-term NO2 exposure with fertility, reproduction,
and pregnancy, as well as with postnatal development. A key uncertainty in understanding the
relationship between these non-respiratory health effects and short- or long-term exposure to
NO2 is co-pollutant confounding, particularly by other roadway pollutants. The available
evidence for non-respiratory health effects does not adequately address whether NO2 has an
independent effect or whether it primarily represents effects related to other or a mixture of
traffic-related pollutants.
The 2016 ISA for Oxides of Nitrogen concluded that people with asthma, children, and
older adults are at increased risk for N02-related health effects. In these groups and lifestages,
NO2 is consistently related to larger effects on outcomes related to asthma exacerbation, for
which there is confidence in the relationship with NO2 exposure.
6,1,1,4 Sulfur Oxides
6.1.1.4.1	Background
Sulfur dioxide (SO2), 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. SO2 andits gas phase oxidation products can dissolve in water droplets and
further oxidize to form sulfuric acid which reacts with ammonia to form sulfates, which are
important components of ambient PM. The health effects of ambient PM are discussed in
Chapter 6.1.1.1.2.
6.1.1.4.2	Health Effects of Sulfur Oxides
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 2008 Integrated Science Assessment
for Sulfur Oxides - Health Criteria (SOx ISA).24 Following an extensive evaluation of health
evidence from epidemiologic and laboratory studies, 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 addition to those with asthma (both children and adults), potentially at-risk groups
include all children and the elderly. In free-breathing laboratory studies involving controlled

-------
human exposures to SO2, respiratory effects have consistently been observed following 5-10 min
exposures at SO2 concentrations > 400 ppb in asthmatics engaged in moderate to heavy levels of
exercise, with respiratory effects occurring at concentrations as low as 200 ppb in some
asthmatics. A clear concentration-response relationship has been demonstrated in these studies
following exposures to SO2 at concentrations between 200 and 1000 ppb, 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 has examined potential
confounding by co-pollutants using multipollutant regression models. These analyses indicate
that although co-pollutant adjustment has varying degrees of influence on the SO2 effect
estimates, the effect of SO2 on respiratory health outcomes appears to be generally robust and
independent of the effects of gaseous and particulate co-pollutants, suggesting that the observed
effects of SO2 on respiratory endpoints occur independent of the effects of other ambient air
pollutants.
Consistent associations between short-term exposure to SO2 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 SO2 on respiratory morbidity, uncertainty remains with respect to the interpretation of these
observed mortality associations due to potential confounding by various co-pollutants.
Therefore, EPA has concluded that the overall evidence is suggestive of a causal relationship
between short-term exposure to SO2 and mortality. Significant associations between short-term
exposure to SO2 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 SO2 exposure and
cardiovascular morbidity.
6.1.1.5 Carbon Monoxide
6.1.1.5.1	Background
Carbon monoxide (CO) is a colorless, odorless gas emitted from combustion processes.
Nationally, particularly in urban areas, the majority of CO emissions to ambient air come from
mobile sources.25
6.1.1.5.2	Health Effects of Carbon Monoxide
Information on the health effects of carbon monoxide (CO) can be found in the January
2010 Integrated Science Assessment for Carbon Monoxide (CO ISA).26 The CO ISA presents

-------
conclusions regarding the presence of causal relationships between CO exposure and categories
of adverse health effects.H This section provides a summary of the health effects associated with
exposure to ambient concentrations of CO, along with the ISA conclusions.1
Controlled human exposure 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 observed 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 CO ISA concludes that a causal relationship is likely to exist 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 central nervous system and behavioral effects following low-level
CO exposures, although the findings have not been consistent across all studies. The CO 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 studies cited in the CO ISA have evaluated the role of CO exposure in birth
outcomes such as preterm birth or cardiac birth defects. There is limited epidemiologic evidence
of a CO-induced effect on preterm births and birth defects, with weak evidence for a decrease in
birth weight. Animal toxicological studies have found perinatal CO exposure to affect birth
weight, as well as other developmental outcomes. The CO ISA concludes the evidence is
suggestive of a causal relationship between long-term exposures to CO and developmental
effects and birth outcomes.
Epidemiologic studies provide evidence of associations between short-term CO
concentrations and respiratory morbidity such as changes in pulmonary function, respiratory
symptoms, and hospital admissions. A limited number of epidemiologic studies considered co-
pollutants 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 CO ISA concludes that the evidence is suggestive
of a causal relationship between short-term CO exposure and respiratory morbidity, and
H The ISA evaluates the health evidence associated with different health effects, assigning one of five "weight of
evidence" determinations: 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.
1 Personal exposure includes contributions from many sources, and in many different environments. Total personal
exposure to CO includes both ambient and non-ambient components; and both components may contribute to
adverse health effects.

-------
inadequate to conclude that a causal relationship exists between long-term exposure and
respiratory morbidity.
Finally, the CO ISA concludes that the epidemiologic evidence is suggestive of a causal
relationship between short-term concentrations of CO and mortality. Epidemiologic evidence
suggests an association exists 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 co-pollutant
models contributes to the uncertainty as to whether CO is acting alone or as an indicator for other
combustion-related pollutants. The CO ISA also concludes that there is not likely to be a causal
relationship between relevant long-term exposures to CO and mortality.
6.1.1.6 Diesel Exhaust
6.1.1.6.1	Background on Diesel Exhaust
Diesel exhaust consists of a complex mixture composed of particulate matter, carbon
dioxide, oxygen, nitrogen, water vapor, carbon monoxide, nitrogen compounds, sulfur
compounds and numerous low-molecular-weight hydrocarbons. A number of these gaseous
hydrocarbon components are individually known to be toxic, including aldehydes, benzene and
1,3-butadiene. The diesel particulate matter present in diesel exhaust consists mostly of fine
particles (< 2.5 |im), of which a significant fraction is ultrafine particles (< 0.1 |im). These
particles have a large surface area which makes them an excellent medium for adsorbing
organics and their small size makes them highly respirable. Many of the organic compounds
present in the gases and on the particles, such as polycyclic organic matter, are individually
known to have mutagenic and carcinogenic properties.
Diesel exhaust varies significantly in chemical composition and particle sizes between
different engine types (heavy-duty, light-duty), engine operating conditions (idle, acceleration,
deceleration), and fuel formulations (high/low sulfur fuel). Also, there are emissions differences
between on-road and nonroad engines because the nonroad engines are generally of older
technology. After being emitted in the engine exhaust, diesel exhaust undergoes dilution as well
as chemical and physical changes in the atmosphere. The lifetime for some of the compounds
present in diesel exhaust ranges from hours to days.
6.1.1.6.2	Health Effects of Diesel Exhaust
In EPA's 2002 Diesel Health Assessment Document (Diesel HAD), exposure to diesel
exhaust was classified as likely to be carcinogenic to humans by inhalation from environmental
exposures, in accordance with the revised draft 1996/1999 EPA cancer guidelines.27'28 A
number of other agencies (National Institute for Occupational Safety and Health, the
International Agency for Research on Cancer, the World Health Organization, California EPA,
and the U.S. Department of Health and Human Services) had made similar hazard classifications
prior to 2002. EPA also concluded in the 2002 Diesel HAD that it was not possible to calculate a
cancer unit risk for diesel exhaust due to limitations in the exposure data for the occupational
groups or the absence of a dose-response relationship.

-------
In the absence of a cancer unit risk, the Diesel HAD sought to provide additional insight
into the significance of the diesel exhaust cancer hazard by estimating possible ranges of risk that
might be present in the population. An exploratory analysis was used to characterize a range of
possible lung cancer risk. The outcome was that environmental risks of cancer from long-term
diesel exhaust exposures could plausibly range from as low as 10"5 to as high as 10"3. Because of
uncertainties, the analysis acknowledged that the risks could be lower than 10"5, and a zero risk
from diesel exhaust exposure could not be ruled out.
Noncancer health effects of acute and chronic exposure to diesel exhaust emissions are
also of concern to EPA. EPA derived a diesel exhaust reference concentration (RfC) from
consideration of four well-conducted chronic rat inhalation studies showing adverse pulmonary
effects. The RfC is 5 |ig/m3 for diesel exhaust measured as diesel particulate matter. This RfC
does not consider allergenic effects such as those associated with asthma or immunologic or the
potential for cardiac effects. There was emerging evidence in 2002, discussed in the Diesel
HAD, that exposure to diesel exhaust can exacerbate these effects, but the exposure-response
data were lacking at that time to derive an RfC based on these then-emerging considerations.
The EPA Diesel HAD states, "With [diesel particulate matter] being a ubiquitous component of
ambient PM, there is an uncertainty about the adequacy of the existing [diesel exhaust]
noncancer database to identify all of the pertinent [diesel exhaust]-caused noncancer health
hazards." The Diesel HAD also notes "that acute exposure to [diesel exhaust] has been
associated with irritation of the eye, nose, and throat, respiratory symptoms (cough and phlegm),
and neurophysiological symptoms such as headache, lightheadedness, nausea, vomiting, and
numbness or tingling of the extremities." The Diesel HAD noted that the cancer and noncancer
hazard conclusions applied to the general use of diesel engines then on the market and as cleaner
engines replace a substantial number of existing ones, the applicability of the conclusions would
need to be reevaluated.
It is important to note that the Diesel HAD also briefly summarizes health effects
associated with ambient PM and discusses EPA's then-annual PM2.5 NAAQS of 15 |ig/m3. In
2012, EPA revised the annual PM2.5 NAAQS to 12 |ig/m3. There is a large and extensive body
of human data showing a wide spectrum of adverse health effects associated with exposure to
ambient PM, of which diesel exhaust is an important component. The PM2.5 NAAQS is
designed to provide protection from the noncancer health effects and premature mortality
attributed to exposure to PM2.5. The contribution of diesel PM to total ambient PM varies in
different regions of the country and also, within a region, from one area to another. The
contribution can be high in near-roadway environments, for example, or in other locations where
diesel engine use is concentrated.
Since 2002, several new studies have been published which continue to report increased
lung cancer risk with occupational exposure to diesel exhaust from older engines. Of particular
note since 2011 are three new epidemiology studies which have examined lung cancer in
occupational populations, for example, truck drivers, underground nonmetal miners and other
diesel motor-related occupations. These studies reported increased risk of lung cancer with
exposure to diesel exhaust with evidence of positive exposure-response relationships to varying
degrees.29'30'31 These newer studies (along with others that have appeared in the scientific
literature) add to the evidence EPA evaluated in the 2002 Diesel HAD and further reinforces the
concern that diesel exhaust exposure likely poses a lung cancer hazard. The findings from these

-------
newer studies do not necessarily apply to newer technology diesel engines since the newer
engines have large reductions in the emission constituents compared to older technology diesel
engines.
In light of the growing body of scientific literature evaluating the health effects of
exposure to diesel exhaust, in June 2012 the World Health Organization's International Agency
for Research on Cancer (IARC), a recognized international authority on the carcinogenic
potential of chemicals and other agents, evaluated the full range of cancer-related health effects
data for diesel engine exhaust. IARC concluded that diesel exhaust should be regarded as
"carcinogenic to humans."32 This designation was an update from its 1988 evaluation that
considered the evidence to be indicative of a "probable human carcinogen."
6.1.1.7 Air Toxics
Heavy-duty vehicle emissions contribute to ambient levels of air toxics that are known or
suspected human or animal carcinogens, or that have noncancer health effects. The population
experiences an elevated risk of cancer and other noncancer health effects from exposure to the
class of pollutants known collectively as "air toxics."33 These compounds include, but are not
limited to, benzene, 1,3-butadiene, formaldehyde, acetaldehyde, acrolein, polycyclic organic
matter, and naphthalene. These compounds were identified as national or regional risk drivers or
contributors in the 2011 National-scale Air Toxics Assessment and have significant inventory
contributions from mobile sources.34
6.1.1.7.1 Health Effects of Benzene
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.35'36'37 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. EPA's IRIS documentation for benzene also lists a range of 2.2 x 10"6 to
7.8 x 10"6 per |ig/m3 as the unit risk estimate (URE) for benzene/'38 The International Agency
for Research on Cancer (IARC) has determined that benzene is a human carcinogen and the U.S.
Department of Health and Human Services (DHHS) has characterized benzene as a known
human carcinogen.39'40
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.41'42 The most sensitive noncancer effect observed in humans, based on current data, is
the depression of the absolute lymphocyte count in blood.43'44 EPA's inhalation reference
concentration (RfC) for benzene is 30 |ig/m3. The RfC is based on suppressed absolute
lymphocyte counts seen in humans under occupational exposure conditions. In addition, recent
work, including studies sponsored by the Health Effects Institute (HEI), provides evidence that
1A unit risk estimate is defined as the increase in the lifetime risk of an individual who is exposed for a lifetime to 1
|ig/m3 benzene in air.

-------
biochemical responses are occurring at lower levels of benzene exposure than previously
known.45'46'47'48 EPA's IRIS program has not yet evaluated these new data. EPA does not
currently have an acute reference concentration for benzene. The Agency for Toxic Substances
and Disease Registry (ATSDR) Minimal Risk Level (MRL) for acute exposure to benzene is 29
|ig/m3 for 1-14 days exposure. 49'K
6.1.1.7.2	Health Effects of 1,3-Butadiene
EPA has characterized 1,3-butadiene as carcinogenic to humans by inhalation.50'51 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.52'53'54 There are numerous studies
consistently demonstrating that 1,3-butadiene is metabolized into genotoxic metabolites by
experimental animals and humans. The specific mechanisms of 1,3-butadiene-induced
carcinogenesis are unknown; however, the scientific evidence strongly suggests that the
carcinogenic effects are mediated by genotoxic metabolites. Animal data suggest that females
may be more sensitive than males for cancer effects associated with 1,3-butadiene exposure;
there are insufficient data in humans from which to draw conclusions about sensitive
subpopulations. The URE for 1,3-butadiene is 3 x 10"5 per |ig/m3.55 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.56 Based on this critical effect and the benchmark concentration methodology, an
RfC for chronic health effects was calculated at 0.9 ppb (approximately 2 |ig/m3).
6.1.1.7.3	Health Effects of Formaldehyde
In 1991, EPA concluded that formaldehyde is a carcinogen based on nasal tumors in
animal bioassays.57 An Inhalation URE for cancer and a Reference Dose for oral noncancer
effects were developed by the agency and posted on the Integrated Risk Information System
(IRIS) database. Since that time, the National Toxicology Program (NTP) and International
Agency for Research on Cancer (IARC) have concluded that formaldehyde is a known human
carcinogen.58'59'60
The conclusions by IARC and NTP reflect the results of epidemiologic research
published since 1991 in combination with previous animal, human and mechanistic evidence.
Research conducted by the National Cancer Institute reported an increased risk of
nasopharyngeal cancer and specific lymphohematopoietic malignancies among workers exposed
to formaldehyde.61'62'63 A National Institute of Occupational Safety and Health study of garment
workers also reported increased risk of death due to leukemia among workers exposed to
formaldehyde.64 Extended follow-up of a cohort of British chemical workers did not report
evidence of an increase in nasopharyngeal or lymphohematopoietic cancers, but a continuing
statistically significant excess in lung cancers was reported.65 Finally, a study of embalmers
reported formaldehyde exposures to be associated with an increased risk of myeloid leukemia
but not brain cancer.66
K A minimal risk level (MRL) is defined as an estimate of the daily human exposure to a hazardous substance that is
likely to be without appreciable risk of adverse noncancer health effects over a specified duration of exposure.

-------
Health effects of formaldehyde in addition to cancer were reviewed by the Agency for
Toxics Substances and Disease Registry in 199967, supplemented in 2010,68 and by the World
Health Organization.69 These organizations reviewed the scientific literature concerning health
effects linked to formaldehyde exposure to evaluate hazards and dose response relationships and
defined exposure concentrations for minimal risk levels (MRLs). The health endpoints reviewed
included sensory irritation of eyes and respiratory tract, reduced pulmonary function, nasal
histopathology, and immune system effects. In addition, research on reproductive and
developmental effects and neurological effects were discussed along with several studies that
suggest that formaldehyde may increase the risk of asthma - particularly in the young.
EPA released a draft Toxicological Review of Formaldehyde - Inhalation Assessment
through the IRIS program for peer review by the National Research Council (NRC) and public
comment in June 2010.70 The draft assessment reviewed more recent research from animal and
human studies on cancer and other health effects. The NRC released their review report in April
201171 (http://www.nap.edu/catalog.php?record_id=13142). EPA is currently developing a
revised draft assessment in response to this review.
6.1.1.7.4	Health Effects of 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.72 The URE in IRIS for acetaldehyde is 2.2 x 10"6 per |ig/m3.73 Acetaldehyde is
reasonably anticipated to be a human carcinogen by the U.S. DHHS in the 13th Report on
Carcinogens and is classified as possibly carcinogenic to humans (Group 2B) by the IARC.74 75
EPA is currently conducting a reassessment of cancer risk from inhalation exposure to
acetaldehyde. Acetaldehyde is currently listed on the IRIS Program Multi-Year Agenda for
reassessment within the next few years.
The primary noncancer effects of exposure to acetaldehyde vapors include irritation of
the eyes, skin, and respiratory tract.76 In short-term (4 week) rat studies, degeneration of
olfactory epithelium was observed at various concentration levels of acetaldehyde exposure.77'78
Data from these studies were used by EPA to develop an inhalation reference concentration of 9
|ig/m3. Some asthmatics have been shown to be a sensitive subpopulation to decrements in
functional expiratory volume (FEV1 test) and bronchoconstriction upon acetaldehyde
inhalation.79
6.1.1.7.5	Health Effects of Acrolein
EPA most recently evaluated the toxicological and health effects literature related to
acrolein in 2003 and concluded 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.80 The IARC determined in 1995 that acrolein was not classifiable as to its
carcinogenicity in humans.81

-------
Lesions to the lungs and upper respiratory tract of rats, rabbits, and hamsters have been
observed after subchronic exposure to acrolein.82 The agency has developed an RfC for acrolein
of 0.02 |ig/m3 and an RfD of 0.5 |ig/kg-day.83
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.84 These
data and additional studies regarding acute effects of human exposure to acrolein are
summarized in EPA's 2003 Toxicological Review of Acrolein.85 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. Acute exposures in animal studies report bronchial hyper-responsiveness. Based on
animal data (more pronounced respiratory irritancy in mice with allergic airway disease in
comparison to non-diseased mice86) and demonstration of similar effects in humans (e.g.,
reduction in respiratory rate), individuals with compromised respiratory function (e.g.,
emphysema, asthma) are expected to be at increased risk of developing adverse responses to
strong respiratory irritants such as acrolein. EPA does not currently have an acute reference
concentration for acrolein. The available health effect reference values for acrolein have been
summarized by EPA and include an ATSDR MRL for acute exposure to acrolein of 7 |ig/m3 for
1-14 days exposure; and Reference Exposure Level (REL) values from the California Office of
Environmental Health Hazard Assessment (OEHHA) for one-hour and 8-hour exposures of 2.5
|ig/m3 and 0.7 |ig/m3, respectively.87
6.1.1.7.6	Health Effects of Polycyclic Organic Matter (POM)
The term polycyclic organic matter (POM) defines a broad class of compounds that
includes the polycyclic aromatic hydrocarbon compounds (PAHs). One of these compounds,
naphthalene, is discussed separately below. POM compounds are formed primarily from
combustion and are present in the atmosphere in gas and particulate form. Cancer is the major
concern from exposure to POM. Epidemiologic studies have reported an increase in lung cancer
in humans exposed to diesel exhaust, coke oven emissions, roofing tar emissions, and cigarette
smoke; all of these mixtures contain POM compounds.8889 Animal studies have reported
respiratory tract tumors from inhalation exposure to benzo[a]pyrene and alimentary tract and
liver tumors from oral exposure to benzo[a]pyrene.90 In 1997 EPA classified seven PAHs
(benzo[a]pyrene, benz[a]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene,
dibenz[a,h]anthracene, and indeno[l,2,3-cd]pyrene) as Group B2, probable human
carcinogens.91 Since that time, studies have found that maternal exposures to PAHs 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 in
preschool children (3 years of age).92,93 These and similar studies are being evaluated as a part of
the ongoing IRIS reassessment of health effects associated with exposure to benzo[a]pyrene.
6.1.1.7.7	Health Effects of 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. Acute (short-term) exposure of humans to naphthalene by inhalation, ingestion, or
dermal contact is associated with hemolytic anemia and damage to the liver and the nervous
system.94 Chronic (long term) exposure of workers and rodents to naphthalene has been reported
to cause cataracts and retinal damage.95 EPA released an external review draft of a reassessment
of the inhalation carcinogenicity of naphthalene based on a number of recent animal
carcinogenicity studies.96 The draft reassessment completed external peer review.97 Based on
external peer review comments received, a revised draft assessment that considers all routes of
exposure, as well as cancer and noncancer effects, is under development. The 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.98
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.99
Naphthalene also causes a number of chronic non-cancer effects in animals, including
abnormal cell changes and growth in respiratory and nasal tissues.100 The current EPA IRIS
assessment includes noncancer data on hyperplasia and metaplasia in nasal tissue that form the
basis of the inhalation RfC of 3 |ig/m3.101 The ATSDR MRL for acute exposure to naphthalene
is 0.6 mg/kg/day.
6.1.1.7.8 Health Effects of Other Air Toxics
In addition to the compounds described above, other compounds in gaseous hydrocarbon
and PM emissions from vehicles will be affected by the rules. 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.102
6.1.1.8 Exposure and Health Effects Associated with Traffic
In addition to health concerns resulting from specific air pollutants, a large number of
studies have examined the health status of populations near major roadways. These studies
frequently have employed exposure metrics that are not specific to individual pollutants, but
rather reflect the large number of different pollutants found in elevation near major roads.
In this section of the RIA, information on health effects associated with air quality near
major roads or traffic in general is summarized. Generally, the section makes use of publications
that systematically review literature on a given health topic. In particular, this section makes
frequent reference of a report of by the Health Effects Institute (HEI) Panel on the Health Effects
of Traffic-Related Air Pollution, published in 2010 as a review of relevant studies.L'103 Other
systematic reviews of relevant literature are cited were appropriate.
L It should be noted that there are no peer reviewed EPA-authored reviews of traffic-related health studies. The HEI
panel primarily used epidemiology studies for inferring whether there was sufficient evidence of a causal association

-------
6.1.1.8.1
Populations near Major Roads
Numerous studies have estimated the size and demographics of populations that live near
major roads. Other studies have estimated the number of schools near major roads, and the
populations of students in such schools.
Every two years, the U.S. Census Bureau's American Housing Survey (AHS) has
reported whether housing units are within 300 feet of an "airport, railroad, or highway with four
or more lanes." The 2009 survey reports that over 22 million homes, or 17 percent of all housing
units in the U.S., were located in such areas. Assuming that populations and housing units are in
the same locations, this corresponds to a population of more than 50 million U.S. residents in
close proximity to high-traffic roadways or other transportation sources. According to the
Central Intelligence Agency's World Factbook, in 2010, the United States had 6,506,204 km or
roadways, 224,792 km of railways, and 15,079 airports. As such, highways represent the
overwhelming majority of transportation facilities described by this factor in the AHS.
The AHS reports are published every two years, and until 2011 recorded whether homes
were located near highways with four or more lanes, railroads, or airports. As such, trends in the
AHS can be reported to describe whether a greater or lesser proportion of homes are located near
major roads over time. Figure 6-1 depicts trends in the number and proportion of homes located
near major transportation sources, which generally indicate large roadways. As the figure
indicates, since 2005, there has been a substantial increase in the number and percentage of
homes located near major transportation sources. As such, the population in close proximity to
these sources, which may be affected by near-road air quality and health concerns, appears to
have increased over time.
exists between a particular health effect and traffic-related air pollution. In its weight-of-evidence determinations,
the panel also placed "considerable weight" on controlled human exposure studies. However, it restricted
consideration of other toxicological studies to whether or not the studies provided "general mechanistic support" for
the inferences of causality made on the basis of epidemiology.

-------
24
22
20
w
0)
E 18
o
La
14
12
10
I

1
¦
j 18.0%
-- 16.0%
14.0%
-- 12.0%
-- 10.0%
8.0%
6.0%
+ 4.0%
2.0%
0.0%
0)
o
ci3
CL
czaMillion Housing Units
-~-Near4+ Lane Highway,
Railroad, or Airport
1997 1999 2001 2003 2005 2007 2009
Yearof American Housing Survey
Figure 6-1 Trends in Populations near Large Highways, Railroads, and Airports
Furthermore, according to data from the 2008 American Time Use Survey (ATUS),
conducted by the Bureau of Labor Statistics (BTS), Americans spend more than an hour
traveling each day, on average.104 Although the ATUS does not indicate their mode of travel,
the majority of trips undertaken nationally is by motor vehicle.105 As such, daily travel activity
brings nearly all residents into a high-exposure microenvironment for part of the day.
6.1.1.8.2
Premature Mortality
The HEI panel report concluded that evidence linking traffic-associated air pollution with
premature mortality from all causes was "suggestive but not sufficient" to infer a causal
relationship. This conclusion was based largely on several long-term studies that "qualitatively"
examined whether or not someone was exposed to traffic-associated air pollution. In addition,
based on several short-term studies of exposure, the panel concluded that there was "suggestive
but not sufficient" evidence to infer a causal relation between traffic-related exposure and
cardiovascular mortality.
6.1.1.8.3 Cardiovascular Effects
6.1.1.8.3.1 Cardiac Physiology
Exposure to traffic-associated pollutants has been associated with changes in cardiac
physiology, including cardiac function. One common measure of cardiac function is heart rate

-------
variability (HRV), an indicator of the heart's ability to respond to variations in stress, reflecting
the nervous system's ability to regulate the heart.M Reduced HRV is associated with adverse
cardiovascular events, such as myocardial infarction, in heart disease patients. The HEI panel
concluded that available evidence provides evidence for a causal association between exposure
to traffic-related pollutants and reduced control of HRV by the nervous system. Overall, the
panel concluded that the evidence was "suggestive but not sufficient" to infer a causal relation
between traffic-related pollutants and cardiac function. Studies suggest that the HRV changes
from traffic-related air pollution result in changes to heart rhythms, which can lead to
arrhythmia.106'107
6.1.1.8.3.2 Heart Attack and Atherosclerosis
The HEI panel concluded that epidemiologic evidence of the association between traffic-
related pollutants and heart attacks and atherosclerosis was "suggestive but not sufficient" to
infer a causal association. In addition, the panel concluded that the toxicology studies they
reviewed provided "suggestive evidence that exposure to traffic emissions, including ambient
and laboratory-generated [PM] and diesel- and gasoline-engine exhaust, alters cardiovascular
function." The panel noted there are few studies of human volunteers exposed to real-world
traffic mixture, which were not entirely consistent. The panel notes that the studies provide
consistent evidence for exposure to PM and impaired cardiovascular responses. In addition to
the HEI study, several other reviews of available evidence conclude that there is evidence
supporting a causal association between traffic-related air pollution and cardiovascular
disease.108
A number of mechanisms for cardiovascular disease are highlighted in the HEI and AHA
report, including modified blood vessel endothelial function (e.g., the ability to dilate),
atherosclerosis, and oxidative stress. The HEI review cites "two well executed studies" in which
hospitalization for acute myocardial infarction (i.e., heart attack) were associated with traffic
exposures and a prospective study finding higher rates of arterial hardening and coronary heart
disease near traffic.
6.1.1.8.4 Respiratory Effects
6.1.1.8.4.1 Asthma
Pediatric asthma and asthma symptoms are the effects that have been evaluated by the
largest number of studies in the epidemiologic literature on the topic. In general, studies
consistently show effects of residential or school exposure to traffic and asthma symptoms, and
the effects are frequently statistically significant. Studies have employed both short-term and
long-term exposure metrics, and a range of different respiratory measures. HEI Special Report
17 (HEI Panel on the Health Effects of Traffic-Related Air Pollution, 2010) concluded that there
is sufficient evidence for a causal association between exposure to traffic-related air pollution
and exacerbation of asthma symptoms in children.
M The autonomic nervous system (ANS) consists of sympathetic and parasympathetic components. The sympathetic
ANS signals body systems to "fight or flight." The parasympathetic ANS signals the body to "rest and digest." In
general, HRV is indicative of parasympathetic control of the heart.

-------
While there is general consistency in studies examining asthma incidence in children, the
available studies employ different definitions of asthma (e.g., self-reported vs. hospital records),
methods of exposure assessment, and population age ranges. As such, the overall evidence,
while supportive of an association between traffic exposure and new onset asthma, are less
consistent than for asthma symptoms. The HEI report determined that evidence is between
"sufficient" and "suggestive" of a causal relationship between exposure to traffic-related air
pollution and incident (new onset) asthma in children (HEI Panel on the Health Effects of
Traffic-Related Air Pollution, 2010). A recent meta-analysis of studies on incident asthma and
air pollution in general, based on studies dominated by traffic-linked exposure metrics, also
concluded that available evidence is consistent with HEI's conclusion (Anderson et al., 2011).
The study reported excess main risk estimates for different pollutants ranging from 7-16 percent
per 10 |j,g/m3 of long-term exposure (random effects models). Other qualitative reviews (Salam
et al., 2008; Braback and Forsberg, 2009) conclude that available evidence is consistent with the
hypothesis that traffic-associated air pollutants are associated with incident asthma.
6.1.1.8.4.2	Chronic Obstructive Pulmonary Disease (COPD)
The HEI panel reviewed available studies examining COPD in the context of traffic-
associated air pollution. Because of how the panel selected studies for inclusion in review, there
were only two studies that they used to review the available evidence. Both studies reported
some positive associations, but not for all traffic metrics. The small number of studies and lack
of consistency across traffic metrics led the panel to conclude that there is insufficient evidence
for traffic-associated air pollution causing COPD.
6.1.1.8.4.3	Allergy
There are numerous human and animal experimental studies that provides strongly
suggestive evidence that traffic-related air pollutants can enhance allergic responses to common
allergens.109'110'111 However, in its review of 16 epidemiologic studies that address traffic-
related air pollution's effect on allergies, the HEI expert panel (HEI, 2010) reported that only
two such studies showed consistently positive associations. As a result, despite the strongly
suggestive experimental evidence, the panel concluded that there is "inadequate/insufficient"
evidence of an association between allergy and traffic-associated air pollution. As noted above,
the HEI panel considered toxicological studies only based on whether or not they provide
mechanistic support for observations and inferences derived from epidemiology.
6.1.1.8.4.4	Lung Fun ction
There are numerous measurements of breathing (spirometry) that indicate the presence or
degree of airway disease, such as asthma and chronic obstructive pulmonary disease (COPD).
Forced vital capacity (FVC) is measured when a patient maximally fills their lungs and then
blows their hardest in completely exhaling. The peak expiratory flow (PEF) is the maximum air
flow achievable during exhalation. The forced expiratory volume in the first second of
exhalation is referred to as FEV i. FEVi and PEF reflect the function of the large airways. FVC
and FEVi, along with their ratio (FVC/FEVi) are used to classify airway obstruction in asthma
and COPD. Measurements of air flow at various times during forced exhalation, such as 25

-------
percent, 50 percent, and 75 percent, are also used. The flow at 75 percent of forced exhalation
(FEF75) reflects the status of small airways, which asthma and COPD affect.
The HEI panel concluded that the available literature suggests that long-term exposure to
traffic-related air pollution is associated with reduced lung function in adolescents and young
adults and that lung function is lower in populations in areas with high traffic-related air
pollutant levels. However, the panel noted the difficulty of disentangling traffic-specific
exposures from urban air pollution in general. The studies reviewed that were more specifically
oriented toward traffic were not consistent in their findings. As a result, the panel found that the
evidence linking lung function and traffic exposure is "inadequate and insufficient" to infer a
causal relationship.
6.1.1.8.5	Reproductive and Developmental Effects
Several studies have reported associations between traffic-related air pollution and
adverse birth outcomes, such as preterm birth and low birth weight. At the time of the HEI
review, the panel concluded that evidence for adverse birth outcomes being causally associated
with traffic-related exposures was "inadequate and insufficient." Only four studies met the
panel's inclusion criteria, and had limited geographic coverage. One study provided evidence of
small but consistently increased risks using multiple exposure metrics. No studies were at the
time available that examined traffic-specific exposures and congenital abnormalities. Since then,
several studies investigating birth outcomes have been published, but no new systematic reviews.
One new meta-analysis of air pollution and congenital abnormalities has been published, though
none of the reviewed studies includes traffic-specific exposure information.
The HEI panel also reviewed toxicological studies of traffic-related air pollutants and
fertility. While numerous studies examining animal or human exposure and sperm count have
been published, the panel concluded that the generally high exposure concentrations employed in
the studies limited the applicability to typical ambient concentrations. Because there was no
overlap in the effects studied by epidemiology and toxicology studies, no synthesis review of the
combined literature was undertaken.
Since the HEI panel's publication, a systematic review and meta-analysis of air pollution
and congenital abnormalities was published.112 In that review, only one study directly included
nearby traffic in its exposure analysis. As such, there are no systematic reviews that specifically
address traffic's impact on congenital abnormalities.
6.1.1.8.6	Cancer
6.1.1.8.6.1 Childhood Cancer
In 2014, Boothe et al. published a systematic review and meta-analysis of studies of
childhood leukemia risks associated for populations near major roads.113 The study concluded
that childhood leukemia was positively associated with residential exposure during childhood,
but not during the prenatal period. Other literature reviews have not concluded that available
evidence supports an association between childhood leukemia and traffic exposure.114,115 For
example, the HEI panel concluded that the available epidemiologic evidence was "inadequate

-------
and insufficient" to infer a causal relationship between traffic-related air pollution and childhood
cancer.
6.1.1.8.6.2 Adult Cancer
Several studies have examined the risk of adult lung cancers in relation to exposure to
traffic-related air pollutants. The HEI panel evaluated four such studies, and rated the available
evidence as "inadequate and insufficient" to infer a causal relation for non-occupational lung
cancer.
6.1.1.8.7 Neurological Effects
The HEI panel found that current toxicologic and epidemiologic literature on the
neurotoxicity of traffic-related air pollution was inadequate for their evaluation. The panel noted
that there were a number of toxicologic studies of traffic-associated pollutants, but found them to
have diverse exposure protocols, animal models, and endpoints, making them unsuitable for
systematic evaluation.
6.1.2 Environmental Effects Associated with Exposure to Non-GHG Pollutants
Along with reducing GHGs, the Phase 2 standards also have an impact on non-GHG
(criteria and air toxic pollutant) emissions. As discussed in Chapter 5, the standards will impact
exhaust emissions of these pollutants from vehicles and will also impact emissions that occur
during the refining and distribution of fuel (upstream sources).
In this section we will discuss the environmental effects associated with non-GHG
pollutants, specifically: particulate matter, ozone, NOx, SOx and air toxics.
6,1,2.1 Visibility Degradation
Visibility can be defined as the degree to which the atmosphere is transparent to visible
light.116 Visibility impairment is caused by light scattering and absorption by suspended
particles and gases. Fine particles with significant light-extinction efficiencies include sulfates,
nitrates, organic carbon, elemental carbon, and soil.117 Visibility is important because it has
direct significance to people's enjoyment of daily activities in all parts of the country.
Individuals value good visibility for the well-being it provides them directly, where they live and
work, and in places where they enjoy recreational opportunities. Visibility is also highly valued
in significant natural areas, such as national parks and wilderness areas, and special emphasis is
given to protecting visibility in these areas. For more information on visibility see the final 2009
PM ISA.118
The extent to which any amount of light extinction affects a person's ability to view a
scene depends on both scene and light characteristics. For example, the appearance of a nearby
object (e.g., a building) is generally less sensitive to a change in light extinction than the
appearance of a similar object at a greater distance. See Figure 6-2 for an illustration of the
important factors affecting visibility.

-------
Optical Characteristics of Illumination
Characteristics of Observer
•	Sunlight (Sun Angle)
•	Cloud Cover (Overcast, Puffy, etc.)
•	Sky
Optical Characteristics of
Optical Characteristics of
•	Color
•	Contrast Detail (Texture!
•	Form
•	Brightness
•	Detection Thresholds
•	Psychological Response to
Incoming Light
•	Value Judgements
Light from clouds
scattered Into
sight path ^
Sunlight y
scattered L|gM reflected
from ground
scattered Into
•	Light Added to Sight Path by
Particles and Casts
•	Image-Forming Light Subtracted
from Sight Path by Scattering
and Absorption
Image-forming
light scattered
out of sight path
Image-forming
light absorbed
Figure 6-2 Important Factors Involved in Seeing a Scenic Vista (Malm, 1999)
EPA is working to address visibility impairment. Reductions in air pollution from
implementation of various programs associated with the Clean Air Act Amendments of 1990
(CAAA) provisions have resulted in substantial improvements in visibility and will continue to
do so in the future. Because trends in haze are closely associated with trends in particulate
sulfate and nitrate due to the relationship between their concentration and light extinction,
visibility trends have improved as emissions of SO2 and NOx have decreased over time due to
air pollution regulations such as the Acid Rain Program.119
In the Clean Air Act Amendments of 1977, Congress recognized visibility's value to
society by establishing a national goal to protect national parks and wilderness areas from
visibility impairment caused by manmade pollution.1^ In 1999, EPA finalized the regional haze
program (64 FR 35714) 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 38680-38681, July 18, 1997). 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. Figure 6-3 shows
the location of the 156 Mandatory Class I Federal areas.
® See Section 169(a) of the Clean Air Act.

-------
NPS Units
FWS Units
* Rainbow Lake, W1 and Bradwell Bay, FL are Class 1 Areas
where visibility is not an important air quality related value
FS Units
Produced by NPS Air Resources Division
Figure 6-3 Mandatory Class I Federal Areas in the U.S.
EPA has also concluded that PM2.5 causes adverse effects on visibility in other areas that
are not targeted by the Regional Haze Rule, such as urban areas, depending on PM2.5
concentrations and other factors such as dry chemical composition and relative humidity (i.e., an
indicator of the water composition of the particles). EPA revised the PM2.5 standards in
December 2012 and established a target level of protection that is expected to be met through
attainment of the existing secondary standards for PM2.5.
6.1.2.2 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., EPA has supported visibility monitoring in
national parks and wilderness areas since 1988. The monitoring network was originally
established at 20 sites, but it has now been expanded to 110 sites that represent all but one of the
156 Mandatory Federal Class I areas across the country (see Figure 6-3). 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 PM10 and PM2.5
mass, and for key constituents of PM2.5, such as sulfate, nitrate, organic and elemental carbon
OC and EC), soil dust, and several other elements. Measurements for specific aerosol
constituents are used to calculate "reconstructed" aerosol light extinction by multiplying the mass
for each constituent by its empirically-derived scattering and/or absorption efficiency, with
adjustment for the relative humidity. The IMPROVE program utilizes both an "original" and a
"revised" reconstruction formula for this purpose, with the latter explicitly accounting for sea salt

-------
concentrations. 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 made principally with a nephelometer to
measure light scattering, some sites also include an aethalometer for light absorption, or a few
sites use a transmissometer, which measures total light extinction. Scene characteristics are
typically recorded using digital or video 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 aero sol-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 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
6.1,2.3 Plant and Ecosystem Effects of Ozone
The welfare effects of ozone can be observed across a variety of scales, i.e. subcellular,
cellular, leaf, whole plant, population and ecosystem. Ozone effects that begin at small spatial
scales, such as the leaf of an individual plant, when they occur at sufficient magnitudes (or to a
sufficient degree) can result in effects being propagated along a continuum to larger and larger
spatial scales. For example, effects at the individual plant level, such as altered rates of leaf gas
exchange, growth and reproduction, can, when widespread, result in broad changes in
ecosystems, such as productivity, carbon storage, water cycling, nutrient cycling, and community
composition.
Ozone can produce both acute and chronic injury in sensitive species depending on the
concentration level and the duration of the exposure.121 In those sensitive species0, effects from
repeated exposure to ozone throughout the growing season of the plant tend to accumulate, so
that even low concentrations experienced for a longer duration have the potential to create
chronic stress on vegetation. 122'p Ozone damage to sensitive species includes impaired
photosynthesis and visible injury to leaves. The impairment of photosynthesis, the process by
which the plant makes carbohydrates (its source of energy and food), can lead to reduced crop
yields, timber production, and plant productivity and growth. Impaired photosynthesis can also
0 73 FR 16491 (March 27, 2008). Only a small percentage of all the plant species growing within the U.S. (over
43,000 species have been catalogued in the USD A PLANTS database) have been studied with respect to ozone
sensitivity.
p The concentration at which ozone levels overwhelm a plant's ability to detoxify or compensate for oxidant
exposure varies. Thus, whether a plant is classified as sensitive or tolerant depends in part on the exposure levels
being considered.

-------
lead to a reduction in root growth and carbohydrate storage below ground, resulting in other,
more subtle plant and ecosystems impacts.123 These latter impacts include increased
susceptibility of plants to insect attack, disease, harsh weather, interspecies competition and
overall decreased plant vigor. The adverse effects of ozone on areas with sensitive species could
potentially lead to species shifts and loss from the affected ecosystemsQ, resulting in a loss or
reduction in associated ecosystem goods and services.124 Additionally, visible ozone injury to
leaves can result in a loss of aesthetic value in areas of special scenic significance like national
parks and wilderness areas and reduced use of sensitive ornamentals in landscaping.125
The Integrated Science Assessment (ISA) for Ozone presents more detailed information
on how ozone affects vegetation and ecosystems.126 The ISA concludes that ambient
concentrations of ozone are associated with a number of adverse welfare effects and
characterizes the weight of evidence for different effects associated with ozone.R The ISA
concludes that visible foliar injury effects on vegetation, reduced vegetation growth, reduced
productivity in terrestrial ecosystems, reduced yield and quality of agricultural crops, and
alteration of below-ground biogeochemical cycles are causally associated with exposure to
ozone. It also concludes that reduced carbon sequestration in terrestrial ecosystems, alteration of
terrestrial ecosystem water cycling, and alteration of terrestrial community composition are
likely to be causally associated with exposure to ozone.
6,1,2,4 Deposition of Particulate Matter, Nitrogen and Sulfur
Particulate matter contributes to adverse effects on vegetation and ecosystems, and to
soiling and materials damage. These welfare effects result predominantly 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 2008 Integrated Science
Assessment for Oxides of Nitrogen and Sulfur- Ecological Criteria (secondary NOx/SOx
ISA).127'128
6.1.2.4.1 Deposition of Nitrogen and Sulfur
Nitrogen and sulfur interactions in the environment are highly complex as shown in
Figure 6-4. Both nitrogen and sulfur are essential, and sometimes limiting, nutrients needed for
growth and productivity of ecosystem components (e.g. algae, plants). In terrestrial and aquatic
ecosystems excesses of nitrogen or sulfur can lead to acidification and nutrient enrichment.129 In
addition, in aquatic ecosystems, sulfur deposition can increase mercury methylation.
Q Per footnote above, ozone impacts could be occurring in areas where plant species sensitive to ozone have not yet
been studied or identified.
R The Ozone ISA evaluates the evidence associated with different ozone related health and welfare effects, assigning
one of five "weight of evidence" determinations: 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 more
information on these levels of evidence, please refer to Table II of the ISA.

-------
Ambient Air
Concentration
Sunlight
Oxidation	Dissolution
SOi	» H2SO4	* 2H* ~SO**"
NO,	*¦ HNOj 	». H*+NOj"
Wet Deposition
H\NH4*,NOj,:
Dry deposition
NO,. NH„ SO,
Deposition
Foliar and
nutrient effects
Acidification of water + Eutrophication
Ecological
Effect
Figure 6-4 Nitrogen and Sulfur Cycling, and Interactions in the Environment
Source: U.S. EPA, 2008c
6.1.2.4.1.1 Ecological Effects of Acidification
Deposition of nitrogen and sulfur can cause acidification, which alters biogeochemistry
and affects animal and plant life in terrestrial and aquatic ecosystems across the U.S. Soil
acidification is a natural process, but is often accelerated by acidifying deposition, which can
decrease concentrations of exchangeable base cations in soils.130 Biological effects of
acidification in terrestrial ecosystems are generally linked to aluminum toxicity and decreased
ability of plant roots to take up base cations.131 Decreases in the acid neutralizing capacity and
increases in inorganic aluminum concentration contribute to declines in zooplankton, macro
invertebrates, and fish species richness in aquatic ecosystems.132
Geology (particularly surficial geology) is the principal factor governing the sensitivity of
terrestrial and aquatic ecosystems to acidification from nitrogen and sulfur deposition.133
Geologic formations having low base cation supply generally underlie the watersheds of acid-
sensitive lakes and streams. Other factors contribute to the sensitivity of soils and surface waters
to acidifying deposition, including topography, soil chemistry, land use, and hydrologic flow
path.134
6.1.2.4.1.1.1 Aquatic Acidification
Aquatic effects of acidification have been well studied in the U.S. and elsewhere at
various trophic levels. These studies indicate that aquatic biota have been affected by
aci dification at virtually all l evels of the food web in acid sensitive aquatic ecosystems. Effects

-------
have been most clearly documented for fish, aquatic insects, other invertebrates, and algae.
Biological effects are primarily attributable to a combination of low pH and high inorganic
aluminum concentrations. Such conditions occur more frequently during rainfall and snowmelt
that cause high flows of water and less commonly during low-flow conditions, except where
chronic acidity conditions are severe. Biological effects of episodes include reduced fish
condition factorA, changes in species composition and declines in aquatic species richness across
multiple taxa, ecosystems and regions.
Because acidification primarily affects the diversity and abundance of aquatic biota, it
also affects the ecosystem services, e.g., recreational and subsistence fishing, that are derived
from the fish and other aquatic life found in these surface waters. In the northeastern United
States, the surface waters affected by acidification are a source of food for some recreational and
subsistence fishermen and for other consumers with particularly high rates of self-caught fish
consumption, such as the Hmong and Chippewa ethnic groups.135'136
6.1.2.4.1.1.2 Terrestrial Acidification
Acidifying deposition has altered major biogeochemical processes in the U.S. by
increasing the nitrogen and sulfur content of soils, accelerating nitrate and sulfate leaching from
soil to drainage waters, depleting base cations (especially calcium and magnesium) from soils,
and increasing the mobility of aluminum. Inorganic aluminum is toxic to some tree roots. Plants
affected by high levels of aluminum from the soil often have reduced root growth, which restricts
the ability of the plant to take up water and nutrients, especially calcium.137 These direct effects
can, in turn, influence the response of these plants to climatic stresses such as droughts and cold
temperatures. They can also influence the sensitivity of plants to other stresses, including insect
pests and disease leading to increased mortality of canopy trees.138 In the U.S., terrestrial effects
of acidification are best described for forested ecosystems (especially red spruce and sugar maple
ecosystems) with additional information on other plant communities, including shrubs and
lichen.139
Both coniferous and deciduous forests throughout the eastern U.S. are experiencing
gradual losses of base cation nutrients from the soil due to accelerated leaching from acidifying
deposition. This change in nutrient availability may reduce the quality of forest nutrition over
the long term. Evidence suggests that red spruce and sugar maple in some areas in the eastern
U.S. have experienced declining health because of this deposition. For red spruce, (Picea
rubens) dieback or decline has been observed across high elevation landscapes of the
northeastern U.S., and to a lesser extent, the southeastern U.S., and acidifying deposition has
been implicated as a causal factor.140
6.1.2.4.1.2 Ecological Effects from Nitrogen Enrichment
6.1.2.4.1.2.1 Aquatic Enrichment
Eutrophication in estuaries is associated with a range of adverse ecological effects
including low dissolved oxygen (DO), harmful algal blooms (HABs), loss of submerged aquatic
vegetation (SAV), and low water clarity. Low DO disrupts aquatic habitats, causing stress to
fish and shellfish, which, in the short-term, can lead to episodic fish kills and, in the long-term,
can damage overall growth in fish and shellfish populations. Low DO also degrades the

-------
aesthetic qualities of surface water. In addition to often being toxic to fish and shellfish, and
leading to fish kills and aesthetic impairments of estuaries, HABs can, in some instances, also be
harmful to human health. S AV provides critical habitat for many aquatic species in estuaries
and, in some instances, can also protect shorelines by reducing wave strength; therefore, declines
in SAV due to nutrient enrichment are an important source of concern. Low water clarity is in
part the result of accumulations of both algae and sediments in estuarine waters. In addition to
contributing to declines in SAV, high levels of turbidity also degrade the aesthetic qualities of
the estuarine environment.
An assessment of estuaries nationwide by the National Oceanic and Atmospheric
Administration (NOAA) concluded that 64 estuaries (out of 99 with available data) suffered
from moderate or high levels of eutrophication due to excessive inputs of both N and
phosphorus.141 For estuaries in the Mid-Atlantic region, the contribution of atmospheric
deposition to total N loads is estimated to range between 10 percent and 58 percent.142 Estuaries
in the eastern United States are an important source of food production, in particular fish and
shellfish production. The estuaries are capable of supporting large stocks of resident commercial
species, and they serve as the breeding grounds and interim habitat for several migratory species.
Eutrophi cation in estuaries may also affect the demand for seafood after well-publicized toxic
blooms, water-based recreation, and erosion protection provided by SAV.
6.1.2.4.1.2.2 Terrestrial Enrichment
Terrestrial enrichment occurs when terrestrial ecosystems receive N loadings in excess of
natural background levels, through either atmospheric deposition or direct application.
Atmospheric N deposition is associated with changes in the types and number of species and
biodiversity in terrestrial systems. Nitrogen enrichment occurs over a long time period; as a
result, it may take as much as 50 years or more to see changes in ecosystem conditions and
indicators. One of the main provisioning services potentially affected by N deposition is grazing
opportunities offered by grasslands for livestock production in the Central U.S. Although N
deposition on these grasslands can offer supplementary nutritive value and promote overall grass
production, there are concerns that fertilization may favor invasive grasses and shift the species
composition away from native grasses. This process may ultimately reduce the productivity of
grasslands for livestock production.
Terrestrial enrichment also affects habitats, for example the Coastal Sage Scrub (CSS)
and Mixed Conifer Forest (MCF) habitats which are an integral part of the California landscape.
Together the ranges of these habitats include the densely populated and valuable coastline and
the mountain areas. Numerous threatened and endangered species at both the state and federal
levels reside in CSS and MCF. Fire regulation is also an important regulating service that could
be affected by nutrient enrichment of the CSS and MCF ecosystems by encouraging growth of
more flammable grasses, increasing fuel loads, and altering the fire cycle.
6.1.2.4.1.3 Vegetation Effects Associated with Gaseous Sulfur Dioxide
Uptake of gaseous sulfur dioxide in a plant canopy is a complex process involving
adsorption to surfaces (leaves, stems, and soil) and absorption into leaves. SO2 penetrates into
leaves through the stomata, although there is evidence for limited pathways via the cuticle.143

-------
Pollutants must be transported from the bulk air to the leaf boundary layer in order to get to the
stomata. When the stomata are closed, as occurs under dark or drought conditions, resistance to
gas uptake is very high and the plant has a very low degree of susceptibility to injury. In
contrast, mosses and lichens do not have a protective cuticle barrier to gaseous pollutants or
stomates and are generally more sensitive to gaseous sulfur and nitrogen than vascular plants.144
Acute foliar injury usually happens within hours of exposure, involves a rapid absorption of a
toxic dose, and involves collapse or necrosis of plant tissues. Another type of visible injury is
termed chronic injury and is usually a result of variable SO2 exposures over the growing season.
Besides foliar injury, chronic exposure to low SO2 concentrations can result in reduced
photosynthesis, growth, and yield of plants.145 These effects are cumulative over the season and
are often not associated with visible foliar injury. As with foliar injury, these effects vary among
species and growing environment. SO2 is also considered the primary factor causing the death of
lichens in many urban and industrial areas.146
6.1.2.4.1.4 Mercury Methylation
Mercury is a persistent, bioaccumulative toxic metal that is emitted in three forms:
gaseous elemental Hg (Hg°), oxidized Hg compounds (Hg+2), and particle-bound Hg (Hgp).
Methylmercury (MeHg) is formed by microbial action in the top layers of sediment and soils,
after Hg has precipitated from the air and deposited into waterbodies or land. Once formed,
MeHg is taken up by aquatic organisms and bioaccumulates up the aquatic food web. Larger
predatory fish may have MeHg concentrations many times, typically on the order of one million
times, that of the concentrations in the freshwater body in which they live. The NOx SOx
ISA—Ecological Criteria concluded that evidence is sufficient to infer a causal relationship
between sulfur deposition and increased mercury methylation in wetlands and aquatic
environments.147 Specifically, there appears to be a relationship between SO42" deposition and
mercury methylation; however, the rate of mercury methylation varies according to several
spatial and biogeochemical factors whose influence has not been fully quantified. Therefore, the
correlation between SO42" deposition and MeHg cannot yet be quantified for the purpose of
interpolating the association across waterbodies or regions. Nevertheless, because changes in
MeHg in ecosystems represent changes in significant human and ecological health risks, the
association between sulfur and mercury cannot be neglected.148
6.1.2.4.2 Deposition of Metallic and Organic Constituents of PM
Several significant ecological effects are associated with deposition of chemical
constituents of ambient PM such as metals and organics.149 The trace metal constituents of PM
include cadmium, copper, chromium, mercury, nickel, zinc, and lead. The organics include
persistent organic pollutants (POPs), polyaromatic hydrocarbons (PAHs) and polybromiated
diphenyl ethers (PBDEs). Exposure to PM for direct effects occur via deposition (e.g., wet, dry
or occult) to vegetation surfaces, while indirect effects occur via deposition to ecosystem soils or
surface waters where the deposited constituents of PM then interacts with biological organisms.
While both fine and coarse-mode particles may affect plants and other organisms, more often the
chemical constituents drive the ecosystem response to PM.150 Ecological effects of PM include
direct effects to metabolic processes of plant foliage; contribution to total metal loading resulting
in alteration of soil biogeochemistry and microbiology, plant and animal growth and

-------
reproduction; and contribution to total organics loading resulting in bioaccumulation and
biomagnification.
Particulate matter can adversely impact plants and ecosystem services provided by plants
by deposition to vegetative surfaces.151 Particulates deposited on the surfaces of leaves and
needles can block light, altering the radiation received by the plant. PM deposition near sources
of heavy deposition can obstruct stomata limiting gas exchange, damage leaf cuticles and
increase plant temperatures.152 Plants growing on roadsides exhibit impact damage from near-
road PM deposition, having higher levels of organics and heavy metals, and accumulate salt from
road de-icing during winter months.153 In addition, atmospheric PM can convert direct solar
radiation to diffuse radiation, which is more uniformly distributed in a tree canopy, allowing
radiation to reach lower leaves.154 Decreases in crop yields (a provisioning service) due to
reductions in solar radiation have been attributed to regional scale air pollution in other counties
with especially severe regional haze.155
In addition to damage to plant surfaces, deposited PM can be taken up by plants from soil
or foliage. Copper, zinc, and nickel have been shown to be directly toxic to vegetation under
field conditions.156 The ability of vegetation to take up heavy metals is dependent upon the
amount, solubility and chemical composition of the deposited PM. Uptake of PM by plants from
soils and vegetative surfaces can disrupt photosynthesis, alter pigments and mineral content,
reduce plant vigor, decrease frost hardiness and impair root development.
Particulate matter can also contain organic air toxic pollutants, including PAHs, which
are a class of polycyclic organic matter (POM). PAHs can accumulate in sediments and
bioaccumulate in freshwater, flora and fauna. The uptake of organics depends on the plant
species, site of deposition, physical and chemical properties of the organic compound and
prevailing environmental conditions.157 Different species can have different uptake rates of
PAHs. For example, zucchini (Cucurbita pepo) accumulated significantly more PAHs than
related plant species.158 PAHs can accumulate to high enough concentrations in some coastal
environments to pose an environmental health threat that includes cancer in fish populations,
toxicity to organisms living in the sediment and risks to those (e.g., migratory birds) that
consume these organisms.159'160 Atmospheric deposition of particles is thought to be the major
source of PAHs to the sediments of Lake Michigan, Chesapeake Bay, Tampa Bay and other
coastal areas of the U.S.161
Contamination of plant leaves by heavy metals can lead to elevated concentrations in the
soil. 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.162'163 Many of the major indirect plant responses to PM deposition are chiefly soil-mediated
and depend on the chemical composition of individual components of deposited PM. Upon
entering the soil environment, PM pollutants can alter ecological processes of energy flow and
nutrient cycling, inhibit nutrient uptake to plants, change microbial community structure and,
affect biodiversity. Accumulation of heavy metals in soils depends on factors such as local soil
characteristics, geologic origin of parent soils, and metal bioavailability. Heavy metals, such as
zinc, copper, and cadmium, and some pesticides can interfere with microorganisms that are
responsible for decomposition of soil litter, an important regulating ecosystem service that serves
as a source of soil nutrients.164 Surface litter decomposition is reduced in soils having high metal

-------
concentrations. Soil communities have associated bacteria, fungi, and invertebrates that are
essential to soil nutrient cycling processes. Changes to the relative species abundance and
community composition are associated with deposited PM to soil biota.165
Atmospheric deposition can be the primary source of some organics and metals to
watersheds. Deposition of PM to surfaces in urban settings increases the metal and organic
component of storm water runoff.166 This atmospherically-associated pollutant burden can then
be toxic to aquatic biota. The contribution of atmospherically deposited PAHs to aquatic food
webs was demonstrated in high elevation mountain lakes with no other anthropogenic
contaminant sources.167 Metals associated with PM deposition limit phytoplankton growth,
affecting aquatic trophic structure. Long-range atmospheric transport of 47 pesticides and
degradation products to the snowpack in seven national parks in the Western U.S. was recently
quantified indicating PM-associated contaminant inputs to receiving waters during spring
snowmelt.168
The recently completed Western Airborne Contaminants Assessment Project (WACAP)
is the most comprehensive database on contaminant transport and PM depositional effects on
sensitive ecosystems in the Western U.S.169 In this project, the transport, fate, and ecological
impacts of anthropogenic contaminants from atmospheric sources were assessed from 2002 to
2007 in seven ecosystem components (air, snow, water, sediment, lichen, conifer needles and
fish) in eight core national parks. The study concluded that bioaccumulation of semi-volatile
organic compounds occurred throughout park ecosystems, an elevational gradient in PM
deposition exists with greater accumulation in higher altitude areas, and contaminants
accumulate in proximity to individual agriculture and industry sources, which is counter to the
original working hypothesis that most of the contaminants would originate from Eastern Europe
and Asia.
6.1.2.4.3 Materials Damage and Soiling
Building materials including metals, stones, cements, and paints undergo natural
weathering processes from exposure to environmental elements (e.g., wind, moisture,
temperature fluctuations, sunlight, etc.). Pollution can worsen and accelerate these effects.
Deposition of PM is associated with both physical damage (materials damage effects) and
impaired aesthetic qualities (soiling effects). Wet and dry deposition of PM 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 stone, concrete and marble.170 The effects of PM are exacerbated by the presence of
acidic gases and can be additive or synergistic due to the complex mixture of pollutants in the air
and surface characteristics of the material. Acidic deposition has been shown to have an effect
on materials including zinc/galvanized steel and other metal, carbonate stone (as monuments and
building facings), and surface coatings (paints).171 The effects on historic buildings and outdoor
works of art are of particular concern because of the uniqueness and irreplaceability of many of
these objects.

-------
6,1,2.5 Environmental Effects of Air Toxics
Emissions from producing, transporting and combusting fuel 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.172 In laboratory experiments, a wide range of tolerance to VOCs has been
observed.173 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.174
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.175'176'177 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.
6.2 Impacts of the Rules on Concentrations of Non-GHG Pollutants
Along with reducing GHGs, the Phase 2 standards also have an impact on non-GHG
(criteria and air toxic pollutant) emissions. As discussed in Chapter 5, the standards will impact
exhaust emissions of these pollutants from vehicles and will also impact emissions that occur
during the refining and distribution of fuel (upstream sources).
This section first discusses current concentrations of non-GHG pollutants and then
discusses the projected impacts of the standards on ambient concentrations of non-GHG
pollutants in 2040. Additional information on the air quality modeling methodology and results
of the air quality modeling can be found in Appendix 6A.
6.2.1 Current Concentrations of Non-GHG Pollutants
Nationally, levels of PM2.5, ozone, NOx, SOx, CO and air toxics are declining.178
However as of April 22, 2016, more than 125 million people lived in counties designated
nonattainment for one or more of the NAAQS, and this figure does not include the people living
in areas with a risk of exceeding the NAAQS in the future.s Many Americans continue to be
exposed to ambient concentrations of air toxics at levels which have the potential to cause
adverse health effects.179 In addition, populations who live, work, or attend school near major
roads experience elevated exposure concentrations to a wide range of air pollutants.180
s Data come from Summary Nonattainment Area Population Exposure Report, current as of April 22, 2016 at:
https://www3.epa.gov/airquality/greenbk/popexp.html.

-------
6,2.1,1 Current Concentrations of Particulate Matter
As described in Chapter 6.1, PM causes adverse health effects, and EPA has set national
standards to provide requisite protection against those health effects. There are two primary
NAAQS for PM2.5: an annual standard (12.0 micrograms per cubic meter (|ig/m3)) and a 24-hour
standard (35 (j,g/m3), and two secondary NAAQS for PM2.5: an annual standard (15.0 (J,g/m3) and
a 24-hour standard (35 (j,g/m3). The initial PM2.5 standards were set in 1997 and revisions to the
standards were finalized in 2006 and in December 2012.
There are many areas of the country that are currently in nonattainment for the annual
and 24-hour PM2.5 NAAQS. In 2005 the EPA designated 39 nonattainment areas for the 1997
PM2.5 NAAQS.181 As of April 22, 2016, more than 23 million people lived in the 7 areas that
are still designated as nonattainment for the 1997 annual PM2.5 NAAQS. These PM2.5
nonattainment areas are comprised of 33 full or partial counties. In December 2014 EPA
designated 14 nonattainment areas for the 2012 PM2.5 NAAQS.182 As of April 22, 2016, 9 of
these areas remain designated as nonattainment, and they are composed of 20 full or partial
counties with a population of over 23 million. On November 13, 2009 and February 3, 2011, the
EPA designated 32 nonattainment areas for the 2006 24-hour PM2.5 NAAQS.183 As of April 22,
2016, 16 of these areas remain designated as nonattainment for the 2006 PM2.5 NAAQS, and
they are composed of 46 full or partial counties with a population of over 32 million. In total,
there are currently 24 PM2.5 nonattainment areas with a population of more than 39 million
people.1 Nonattainment areas for the PM2.5 NAAQS are pictured in Figure 6-5.
T The 39 million total is calculated by summing, without double counting, the 1997, 2006 and 2012 PM2 5
nonattainment populations contained in the Summary Nonattainment Area Population Exposure report
(https://www3.epa.gov/airquality/greenbk/popexp.html). If there is a population associated with more than one of
the 1997, 2006 and 2012 nonattainment areas, and they are not the same, then the larger of the populations is
included in the sum.

-------
Counties Designated Nonattainment
for PM-2.5 (1997, 2006, and/or 2012 Standards)
Designated Nonattainment
All three PM-2.5 Standards
I I Both 2006 and 2012 PM-2.5
I I Both 1997 and 2006 PM-2.5
| 2012 PM-2.5 only
I I 2006 PM-2.5 only
I I 1997 PM-2.5 only
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
For PM-2.5 (1997 Standard) Chattanooga TN-GA-AL nonattainment area, the Georgia portion was
redesignated on December 19.2014 and the Alabama portion was redesignated on December
22, 2014. The Tennessee portion has not been redesignated. The entire area is not considered in
maintenance until all states in a multi-state area are redesignated.
Figure 6-5 I'M: ? Nonattainment Areas
The EPA has already adopted many mobile source emission control programs that are
expected to reduce ambient PM concentrations. As a result of these and other federal, state and
local programs, the number of areas that fail to meet the PM2.5 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 PM2.5 NAAQS well into the future. States will
need to meet the 2006 24-hour standards in the 2015-2019 timeframe and the 2012 primary
annual standard in the 2021-2025 timeframe. The emission reductions and improvements in
ambient PM2.5 concentrations from this action, which will take effect as early as model year
2018, will be helpful to states as they work to attain and maintain the PM2.5 NAAQS.u The
standards can assist areas with attainment dates in 2018 and beyond in attaining the NAAQS as
D The final Phase 2 trailer standards and PM controls for APUs begin with model year 2018.

-------
expeditiously as practicable and may relieve areas with already stringent local regulations from
some of the burden associated with adopting additional local controls.
6.2.1.2 Current Concentrations of Ozone
As described in Chapter6.1, ozone causes adverse health effects, and EPA has set
national ambient air quality standards to protect against those health effects. The primary and
secondary NAAQS for ozone are 8-hour standards with a level of 0.07 ppm. The most recent
revision to the ozone standards was in 2015; the previous 8-hour ozone primary standard, set in
2008, had a level of 0.075 ppm. Nonattainment designations for the 2008 ozone standard were
finalized on April 30, 2012, and May 31, 2012.184 As of April 22, 2016, there were 44 ozone
nonattainment areas for the 2008 ozone NAAQS, composed of 216 full or partial counties, with a
population of more than 120 million. Nonattainment areas for the 2008 ozone NAAQS are
pictured in Figure 6-6. In addition, EPA plans to finalize nonattainment areas for the 2015 ozone
NAAQS in October 2017.
8-Hour Ozone Nonattainment Areas (2008 Standard)
8-hour Ozone Classification
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.
Figure 6-6 8-hour Ozone Nonattainment Areas (2008 Standard)
States with ozone nonattainment areas are required to take action to bring those areas into
attainment. The attainment date assigned to an ozone nonattainment area is based on the area's
classification. Most ozone nonattainment areas were required to attain the 1997 8-hour ozone
NAAQS in the 2007 to 2013 time frame and then to maintain it thereafter. The attainment dates

-------
for areas designated nonattainment for the 2008 8-hour ozone NAAQS are in the 2015 to 2032
timeframe, depending on the severity of the problem in each area. Nonattainment area
attainment dates associated with areas designated for the 2015 NAAQS will be in the 2020-2037
timeframe, depending on the severity of the problem in each area.185
EPA has already adopted many emission control programs that are expected to reduce
ambient ozone levels. 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 emission reductions from this action, which will take effect as
early as model year 2018, will be helpful to states as they work to attain and maintain the ozone
NAAQS.v The standards can assist areas with attainment dates in 2018 and beyond in attaining
the NAAQS as expeditiously as practicable and may relieve areas with already stringent local
regulations from some of the burden associated with adopting additional local controls
6.2.1.3	Current Concentrations of Nitrogen Oxides
EPA most recently completed a review of the primary NAAQS for NO2 in January 2010.
There are two primary NAAQS for NO2: an annual standard (53 ppb) and a 1-hour standard (100
ppb). EPA promulgated area designations in the Federal Register on February 17, 2012. In this
initial round of designations, all areas of the country were designated as
"unclassifiable/attainment" for the 2010 NO2 NAAQS based on data from the existing air quality
monitoring network. EPA and state agencies are working to establish an expanded network of
NO2 monitors, expected to be deployed in the 2013-2017 time frame. Once three years of air
quality data have been collected from the expanded network, EPA will be able to evaluate NO2
air quality in additional locations.186'187
6.2.1.4	Current Concentrations of Sulfur Oxides
EPA most recently completed a review of the primary SO2 NAAQS in June 2010. The
current primary NAAQS for SO2 is a 1-hour standard of 75 ppb. EPA finalized the initial area
designations for 29 nonattainment areas in 16 states in a notice published in the Federal Register
on August 5, 2013. In this first round of designations, EPA only designated nonattainment areas
that were violating the standard based on existing air quality monitoring data provided by the
states. The agency did not have sufficient information to designate any area as "attainment" or
make final decisions about areas for which additional modeling or monitoring is needed (78 FR
47191, August 5, 2013). On March 2, 2015, the U.S. District Court for the Northern District of
California accepted, as an enforceable order, an agreement between the EPA and Sierra Club and
Natural Resources Defense Council to resolve litigation concerning the deadline for completing
designations.w The court's order directs the EPA to complete designations for all remaining
areas in the country in up to three additional rounds: the first round by July 2, 2016, the second
round by December 31, 2017, and the final round by December 31, 2020.
v The final Phase 2 trailer standards begin with model year 2018.
w Sierra Club v. McCarthy, No. 3-13-CV-3953 (SI) (N.D. Cal. Mar. 2, 2015).

-------
6.2.1.5	Current Concentrations of Carbon Monoxide
There are two primary NAAQS for CO: an 8-hour standard (9 ppm) and a 1-hour
standard (35 ppm). The primary NAAQS for CO were retained in August 2011. There are
currently no CO nonattainment areas; as of September 27, 2010, all CO nonattainment areas
have been redesignated to attainment. The designations were based on the existing community-
wide monitoring network. EPA is making changes to the ambient air monitoring requirements
for CO. The new requirements are expected to result in approximately 52 CO monitors
operating near roads within 52 urban areas by January 2015 (76 FR 54294, August 31, 2011).
6.2.1.6	Current Concentrations of Diesel Exhaust PM (DPM)
Because DPM is part of overall ambient PM and cannot be easily distinguished from
overall PM, we do not have direct measurements of DPM in the ambient air. DPM
concentrations are estimated using ambient air quality modeling based on DPM emission
inventories. DPM emission inventories are computed as the exhaust PM emissions from mobile
sources combusting diesel or residual oil fuel. DPM concentrations were recently estimated as
part of the 2011 NATA.188
Concentrations of DPM were calculated at the census tract level in the 2011 NATA.
Figure 6-7 below summarizes the distribution of ambient DPM concentrations at the national
scale. Areas with high concentrations are clustered in the Northeast, Great Lake States,
California, and the Gulf Coast States, and are also distributed throughout the rest of the U.S.
Table 6-1 presents a distribution of ambient DPM concentrations across the country. The
median DPM concentration calculated nationwide is 0.76 [j,g/m3. Half of the DPM can be
attributed to heavy-duty diesel vehicles.

-------
2011 NATA Census Tract Diesel PM Ambient Concentration (|jg/m3)
J


W


W hr IV 1§r -




-jam*


r . M
"T^r t 1


r *

<


"* i V ¦ " v '
*' ' \ ¦ WT- t


- *
t , ¦
i

Diesel PM Cone
(Mg/m3)
^ 1
AK
0.0 - 0.09
0.09 - 0.4

HI
0.4 - 1
¦ 1-3
H 3-8
>8
PR & VI



Figure 6-7 Estimated County Ambient Concentration of Diesel Particulate Matter
Table 6-1 Distribution of Census Tract Ambient Concentrations of DPM at the National Scale in 2011
NATAa

AMBIENT CONCENTRATION
(MG/M3)
5th Percentile
0.15
25th Percentile
0.39
50th Percentile
0.76
75th Percentile
1.24
95th Percentile
2.37
Heavy-Duty Vehicle Contribution to Median Census Tract
Concentrations
50%
Note:
a This table is generated from data contained in the diesel particulate matter Microsoft Access database file found in
the results section of the 2011 NATA webpage (https://www3.epa.gov/national-air-toxics-assessment/2011-nata-
assessment-results#pollutant).

-------
6.2.1.7	Current Concentrations 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.189 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 EPA's most recent Mobile Source Air
Toxics (MSAT) Rule.190 In order to identify and prioritize air toxics, emission source types and
locations which are of greatest potential concern, EPA conducts the National-Scale Air Toxics
Assessment (NATA). The most recent NATA was conducted for calendar year 2011, and was
released in December 2015.191 NATA for 2011 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
4)	Characterizing potential public health risk due to inhalation of air toxics including
both cancer and noncancer effects
According to the NATA for 2011, mobile sources were responsible for 50 percent of
outdoor anthropogenic toxic emissions and were the largest contributor to cancer and noncancer
risk from directly emitted pollutants.x'192 Mobile sources are also large contributors to precursor
emissions which react to form secondary concentrations of air toxics. Formaldehyde is the
largest contributor to cancer risk of all 71 pollutants quantitatively assessed in the 2011 NATA.
Mobile sources were responsible for more than 25 percent of primary anthropogenic emissions
of this pollutant in 2011 and are major contributors to formaldehyde precursor emissions.
Benzene is also a large contributor to cancer risk, and mobile sources account for almost 80
percent of ambient exposure. Over the years, EPA has implemented a number of mobile source
and fuel controls which have resulted in VOC reductions, which also reduced formaldehyde,
benzene and other air toxic emissions.
6.2.1.8	Current Visibility Levels
Designated PM2.5 nonattainment areas indicate that, as of October 1, 2015, over 46 million
people live in nonattainment areas for the PM2.5 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, these areas continue to suffer from visibility impairment.193'194 Calculated from light
extinction efficiencies from Trijonis et al. (1987, 1988), annual average visual range under
natural conditions in the East is estimated to be 150 km ± 45 km (i.e., 65 to 120 miles) and 230
km ± 35 km (i.e., 120 to 165 miles) in the West.195,196'197 In summary, visibility impairment is
x NATA also includes estimates of risk attributable to background concentrations, which includes contributions
from long-range transport, persistent air toxics, and natural sources; as well as secondary concentrations, where
toxics are formed via secondary formation. Mobile sources substantially contribute to long-range transport and
secondarily formed air toxics.

-------
experienced throughout the U.S., in multi-state regions, urban areas, and remote Mandatory
Class I Federal areas.
6,2,1.9 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 25 years. The
data show that reductions were more substantial for sulfur compounds than for nitrogen
compounds. At 34 long-term monitoring sites in the eastern U.S., where data are most abundant,
average total sulfur deposition decreased by 75 percent between 1989-1991 and 2011-2013,
while average total nitrogen deposition decreased by 39 percent over the same time frames.198
Although total nitrogen and sulfur deposition has decreased over time, many areas continue to be
negatively impacted by deposition.
6.2.2 Projected Concentrations of Non-GHG Pollutants
Reductions in emissions of NOx, VOC, PM2.5 and air toxics expected as a result of the
Phase 2 standards will lead to improvements in air quality, specifically decreases in ambient
concentrations of PM2.5, ozone, NO2 and air toxics, as well as better visibility and reduced
deposition.
Emissions and air quality modeling decisions are made early in the analytical process
because of the time and resources associated with full-scale photochemical air quality modeling.
As a result, the inventories used in the air quality modeling and the benefits modeling are
different from the final emissions inventories. The air quality inventories and the final
inventories are consistent in many ways, but there are some important differences which are
discussed in Chapter 6.2.2.3. Chapter 5.5.2.3 of the RIA also has more detail on the differences
between the air quality and final inventories.
6,2,2,1 Air Quality Modeling Results
This section summarizes the results of our air quality modeling, and more detail is available in
Appendix 6. A to the RIA. Specifically, for the year 2040 we compare a reference scenario (a
scenario without the standards) to a control scenario that includes the standards in the air quality
inventory. The standards in the air quality inventory are based on the Phase 2 proposal. As
mentioned above, the inventories used for the air quality modeling and the final inventories are
consistent in many ways but there are some important differences. For example, the air quality
modeling inventory predicted increases in downstream PM2.5 emissions that we do not expect to
occur. The air quality modeling inventory also predicts larger reductions in NOx emissions than
the final inventory. The implications of these differences are noted in the following discussion
of the air quality modeling results.
6.2.2.1.1 Particulate Matter
The air quality modeling indicates that for the majority of the country, annual and 24-
hour PM2.5 design values (DV) will decrease due to these standards. The magnitude of PM2.5

-------
reductions that will actually result from the final standards is difficult to predict because of the
differences between the air quality modeling inventory and the final inventory. However, we do
expect reductions in ambient concentrations of PM2.5, because the final standards will decrease
primary PM2.5, NOx, SOx and VOC emissions.
As described in Section 5.5.2.3, the air quality modeling used inventories that do not
reflect the new requirements for controlling PM2.5 emissions from APUs installed in new tractors
and therefore show increases in downstream PM2.5 emissions that we now do not expect to
occur. Although in most areas this direct PM2.5 increase is outweighed by reductions in
secondary PM2.5, the air quality modeling does predict ambient PM2.5 increases in a few places.
We do not expect these increases in PM2.5 DV to actually occur, because there will be no
increases in downstream PM2.5 emissions. The air quality inventories and the final rule
inventories also have different assumptions about the usage of diesel-powered APUs. The air
quality inventories assumed more widespread usage of diesel-powered APUs than was assumed
for the final rule. As a result, the NOx reductions in the air quality inventories are larger than we
expect to occur, and the air quality modeling overestimates the reductions in ambient PM2.5 due
to secondary nitrate formation.
6.2.2.1.2	Ozone
EPA expects reductions in ambient ozone concentrations due to these final standards. Air
quality modeling results indicate that 8-hour ozone DV will be reduced across the country.
However, the magnitude of the reductions that will actually result from the final standards is
difficult to estimate because the air quality modeling inventories included larger NOx emission
reductions than we now expect to occur. As described in Chapter 5.5.2.3, the air quality
inventories and the final rule inventories make different assumptions about the usage of diesel-
powered APUs. The air quality inventories assumed more widespread usage of diesel-powered
APUs than was assumed for the final rule, and as a result the NOx reductions and 8-hour ozone
reductions are overestimated in the air quality modeling. While we expect the reductions in
upstream and downstream NOx and VOC emissions to result in decreased 8-hour ozone DVs,
the complex and non-linear chemistry governing ozone formation prevents us from estimating
the magnitude without additional air quality modeling.
Maps and summary tables of the projected impacts of the air quality inventories on 8-
hour ozone DV are included in Appendix 6.A.
6.2.2.1.3	Nitrogen Dioxide
EPA expects reductions in ambient nitrogen dioxide (NO2) concentrations due to these final
standards. Air quality modeling results indicate that annual average NO2 concentrations will be
reduced across the country. However, the magnitude of the reductions that will actually result
from the final standards is difficult to estimate because the air quality modeling inventories
included larger NOx emission reductions than we now expect to occur. As described in Chapter
5.5.2.3, the air quality inventories and the final rule inventories make different assumptions
about the usage of diesel-powered APUs. The air quality inventories assumed more widespread
usage of diesel-powered APUs than was assumed for the final rule, and as a result the reductions
in ambient NO2 concentrations are overestimated in the air quality modeling. Appendix 6A

-------
includes maps of absolute and percent change in NO2 concentrations using air quality
inventories.
6.2.2.1.4	Air Toxics
In this section, we describe results of our modeling of air toxics concentrations in 2040 with the
Phase 2 standards included in the air quality inventory. 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 the 2011 NATA assessment and
were also likely to be more significantly impacted by the standards. These compounds include
benzene, 1,3-butadiene, formaldehyde, acetaldehyde, and acrolein.
Our modeling indicates that the standards will have relatively little impact on national average
ambient concentrations of the modeled air toxics. Annual absolute changes in ambient
concentrations are generally less than 0.2 |ig/m3 for benzene, formaldehyde, and acetaldehyde
and less than 0.005 |ig/m3 for acrolein and 1,3-butadiene. Naphthalene changes are in the range
of 0.005 |ig/m3 along major roadways and in urban areas.
Appendix 6A includes air toxics concentration maps as well as population metrics, including the
population living in areas with increases or decreases in concentrations of various magnitudes.
6.2.2.1.5	Visibility
Air quality modeling was used to project visibility conditions in 135 Mandatory Class I
Federal areas across the U.S. The results show that in 2040 all the modeled areas would
continue to have annual average deciview levels above background^ As described in Chapter
5.5.2.3, the air quality modeling used inventories that do not reflect the new requirements for
controlling PM2.5 emissions from APUs installed in new tractors and therefore show increases in
downstream PM2.5 emissions that we now do not expect to occur. Although in most areas this
direct PM2.5 increase is outweighed by reductions in secondary PM2.5, the air quality modeling
does predict visibility to decrease in one area. We do not expect this decrease in visibility to
actually occur, because there will be no increases in downstream PM2.5 emissions. The air
quality inventories and the final rule inventories also have different assumptions about the usage
of diesel-powered APUs. The air quality inventories assumed more widespread usage of diesel-
powered APUs than was assumed for the final rule. As a result, the NOx reductions in the air
quality inventories are larger than we expect to occur, and the air quality modeling overestimates
the reductions in ambient PM2.5 due to secondary nitrate formation. Appendix 6A contains the
full visibility results from 2040 for the 135 analyzed areas.
Y The level of visibility impairment in an area is based on the light-extinction coefficient and a unit less 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.

-------
6.2.2.1.6 Deposition of Nitrogen and Sulfur
Air quality modeling results indicate that nitrogen and sulfur deposition will be reduced
in many areas of the country. The decreases in nitrogen and sulfur deposition are likely due to
the projected reductions in emissions. As described in Chapter 6.2.2.3.1, the NOx reductions
assumed in the air quality inventories are larger than we expect to occur and reductions in
nitrogen deposition are over-estimated in the air quality modeling. While the magnitude of the
reductions in nitrogen deposition from the final rule is difficult to estimate, EPA does expect
reductions in nitrogen deposition due to these final standards.
Maps of the projected impacts of the air quality inventories on nitrogen and sulfur deposition are
included in Appendix 6. A.
6.3 Changes in Atmospheric CO2 Concentrations, Global Mean
Temperature, Sea Level Rise, and Ocean pH Associated with the
Program's GHG Emissions Reductions
6.3.1 Introduction
The impact of GHG emissions on the climate has been reviewed in the 2009
Endangerment and Cause or Contribute Findings for Greenhouse Gases under section 202(a) of
the Clean Air Act, the 2012-2016 light-duty vehicle rulemaking, the 2014-2018 heavy-duty
vehicle GHG rulemaking, and the 2017-2025 light-duty vehicle rulemaking. See 74 FR at
66496; 75 FR at 25491; 76 FR at 57294; 77 FR at 62894. This section briefly discusses again
some of the climate impact context for transportation emissions.
Once emitted, GHGs that are the subject of this regulation can remain in the atmosphere
for decades to millennia, meaning that 1) their concentrations become well-mixed throughout the
global atmosphere regardless of emission origin, and 2) their effects on climate are long lasting.
GHG emissions come mainly from the combustion of fossil fuels (coal, oil, and gas), with
additional contributions from the clearing of forests, agricultural activities, cement production,
and some industrial activities. Transportation activities, in aggregate, were the second largest
contributor to total U.S. GHG emissions in 2010 (27 percent of total emissions).2
EPA Administrator relied on thorough and peer-reviewed assessments of climate change
science prepared by the Intergovernmental Panel on Climate Change ("IPCC"), the United States
Global Change Research Program ("USGCRP"), and the National Research Council of the
National Academies ("NRC")AA as the primary scientific and technical basis for the
Endangerment and Cause or Contribute Findings for Greenhouse Gases under section 202(a) of
the Clean Air Act (74 FR 66496, December 15, 2009). These assessments comprehensively
address the scientific issues EPA Administrator had to examine, providing her both data and
information on a wide range of issues pertinent to the Endangerment Finding. These
z U.S. EPA (2012) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2010. EPA 430-R-12-001.
Available at http://epa.gov/climatechange/emissions/downloadsl2/US-GHG-Inventory-2012-Main-Text.pdf
A A For a complete list of core references from IPCC, USGCRP/CCSP, NRC and others relied upon for development
of the TSD forEPA's Endangerment and Cause or Contribute Findings see Section 1(b), specifically, Table 1.1 of
the TSD. (Docket EPA-HQ-OAR-2010-0799).

-------
assessments have been rigorously reviewed by the expert community, and also by United States
government agencies and scientists, including by EPA itself.
Based on these assessments, EPA Administrator determined that the emissions from new
motor vehicles and engines contributes to elevated concentrations of greenhouse gases, that these
greenhouse gases cause warming; that the recent warming has been attributed to the increase in
greenhouse gases; and that warming of the climate endangers the public health and welfare of
current and future generations. The D.C. Circuit has emphatically upheld the reasonableness of
these findings. Coalition for Responsible Regulation v. EPA. 684 F. 3d 102,121 (D.C. Cir.
2012) upholding all of EPA's findings and stating "EPA had before it substantial record evidence
that anthropogenic emissions of greenhouse gases 'very likely' caused warming of the climate
over the last several decades. EPA further had evidence of current and future effects of this
warming on public health and welfare. Relying again upon substantial scientific evidence, EPA
determined that anthropogenically induced climate change threatens both public health and
public welfare. It found that extreme weather events, changes in air quality, increases in food-
and water-borne pathogens, and increases in temperatures are likely to have adverse health
effects. The record also supports EPA's conclusion that climate change endangers human
welfare by creating risk to food production and agriculture, forestry, energy, infrastructure,
ecosystems, and wildlife. Substantial evidence further supported EPA's conclusion that the
warming resulting from the greenhouse gas emissions could be expected to create risks to water
resources and in general to coastal areas as a result of expected increase in sea level.")
A number of major peer-reviewed scientific assessments have been released since the
administrative record concerning the Endangerment Finding closed following EPA's 2010
Reconsideration Denial199. These assessments include the "Special Report on Managing the
Risks of Extreme Events and Disasters to Advance Climate Change Adaptation"200, the 2013-14
Fifth Assessment Report (AR5)201, the 2014 National Climate Assessment report202, the "Ocean
Acidification: A National Strategy to Meet the Challenges of a Changing Ocean"203, "Report on
Climate Stabilization Targets: Emissions, Concentrations, and Impacts over Decades to
Millennia"204, "National Security Implications for U.S. Naval Forces" (National Security
Implications)205, "Understanding Earth's Deep Past: Lessons for Our Climate Future"206, "Sea
Level Rise for the Coasts of California, Oregon, and Washington: Past, Present, and Future"207,
"Climate and Social Stress: Implications for Security Analysis"208, and "Abrupt Impacts of
Climate Change" (Abrupt Impacts) assessments209.
EPA has reviewed these assessments and finds that in general, the improved
understanding of the climate system they present are consistent with the assessments underlying
the 2009 Endangerment Finding.
The most recent assessments to be released were the IPCC AR5 assessments between
September 2013 and April 2014, the NRC Abrupt Impacts assessment in December of 2013, and
the U.S. National Climate Assessment in May of 2014. The NRC Abrupt Impacts report
examines the potential for tipping points, thresholds beyond which major and rapid changes
occur in the Earth's climate system or other systems impacted by the climate. The Abrupt
Impacts report did find less cause for concern than some previous assessments regarding some
abrupt events within the next century such as disruption of the Atlantic Meridional Overturning
Circulation (AMOC) and sudden releases of high-latitude methane from hydrates and

-------
permafrost, but found that the potential for abrupt changes in ecosystems, weather and climate
extremes, and groundwater supplies critical for agriculture now seem more likely, severe, and
imminent. The assessment found that some abrupt changes were already underway (Arctic sea
ice retreat and increases in extinction risk due to the speed of climate change), but cautioned that
even abrupt changes such as the AMOC disruption that are not expected in this century can have
severe impacts when they happen.
The IPCC AR5 assessments are also generally consistent with the underlying science
supporting the 2009 Endangerment Finding. For example, confidence in attributing recent
warming to human causes has increased: the IPCC stated that it is extremely likely (>95 percent
confidence) that human influences have been the dominant cause of recent warming. Moreover,
the IPCC found that the last 30 years were likely (>66 percent confidence) the warmest 30 year
period in the Northern Hemisphere of the past 1400 years, that the rate of ice loss of worldwide
glaciers and the Greenland and Antarctic ice sheets has likely increased, that there is medium
confidence that the recent summer sea ice retreat in the Arctic is larger than has been in 1450
years, and that concentrations of carbon dioxide and several other of the major greenhouse gases
are higher than they have been in at least 800,000 years. Climate-change induced impacts have
been observed in changing precipitation patterns, melting snow and ice, species migration,
negative impacts on crops, increased heat and decreased cold mortality, and altered ranges for
water-borne illnesses and disease vectors. Additional risks from future changes include death,
injury, and disrupted livelihoods in coastal zones and regions vulnerable to inland flooding, food
insecurity linked to warming, drought, and flooding, especially for poor populations, reduced
access to drinking and irrigation water for those with minimal capital in semi-arid regions, and
decreased biodiversity in marine ecosystems, especially in the Arctic and tropics, with
implications for coastal livelihoods. The IPCC determined that "[cjontinued emissions of
greenhouse gases will cause further warming and changes in all components of the climate
system. Limiting climate change will require substantial and sustained reductions of greenhouse
gases emissions."
Finally, the recently released National Climate Assessment stated, "Climate change is
already affecting the American people in far reaching ways. Certain types of extreme weather
events with links to climate change have become more frequent and/or intense, including
prolonged periods of heat, heavy downpours, and, in some regions, floods and droughts. In
addition, warming is causing sea level to rise and glaciers and Arctic sea ice to melt, and oceans
are becoming more acidic as they absorb carbon dioxide. These and other aspects of climate
change are disrupting people's lives and damaging some sectors of our economy."
Assessments from these bodies represent the current state of knowledge,
comprehensively cover and synthesize thousands of individual studies to obtain the majority
conclusions from the body of scientific literature and undergo a rigorous and exacting standard
of review by the peer expert community and U.S. government.
Based on modeling analysis performed by EPA, reductions in CO2 and other GHG
emissions associated with these final rules will affect future climate change. 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
millennia, depending on the gas. This section provides estimates of the projected change in

-------
atmospheric CO2 concentrations based on the emission reductions estimated for these final rules,
compared to the reference case. In addition, this section analyzes the response to the changes in
GHG concentrations of the following climate-related variables: global mean temperature, sea
level rise, and ocean pH. See Chapter 5 in this RIA for the estimated net GHG emissions
reductions over time.
6.3.2 Projected Change in Atmospheric CO2 Concentrations, Global Mean Surface
Temperature and Sea Level Rise
To assess the impact of the emissions reductions resulting from the final rules, EPA
estimated changes in projected atmospheric CO2 concentrations, global mean surface
temperature and sea-level rise to 2100 using the GCAM (Global Change Assessment Model,
formerly MiniCAM), integrated assessment modelBB'210 coupled with the MAGICC (Model for
the Assessment of Greenhouse-gas Induced Climate Change) simple climate model.cc'211,212
GCAM was used to create the globally and temporally consistent set of climate relevant
emissions required for running MAGICC. MAGICC was then used to estimate the projected
change in relevant climate variables over time. Given the magnitude of the estimated emissions
reductions associated with the rules, a simple climate model such as MAGICC is appropriate for
estimating the atmospheric and climate response.
6,3.2.1 Methodology
Emissions reductions associated with the rules were evaluated with respect to a baseline
reference case. An emissions scenario was developed by applying the estimated emissions
reductions from the final program relative to the baseline to the GCAM reference (no climate
policy) scenario (used as the basis for the Representative Concentration Pathway RCP4.5).213
Specifically, the annual CO2, N2O, CH4, NOx and SO2 emissions reductions estimated from the
final program were applied as net reductions to the GCAM global baseline net emissions for
each substance. The emissions reductions past 2050 for all emissions were scaled with total U.S.
road transportation fuel consumption from the GCAM reference scenario. This was chosen as a
simple scale factor given that both direct and upstream emissions changes are included in the
emissions reduction scenario provided. Road transport fuel consumption past 2050 does not
change significantly and thus emissions reductions remain relatively constant from 2050 through
2100.
BB GCAM 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. GCAM 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.
cc 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 (N20), reactive gases (CO, NOx, VOCs), the halocarbons (e.g. HCFCs, HFCs, PFCs) and
sulfur dioxide (SO2). MAGICC emulates the global-mean temperature responses of more sophisticated coupled
Atmosphere/Ocean General Circulation Models (AOGCMs) with high accuracy.

-------
The GCAM reference scenario214 depicts a world in which global population reaches a
maximum of more than 9 billion in 2065 and then declines to 8.7 billion in 2100 while global
GDP grows by an order of magnitude and global energy consumption triples. The reference
scenario includes no explicit policies to limit carbon emissions, and therefore fossil fuels
continue to dominate global energy consumption, despite substantial growth in nuclear and
renewable energy. Atmospheric CO2 concentrations rise throughout the century and reach 760
to 820 ppmv by 2100, depending on climatic parameters, with total radiative forcing increasing
more than 5 Watts per square meter (W/m2) above 1990 levels by 2100. Forest land declines in
the reference scenario to accommodate increases in land use for food and bioenergy crops. Even
with the assumed agricultural productivity increases, the amount of land devoted to crops
increases in the first half of the century due to increases in population and income (higher
income drives increases in land-intensive meat consumption). After 2050 the rate of growth in
food demand slows, in part due to declining population. As a result the amount of cropland and
also land use change (LUC) emissions decline as agricultural crop productivity continues to
increase.
The GCAM reference scenario uses non-CC>2 and pollutant emissions implemented as
described in Smith and Wigley (2006); land-use change emissions as described in Wise et al.
(2009); and updated base-year estimates of global GHG emissions. This scenario was created as
part of the Climate Change Science Program (CCSP) effort to develop a set of long-term global
emissions scenarios that incorporate an update of economic and technology data and utilize
improved scenario development tools compared to the IPCC Special Report on Emissions
Scenarios (SRES) (IPCC 2000).
Using MAGICC 5.3 v2,215 the change in atmospheric CO2 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 the final program.
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 a range of
plausible climate sensitivities, 1.5°C to 6.0°C.DD The range as illustrated in Chapter 10, Box
10.2, Figure 2 of the IPCC's Working Group I is approximately consistent with the 10-90
percent probability distribution of the individual cumulative distributions of climate
sensitivity.216 Other uncertainties, such as uncertainties regarding the carbon cycle, ocean heat
uptake, or aerosol forcing, were not addressed.
MAGICC calculates the forcing response at the global scale from changes in atmospheric
concentrations of CO2, CH4, N2O, HFCs, and tropospheric ozone. It also includes the effects of
temperature changes on stratospheric ozone and the effects of CH4 emissions on stratospheric
water vapor. Changes in CH4, NOx, VOC, and CO emissions affect both O3 concentrations and
CH4 concentrations. MAGICC includes the relative climate forcing effects of changes in sulfate
DD 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 most
recent IPCC AR5 assessment states that climate sensitivity is "likely" to be in the range of 1.5°C to 4.5°C,
"extremely unlikely" to be less than 1°C, and "very unlikely" to be greater than 6 °C." Intergovernmental Panel on
Climate Change (IPCC). 2013. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I
to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA.

-------
concentrations due to changing SO2 emissions, including both the direct effect of sulfate
particles and the indirect effects related to cloud interactions. However, MAGICC does not
calculate the effect of changes in concentrations of other aerosols such as nitrates, black carbon,
or organic carbon, making the assumption that the sulfate cooling effect is a proxy for the sum of
all the aerosol effects. Therefore, the climate effects of changes in PM2.5 emissions and
precursors (besides SO2) presented in Chapter 5 were not included in the calculations in this
chapter. MAGICC also calculates all climate effects at the global scale. This global scale
captures the climate effects of the long-lived, well-mixed greenhouse gases, but does not address
the fact that short-lived climate forcers such as aerosols and ozone can have effects that vary
with location and timing of emissions. Black carbon in particular is known to cause a positive
forcing or warming effect by absorbing incoming solar radiation, but there are uncertainties
about the magnitude of that warming effect and the interaction of black carbon (and other co-
emitted aerosol species) with clouds. See 77 FR 38890, 38991-993 (June 29, 2012). While black
carbon is likely to be an important contributor to climate change, it would be premature to
include quantification of black carbon climate impacts in an analysis of the final standards at this
time. See generally, EPA, Response to Comments to the Endangerment Finding Vol. 9 Section
9.1.6.1, the discussion of black carbon in the endangerment finding at 74 FR at 66520, EPA's
discussion in the recent proposal to revise the PM NAAQS (77 FR at 38991-993), and the
recently published EPA Report to Congress on Black Carbon. Additionally, the magnitude of
PM2.5 emissions changes (and therefore, black carbon emission changes) related to these
standards are small in comparison to the changes in the pollutants which have been included in
the MAGICC model simulations.
To compute the changes in atmospheric CO2 concentration, global mean temperature,
and sea level rise specifically attributable to the impacts of the standards, the difference in
emissions between the final program and the baseline scenario was subtracted from the GCAM
reference emissions scenario. As a result of the final program's emissions reductions relative to
the baseline case, by 2100 the concentration of atmospheric CO2 is projected to be reduced by
approximately 1.2 to 1.3 parts per million by volume (ppmv), the global mean temperature is
projected to be reduced by approximately 0.0027 to 0.0065°C, and global mean sea level rise is
projected to be reduced by approximately 0.026 to 0.058 cm. For sea level rise, the calculations
in MAGICC do not include the possible effects of accelerated ice flow in Greenland and/or
Antarctica; including these effects would show correspondingly larger benefits of mitigation.
Figure 6-8 provides the results over time for the estimated reductions in atmospheric CO2
concentration associated with the final program compared to the baseline scenario. Figure 6-9
provides the estimated change in projected global mean temperatures associated with the final
program. Figure 6-10 provides the estimated reductions in global mean sea level rise associated
with the final program. The range of reductions in global mean temperature and sea level rise
due to uncertainty in climate sensitivity is larger than that for CO2 concentrations because CO2
concentrations are only weakly coupled to climate sensitivity through the dependence on
temperature of the rate of ocean absorption of CO2, whereas the magnitude of temperature
change response to CO2 changes (and therefore sea level rise) is more tightly coupled to climate
sensitivity in the MAGICC model.

-------
>
E
Q.
Q.
Change in C02 Concentration
(Final Rule - Baseline)
0.0 ]

-0.2


-0.4
—CS 1.5
-0.6
" -^CS 2.0
-0.8
- -^CS 2.5
-1.0
_ ->^CS 3.0
-1.2
—W^CS 4.5
—CS 6.0
-1.4

2000
2020
2040
2060
2080
2100
Figure 6-8 Estimated Projected Reductions in Atmospheric CO2 Concentrations (parts per million by
volume) from the Baseline for the Heavy-Duty Final Program (climate sensitivity (CS) cases ranging from
1.5-6°C)
a>
a>
bjo

-------
Change in Global Mean Sea Level Rise
(Final Rule - Baseline)
£ -0.03
u
-0.04
-0.06
-0.05
-0.01
-0.02
0.00
CS 6.0
-0.07
2000
2020
2040
2060
2080
2100
Figure 6-10 Estimated Projected Reductions in Global Mean Sea Level Rise from the Baseline for the Heavy-
Duty Final Program (climate sensitivity (CS) cases ranging from 1.5-6°C)
The results in Figure 6-9 and Figure 6-10 show reductions in the projected global mean
temperature and sea level respectively, across all climate sensitivities. The projected reductions
are small relative to the change in temperature (1.8 - 4.8 °C) and sea level rise (23 - 56 cm) from
1990 to 2100 from the MAGICC simulations for the GCAM reference case. However, this is to
be expected given the magnitude of emissions reductions expected from the rules in the context
of global emissions. These reductions are quantifiable, directionally consistent, and will
contribute to reducing the risks associated with climate change. Notably, these effects are
occurring everywhere around the globe, so benefits that appear to be marginal for any one
location, such as a reduction in sea level rise of half a millimeter, can be sizable when the effects
are summed along thousands of miles of coastline. Climate change is a global phenomenon and
EPA recognizes that this one national action alone will not prevent it; EPA notes this would be
true for any given GHG mitigation action when taken alone or when considered in isolation.
EPA also notes that a substantial portion of CO2 emitted into the atmosphere is not removed by
natural processes for millennia, and therefore each unit of CO2 not emitted into the atmosphere
due to this rule avoids essentially permanent climate change on centennial time scales. Again, it
should be noted that the calculations in MAGICC do not include the possible effects of
accelerated ice flow in Greenland and/or Antarctica: the recent NRC report estimated a likely sea
level increase for the A1B SRES scenario of 0.5 to 1.0 meters, almost double the estimate from
MAGICC, so projected reductions in sea level rise may be similarly underestimated.217 If other
uncertainties besides climate sensitivity were included in the analysis, the resulting ranges of
projected changes would likely be slightly larger.
6.3.3 Estimated Projected Change in Ocean pH
For these rules, EPA analyzes another key climate-related variable and calculates
projected change in ocean pH for tropical waters. For this analysis, changes in ocean pH are
related to the change in the atmospheric concentration of carbon dioxide (CO2) resulting from
the emissions reductions associated with the final program. EPA used the program developed

-------
for CO2 System Calculations C02SYS,218 version 1.05, a program which performs calculations
relating parameters of the carbon dioxide (CO2) system in seawater. The program was
developed by Ernie Lewis at Brookhaven National Laboratory and Doug Wallace at the Institut
fur Meereskunde in Germany, supported by the U.S. Department of Energy, Office of Biological
and Environmental Research, under Contract No. DE-AC02-76CH00016.
The C02SYS program uses two of the four measurable parameters of the CO2 system
[total alkalinity (TA), total inorganic CO2 (TC), pH, and either fugacity (fCCh) or partial
pressure of CO2 (pCCh)] to calculate the other two parameters given a specific set of input
conditions (temperature and pressure) and output conditions chosen by the user. EPA utilized
the Excel version (Pierrot et al. 2006)219 of the program to compute pH for three scenarios: the
baseline scenario at a climate sensitivity of 3 degrees for which the CO2 concentrations was
calculated to be 784.87 in 2100, the final program relative to the baseline with a CO2
concentration of 783.62, and a calculation for 1990 with a CO2 concentration of 353.63.
Using the set of seawater parameters detailed below, EPA calculated pH levels for the
three scenarios. The pH of the emissions standards relative to the baseline scenario pH was
+0.0006 units (more basic). For comparison, the difference between the baseline scenario in
2100 and the pH in 1990 was -0.30 pH units (more acidic).
The C02SYS program required the input of a number of variables and constants for each
scenario for calculating the result for both the reference case and the final program's emissions
reduction baseline cases. EPA used the following inputs, with justification and references for
these inputs provided in brackets:
1)	Input mode: Single-input
2)	Choice of constants: Mehrbach et al. (1973)220, refit by Dickson and Millero (1987)221
3)	Choice of fCCh or pCCh: pCCh
4)	Choice of KS04: Dickson (1990)222 Choice of KS04: Dickson (1990)223
5)	Choice of pH scale: Total scale Choice of pH scale: Total scale
6)	[B]t value: Uppstrom, 1974
The program provides several choices of constants for saltwater that are needed for the
calculations. EPA calculated pH values using all choices and found that in all cases the choice
had an indistinguishable effect on the results. In addition, EPA ran the model using a variety of
other required input values to test whether the model was sensitive to these inputs. EPA found
the model was not sensitive to these inputs in terms of the incremental change in pH calculated
for each climate sensitivity case. The input values are derived from certified reference materials
of sterilized natural sea water (Dickson, 2003, 2005, and 2009).224 Based on the projected
atmospheric CO2 concentration reductions that would result from the final program's baseline
case (1.3 ppmv for a climate sensitivity of 3.0), the modeling program calculates an increase in
ocean pH of approximately 0.0006 pH units in 2100. Thus, this analysis indicates the projected
decrease in atmospheric CO2 concentrations from the final program yields an increase in ocean
pH. Table 6-2 contains the projected changes in ocean pH based the change in atmospheric CO2
concentrations which were derived from the MAGICC modeling.

-------
Table 6-2 Impact of the Rule's GHG Emissions Reductions on Ocean pH
CLIMATE
SENSITIVITY
DIFFERENCE
INC02a
YEAR
PROIECTED
CHANGE
3.0
-1.3 ppmv
2100
0.0006
Note:
" Represents the change in atmospheric CO2 concentrations in 2100 based on the difference
from the rule relative to the base case from the GCAM reference scenario used in the
MAGICC modeling.
6.3.4 Summary of Climate Analyses
EPA's analysis of the impact of the final program's emissions reductions on global
climate conditions is intended to quantify these potential reductions using the best available
science. While EPA's modeling results of the impact of the final program alone show small
differences in climate effects (CO2 concentration, global mean temperature, sea level rise, and
ocean pH), in comparison to the total projected changes, they yield results that are repeatable and
directionally consistent within the modeling frameworks used. The results are summarized in
Table 6-3, Impact of GHG Emissions Reductions on Projected Changes in Global Climate
Associated with the Final Program.
These projected reductions are proportionally representative of changes to U.S. GHG
emissions in the transportation sector. While not formally estimated for this final program, a
reduction in projected global mean temperature and sea level rise implies a reduction in the risks
associated with climate change. The figures for these variables illustrate that across a range of
climate sensitivities projected global mean temperature and sea level rise increase less in the
final program scenario than in the reference (no climate policy) case. The benefits of GHG
emissions reductions can be characterized both qualitatively and quantitatively, some of which
can be monetized (see Chapter 9). 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).
Table 6-3 Impact of GHG Emissions Reductions on Projected Changes in Global Climate Associated with the
Final Program (Based on a Range of Climate Sensitivities from 1.5-6°C)
VARIABLE
UNITS
YEAR
PROIECTED CHANGE
Atmospheric CO2
Concentration
ppmv
2100
-1.2 to -1.3
Global Mean Surface
Temperature
°C
2100
-0.0027 to -0.0065
Sea Level Rise
cm
2100
-0.026 to-0.058
Ocean pH
pH units
2100
+0.00063
Note:
a The value for projected change in ocean pH is based on a climate sensitivity of 3.0.

-------
Appendix 6.A to Chapter 6 - Air Quality Modeling Results
6A.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 air quality
modeling methodology is included in the Air Quality Modeling Technical Support Document
(AQM TSD) found in the docket for this rule.
6A1.1 Air Quality Modeling Analysis Overview
A national-scale air quality modeling analysis was performed to estimate future year 8-
hour ozone concentrations, annual PM2.5 concentrations, 24-hour PM2.5 concentrations, annual
NO2 concentrations, air toxics concentrations, visibility levels and nitrogen and sulfur deposition
levels for 2040. The 2011-based CMAQ modeling platform was used as the basis for the air
quality modeling for this 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 2011. 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 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.225'226'227 The CMAQ model version 5.1, which was an
upcoming new community version in late 2015, was most recently peer-reviewed in September
of 2015 for the U.S. EPA.228 The CMAQ model is a well-known and well-respected tool and has
been used in numerous national and international applications.229'230'231 This 2011 multi-
pollutant modeling platform used the most recent multi-pollutant CMAQ code available at the
time of air quality modeling (CMAQ version 5.0.2 multi pollutant versionEE).
EE CMAQ version 5.0.2 was released in April 2014. It is available from the Community Modeling and Analysis
System (CMAS) as well as previous peer-review reports at: http://www.cmascenter.org.

-------
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 v5.0.2 which reflects updates to version 5.0 to improve the
underlying science algorithms as well as include new diagnostic/scientific modules which are
detailed at http:www.cmascenter.org.232-233-234 Chapter 6A1.6 of this RIA discusses the chemical
mechanism and SOA formation.
6A1.2 Model Domain and Configuration
The CMAQ modeling domain encompasses all of the lower 48 states and portions of
Canada and Mexico, see Figure 6A-1. The modeling domain is made up of a 12 kilometer (km)
grid and contains 25 vertical layers with the top of the modeling domain at about 17,600 meters,
or 50 millibars (mb) of atmospheric pressure.
12US2 domain \ V \
x,y origin: -2412000iri,^1621
col: 396 row:246 / ,

Figure 6A-1 Map of the CMAQ 12-km US Modeling Domain
6A1.3 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 Weather
Research and Forecasting Model (WRF) version 3.4, Advanced Research WRF (ARW) core 235
for the entire year of 2011 over model domains that are slightly larger than those shown in
Figure 6A-1. The WRF Model is a next-generation mesoscale numerical weather prediction
system developed for both operational forecasting and atmospheric research applications
(http://wrf-model.org). The meteorology for the national 12 km grid was developed by EPA and
are described in more detail within the AQM TSD. The meteorological outputs from WRF were
processed to create model-ready inputs for CMAQ using the Meteorology-Chemistry Interface
Processor (MCIP) version 4.1.3. Outputs include: horizontal wind components (i.e., speed and
direction), temperature, moisture, vertical diffusion rates, and rainfall rates for each grid cell in
each vertical layer.236 The 2011 CMAQ meteorological inputs will be derived from Version 3.4
of the Weather Research Forecasting Model (WRF).237 These inputs included hourly-varying
horizontal wind components (i.e., speed and direction), temperature, moisture, vertical diffusion
rates, and rainfall rates for each grid cell in each vertical layer. Details of the annual 2011
meteorological model simulation and evaluation will be described in more detail within the air
quality modeling technical support document.
The lateral boundary and initial species concentrations are provided by a three-
dimensional global atmospheric chemistry model, the GEOS-Chem model238 (standard version
8-03-02 with 8-02-01 chemistry). The global GEOS-Chem model simulates atmospheric
chemical and physical processes driven by assimilated meteorological observations from the
NASA's Goddard Earth Observing System (GEOS-5; additional information available at:
http://gmao.gsfc.nasa.gov/GEOS/ and http://wiki.seas.harvard.edu/geos-chem/index.php/GEOS-
5). This model was run for 2011 with a grid resolution of 2.0 degrees x 2.5 degrees (latitude-
longitude). The predictions were used to provide one-way dynamic boundary conditions at one-
hour intervals and an initial concentration field for the CMAQ simulations. A GEOS-Chem
evaluation was conducted for the purpose of validating the 2011 GEOS-Chem simulation for
predicting selected measurements relevant to their use as boundary conditions for CMAQ. This
evaluation included using satellite retrievals paired with GEOS-Chem grid cells.239 More
information is available about the GEOS-Chem model and other applications using this tool
at: http://acmg. seas.harvard.edu/geos.
The emissions inputs used for the 2011 base year and 2040 reference and control
scenarios analyzed for this rule are summarized in Chapter 5 of this RIA and described in more
detail in the Emission Inventories for Air Quality Modeling Technical Support Document (IAQ
TSD).
6A1.4 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 2011 state/local monitoring data in order to estimate the ability of
the CMAQ modeling system to replicate base year concentrations. The evaluation included
statistical measures of model performance based upon model-predicted versus observed
concentrations that were paired in space and time. Model performance statistics were calculated
for several spatial scales and temporal periods. Statistics were calculated for individual

-------
monitoring sites and for each of nine climate regions of the 12-km U.S. modeling domain. The
regions include the Northeast, Ohio Valley, Upper Midwest, Southeast, South, Southwest,
Northern Rockies, Northwest and West11, which are defined based upon the states contained
within the National Oceanic and Atmospheric Administration (NOAA) climate regions as were
originally identified in Karl and Koss (1984).240
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.00 Overall, the
performance for the 2011 modeling platform is within the range or close to that of these other
applications. 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 2011 CMAQ model
performance evaluation is available within the AQM TSD found in the docket of this rule.
6A1.5 Model Simulation Scenarios
As part of our analysis for this rulemaking, the CMAQ modeling system was used to
calculate 8-hour ozone concentrations, daily and annual PM2.5 concentrations, annual NO2
concentrations, annual and seasonal (summer and winter) air toxics concentrations, visibility
levels and annual nitrogen and sulfur deposition total levels for each of the following emissions
scenarios:
-	2011 Base year
-	2040 Phase 2 reference case
-	2040 Phase 2 control case
As mentioned above, the inventories used for the air quality modeling and the final
inventories are consistent in many ways but there are some important differences. For example,
EPA is adopting Phase 1 and Phase 2 requirements to control PM2.5 emissions from APUs
installed in new tractors, therefore we do not expect increases in PM2.5 emissions from the Phase
2 program; however, the air quality inventories do not reflect these requirements and therefore
show increases in downstream PM2.5 emissions. Chapter 5.5.2.3 of the RIA has more detail on
the differences between the air quality and final inventories. The IAQ TSD, found in the docket
for this rule (EPA-HQ-OAR-2014-0827), also contains a detailed discussion of the emissions
inputs used in our air quality modeling.
FF The nine climate regions are defined by States where: Northeast includes CT, DE, ME, MA, MD, NH, NJ, NY,
PA, RI, and VT; Ohio Valley includes IL, IN, KY, MO, OH, TN, and WV; Upper Midwest includes IA, MI, MN,
and WI; Southeast includes AL, FL, GA, NC, SC, and VA; South includes AR, KS, LA, MS, OK, and TX;
Southwest includes AZ, CO, NM, and UT; Northern Rockies includes MT, NE, ND, SD, WY; Northwest includes
ID, OR, and WA; and West includes CA and NV. Note most monitoring sites in the West region are located in
California, therefore for the West will be mostly representative of California ozone air quality.
GG 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.

-------
We use the predictions from the model in a relative sense by combining the 2011 base-
year predictions with predictions from each future-year scenario and applying these modeled
ratios to ambient air quality observations to estimate 8-hour ozone concentrations, daily and
annual PM2.5 concentrations, annual NO2 concentrations and visibility impairment for each of
the 2040 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., 2009-2013).
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 |ig/m3). More complete details of the 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)."241 For this latest analysis,
several datasets and techniques were updated. These changes are fully described within the
technical support document for the Final Transport Rule AQM TSD.242 The projected 8-hour
ozone design values were calculated using the approach identified in EPA's guidance on air
quality modeling attainment demonstrations.243
Additionally, we conducted an analysis to compare the absolute and percent differences
between the future year reference and control cases for annual and seasonal formaldehyde,
acetaldehyde, benzene, 1,3-butadiene, naphthalene, 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.
6A1.6 Chemical Mechanisms in Modeling
This analysis looks at air quality impacts of criteria pollutants including NOx, VOC, CO,
PM2.5, SO2, and air toxics, specifically benzene, 1,3-butadiene, formaldehyde, acetaldehyde,
naphthalene and acrolein. The air toxics were added as explicit model species to the carbon
bond 5 (CB05) mechanisms used in CMAQv5.0.1.244 Emissions of all the pollutants included in
the rule inventories, except ethanol, were generated using the Motor Vehicle Emissions
Simulator (MOVES) VOC emissions and toxic-to-VOC ratios calculated using EPAct data.245
Ethanol emissions for air quality modeling were based on speciation of VOC using different
ethanol profiles (E0, E10 and E85) (see Inventory for Air Quality Modeling Technical Support
Document for more information). 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 5.3.5.
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.246
Complete combustion of ethanol in fuel produces carbon dioxide (CO2) and water
(H20). 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, alkenes and other
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 other
hydrocarbons, particularly alkenes.
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 (HN03) without PAN formation.™
NO2 + "OH + M —~ HN03 + M	k=1.19x 10-11 cm3molecule-ls-l
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 alkenes, alkanes, and
aromatics. 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.
6A1.6.1 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(0)00-].n When NOx is present in the atmosphere this radical species can then further
react with nitric oxide (NO), to produce formaldehyde (HCHO), or with nitrogen dioxide (NO2),
1111 All rate coefficients are listed at 298 K and, if applicable, 1 bar of air.
11 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).

-------
to produce PAN [CH3C(0)00N02]. An overview of these reactions and the corresponding
reaction rates are provided below.JJ
CH3CHO + -OH -> CH3CO + H20 k = 1.5 x 10"11 cir^molecule'V1 247
CH3CO + O2 + M —~ CH3C(0)00- + M
CH3C(0)00- + NO -> CH3C(0)0- + NO 2	k = 2.0 X 10"11 cn^moleculeV1 248
CH3C(0)O -> -ch3 + co2
•CH3 + 02 + M -> CH300- + M
CH300- + NO -> CH30- + N02
CH30- + 02 -> HCHO + H02
CH3C(0)00- +NO2 + M -> CH3C(0)00N02 + M k = 1.0 X 10"11 cn^moleculeV
1 249
Acetaldehyde can react with the N03 radical, ground state oxygen atom (03P) and
chlorine, although these reactions are much slower. Acetaldehyde can also photolyze (hv),
which predominantly produces -CH3 (which reacts as shown above to form CH30O) and HCO
(which rapidly forms HO2 and CO):
CHsCHO + hv +2 02 -> CH300- +HO2 + CO	X = 240-380 nm 250
As mentioned above, CH300- can react in the atmosphere to produce formaldehyde
(HCHO). Formaldehyde is also a product of hydrocarbon combustion. In the atmosphere, the
most important reactions of formaldehyde are photolysis and reaction with the OH, with
atmospheric lifetimes of approximately 3 hours and 13 hours, respectively.251 Formaldehyde can
also react with N03 radical, ground state oxygen atom (03P) and chlorine, although these
reactions are much slower. 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 an important source of new hydroperoxy radicals
(HO2), which can lead to ozone formation and regenerate OH radicals.
HCHO + hv + 2 02 -> 2 H02 + CO	X = 240-360 nm 252
ho2+no^no2+ oh
Photolysis of HCHO can also proceed by a competing pathway which makes only stable
products: H2 and CO.
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
11 All rate coefficients are listed at 298 K and, if applicable, 1 bar of air.

-------
acetaldehyde in the air. Acetaldehyde is represented explicitly in the CB05 chemical
mechanism253'254 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.255
The decay reactions of acetaldehyde are fewer in number and can be characterized well
because they are explicit representations. In CB05, acetaldehyde can photolyze or react with
molecular oxygen (O (3P)), hydroxyl radical (OH), or nitrate radicals. The reaction rates are
based on expert recommendations,256 and the photolysis rate is from IUPAC recommendations.
In CMAQ v5.0, 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 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.257 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 can in some instances increase acetaldehyde, but because they also are a sink of
radicals, their effect is smaller. 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 (CH3CH2OH) 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).258 This reaction produces
acetaldehyde (CH3CHO) with a 90 percent yield.259 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. ^ For example, an
atmospheric lifetime for acetaldehyde under nominal oxidant conditions, OH of 1.0 x 10"
6 cir^molecule'V1, would be 3.5 days.
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
KK All rate coefficients are listed at 298 K and, if applicable, 1 bar of air.

-------
formaldehyde (HCHO), oxidizing NO to NO2 (represented by model species XO2) and creating
glycoaldehyde, which is represented as ALDX:
CH3CHOH + OH -> H02 + 0.90 CH3CHO + 0.05 ALDX + 0.10 HCHO + 0.10 X02
6A1.6.2 Organic Aerosols
Organic aerosol (OA) can be classified as either primary or secondary depending on
whether it is emitted into the atmosphere as a particle (primary organic aerosol, POA) or formed
in the atmosphere (SOA). SOA precursors include volatile organic compounds (VOCs) as well
as low-volatility compounds that can react to form even lower volatility compounds. Current
research suggests SOA contributes significantly to ambient OA concentrations, and in Southeast
and Midwest States may make up more than 50 percent (although the contribution varies from
area to area) of the organic fraction of PM2.5 during the summer (but less in the winter).260'261 A
wide range of laboratory studies conducted over the past twenty years show that anthropogenic
aromatic hydrocarbons and long-chain alkanes, along with biogenic isoprene, monoterpenes, and
sesquiterpenes, contribute to SOA formation.262'263'264'265'266 Modeling studies, as well as carbon
isotope measurements, indicate that a significant fraction of SOA results from the oxidation of
biogenic hydrocarbons.267'268 Based on parameters derived from laboratory chamber
experiments, SOA chemical mechanisms have been developed and integrated into air quality
models such as the CMAQ model and have been used to predict OA concentrations.269
Secondary organic aerosol (SOA) chemistry in CMAQ v5.0 is largely based on
recommendations of Edney et al. (2007) and Carlton et al. (2008) as initially implemented in
CMAQ v4.7.270'271'272 In previous versions of CMAQ, all SOA was semivolatile and resulted
from the oxidation of compounds emitted entirely in the gas-phase. Starting with CMAQ v4.7,
parameters in existing pathways were revised and new formation mechanisms were added.
Some of the new pathways, such as low-NOx oxidation of aromatics and particle-phase
oligomerization, result in nonvolatile SOA.
New to CMAQ v5.0 is the heterogeneous oxidation of primary organic aerosol (POA).273
Specifically, primary organic aerosol is tracked separately in terms of its carbon and non-carbon
organic matter. Non-carbon organic matter (such as oxygen and hydrogen) is added to the
reduced carbon as a result of heterogeneous reaction with OH. Diesel POA is emitted with an
organic matter to organic carbon (OM/OC) ratio of 1.25. The ratio increases due to exposure
with OH. In the absence of removal, this oxidation process results in increasing organic aerosol
concentrations. These OM/OC ratios assist with post-processing of model output for comparison
with measured OC from routine networks.
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,
currently accounting for almost 40 percent of anthropogenic VOC,274 mobile sources are also an
important source of SOA, particularly in populated areas.

-------
Toluene is an important contributor to anthropogenic SOA.275'276 Mobile sources are the
most significant contributor to ambient toluene concentrations as shown by analyses done for the
2011 National Air Toxics Assessment (NATA)277 and the Mobile Source Air Toxics (MSAT)
Rule.278 The 2011 NATA indicates that onroad and nonroad mobile sources accounted for
around 50 percent (1.35 |ig/m3) of the total average nationwide ambient concentration of toluene
(2.61 |ig/m3.
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.
In addition to toluene, other mobile-source hydrocarbons such as benzene, xylene, and
alkanes form SOA. Similar to toluene, the SOA produced by benzene and xylene from low-NOx
pathways is expected to be less volatile and be produced in higher yields than SOA from high-
NOx conditions.279 Oxidation of alkanes with longer chains as well as cyclic alkanes form SOA
with relatively higher yields than small straight-chain alkanes.280
It is unlikely that ethanol would form SOA directly or affect SOA formation indirectly
through changes in the radical populations due to increasing ethanol exhaust. Nevertheless,
scientists at the U.S. EPA's Office of Research and Development recently directed experiments
to investigate ethanol's SOA forming potential.281 The experiments were conducted under
conditions where peroxy radical reactions would dominate over reaction with NO (i.e.,
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 could form. As expected, no SOA was produced. From these
experiments, the upper limit for the aerosol yield is less than 0.01 percent based on scanning
mobility particle sizer (SMPS) data. Given the lack of aerosol formation found in these initial
smog chamber experiments, these data were not published.
In general, measurements of OA represent the sum of POA and SOA and the fraction of
aerosol that is secondary in nature can only be estimated. One of the most widely applied method
of estimating total ambient SOA concentrations is the EC tracer method using ambient data
which estimates the OC/EC ratio in primary source emissions.282'283 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 POA from the
measured OC concentration.284 Aerosol mass spectrometer (AMS) measurements along with
positive matrix factorization (PMF) can also be used to identify surrogates for POA and SOA in
ambient as well as chamber experiments. Such methods, however, may not be quantitatively
accurate and provide limited 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 complete compositional data from
the precursors has made the identification of aromatically-derived SOA in ambient samples
challenging, which in turn has prevented observation-based measurements of individual SOA
source contributions to ambient PM levels.
As a first step in estimating ambient SOA concentrations, EPA has developed a tracer-
based method.285'286 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 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 SOA chemical formation mechanisms.287,288'289'290
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 limited 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.291 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, ^-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 environmental
smog chambers. These unstudied compounds could produce SOA species that are being used as
tracers for other VOCs thus overestimating 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.292'293 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.

-------
The aromatic tracer compounds and their mass fractions have been used to estimate
monthly ambient aromatic SOA concentrations from March 2004 to February 2005 in five U.S.
Midwestern cities.294 The annual tracer-based SOA concentration estimates were 0.15, 0.18,
0.13, 0.15, and 0.19 jag 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 percent of the total SOA
concentration. Thus, this work suggests that we are finding ambient PM levels on an annual
basis of about 0.15 [j,g/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 xylenes.295
Over the past decade a variety of modeling studies have been conducted to predict
ambient SOA levels. While early studies focused on the contribution of biogenic monoterpenes,
additional precursors, such as sesquiterpenes, isoprene, benzene, toluene, and xylene, have been
implemented in atmospheric models such as GEOS-Chem, PMCAMx, and
CMAQ.296'291'298'299'300-301-302 Studies have indicated that ambient OC levels may be
underestimated by current model parameterizations.303 In general, modeling studies focus on
comparing the sum of the POA and SOA concentrations with ambient OC or estimated OA
concentrations. Without a method to attribute measured OC to different sources or precursors,
identifying causes of the underestimates in modeled OC via model/measurement comparisons
can be challenging. However, analysis of SOA concentrations in Pasadena and Bakersfield,
California during 2010 indicate CMAQ-predicted SOA from toluene and xylene is
underestimated despite overestimates of the VOC precursors.304 In addition, CMAQ-predicted
aromatic SOA was underestimated in the Midwest US despite reasonable predictions of primary
organic aerosol tracers, implying underestimated SOA yields.305
6A1.6.3 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
downwind of NOx sources. PAN is a reservoir and carrier of NOx and is the product of acetyl
radicals reacting with NO2 in the atmosphere. One source of PAN is the photooxidation of
acetaldehyde, but many VOCs have the potential for forming acetyl radicals and therefore PAN
or a PAN-type compound.LL PAN can undergo thermal decomposition with a lifetime of
approximately 1 hour at 298K or 148 days at 250K. MM
CH3C(0)00N02 + M -> CH3C(0)00- + NOi + M	k = 3.3 x 10"4 s"1 306
The reaction above shows how NO2 is released in the thermal decomposition of PAN,
along with a peroxy radical which can oxidize NO to NO2 and form other species that convert
LL Many aromatic hydrocarbons, particularly those present in high percentages in gasoline (toluene, m-, 0-, p-xylene,
and 1,3,5-, 1,2,4-trimethylbenzene), form methylglyoxal and biacetyl, 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
withm-, p-xylene and 1,2,4- and 1,3,5-Trimethylbenzene. J. Atmos. Chem., 34: 339- 364).
1414 All rate coefficients are listed at 298 K and, if applicable, 1 bar of air.

-------
NO to N02 through photochemical reactions, as previously shown in Chapter 6.2.2.2.1. NO2
further photolyzes to produce ozone (O3).
NOi + hv -> NO + 0(3P)	X = 300-800 nm 307
0(3P) + 02 + M^ Os + 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.308 A discussion of CB05
mechanisms for ozone formation can be found in Yarwood et al. (2005).309
Another important way that ethanol fuels contribute to ozone formation is by increasing
the formation of new radicals through increases in formaldehyde and acetaldehyde. The
photolysis of both aldehydes results in up to two molecules of either hydroperoxy radical or
methylperoxy radical, both of which oxidize NO to NO2 leading to ozone formation.
6A1.6.4 Uncertainties Associated with Chemical Mechanisms
A key source of uncertainty with respect to the air quality modeling results is the
photochemical mechanisms in CMAQ. Pollutants such as ozone, PM, acetaldehyde,
formaldehyde, and acrolein 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 previous Chapters
6A. 1.6.1 through 6A1.6.3.
For PM, there are a number of uncertainties associated with SO A formation that should
be addressed explicitly. As mentioned in Chapter 6A.1.6.2, a large number of VOCs emitted
into the atmosphere, which have the potential to form SOA, have not yet been studied in detail.
Not only have known VOCs not been studied in detail, but unknown (or unmeasured) VOCs can
also produce SOA. This makes reconciling SOA from combustion sources extremely difficult.
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.310
6A.2 Air Quality Modeling Results
6A2.1 Annual PM2.5 Results
The air quality modeling indicates that for the majority of the country, annual PM2.5
design values (DV) will decrease due to these standards. The decreases in annual PM2.5 DV, less
than 0.05 |ig/m3, are likely due to the projected reductions in upstream primary PM2.5 emissions,
and reductions in both upstream and downstream NOx, SOx and VOCs. As described in
Chapter 5.5.2.3, the air quality modeling used inventories that do not reflect the new

-------
requirements for controlling PM2.5 emissions from APUs installed in new tractors and therefore
show increases in downstream PM2.5 emissions. Although in most areas this direct PM2.5
increase is outweighed by reductions in secondary PM2.5, the air quality modeling does predict
ambient PM2.5 increases in a few places. EPA is adopting Phase 1 and Phase 2 requirements to
control PM2.5 emissions from APUs installed in new tractors, therefore we do not expect to
actually see increases in PM2.5 DV from the Phase 2 program. In addition, assumptions about
the usage of diesel-powered APUs also differs between the air quality inventories and the final
rule inventories. The air quality inventories assumed more widespread usage of diesel-powered
APUs than was assumed for the final rule. The APU assumptions mean that the NOx reductions
assumed in the air quality inventories are larger than we expect to occur and reductions in
ambient PM2.5 due to secondary nitrate formation are over-estimated in the air quality modeling.
The magnitude of the reductions in PM2.5 DV from the final rule inventories is difficult
to estimate due to the differences in the air quality inventories, namely overestimation of nitrate
reductions and underestimation of direct PM2.5 reductions. However, EPA does expect
reductions in ambient concentrations of PM2.5 due to these final standards. Maps and summary
tables of the projected impacts of the air quality inventories on PM2.5 DV are presented below.
Figure 6A-2 presents the changes in annual PM2.5 design values in 2040.^
An annual PM2 5 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.

-------
n >= "0.01 to <=0.0
~~I >0.0 to <=0.01
J > 0.01 to <= 0.05
J >0.05 to <=0.10
329
Difference in Annual PM2.5 DV
2040ei_hdghgp2_ctl minus 2040ei_hdghgp2_ref
Figure 6A-2 Projected Change in 2040 Annual PM2.5 Design Values Using Air Quality Inventories
Table 6A-1 presents the average change in 2040 annual PM2.5 design values for: (1) all
counties with 2011 baseline design values, (2) counties with 2011 baseline design values that
exceeded the 2012 annual PM2.5 standard, (3) counties with 2011 baseline design values that did
not exceed the 2012 standard, but were within 10 percent of it, (4) counties with 2040 design
values that exceeded the 2012 annual PM2.5 standard, and (5) counties with 2040 design values
that did not exceed the standard, but were within 10 percent of it. Counties within 10 percent 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 PM2.5
NAAQS.

-------
Table 6A-1 Average Change in 2040 Annual PM2.5 Design Values using Air Quality Inventories

Number
of US
Counties
2040
Population
Change in
2040 design
value
(Hg/m3)
All
508

-0.01
All, population-weighted
234,351,941
-0.01
Counties whose 2011 base year is violating the 2012 annual PM2.5 standard
43

-0.02
Counties whose 2011 base year is violating the 2012 annual PM2.5 standard,
population-weighted
41,555,813
-0.02
Counties whose 2011 base year is within 10 percent of the 2012 annual PM2 5
standard
77

-0.01
Counties whose 2011 base year is within 10 percent of the 2012 annual PM2 5
standard, population-weighted
32,091,156
0.00
Counties whose 2040 control case is violating the 2012 annual PM2 5 standard
9

-0.02
Counties whose 2040 control case is violating the 2012 annual PM2 5 standard,
population-weighted
8,575,947
-0.02
Counties whose 2040 control case is within 10% of the 2012 annual PM2 5 standard
5

0.00
Counties whose 2040 control case is within 10% of the 2012 annual PM2 5
standard, population-weighted
6,951,178
-0.01
Notes:
a Averages are over counties with 2011 modeled design values
b Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. (2011). 2012 Complete
Economic and Demographic Data Source (CEDDS).
There are 9 counties, all in California, that are projected to have annual PM2.5 design
values above the NAAQS in 2040 without the Phase 2 standards or any other additional
standards in place. Table 6A-2 below presents the changes in design values for these counties.
Table 6A-2 Change in Annual PM2.5 Design Values (jig/im) using Air Quality Inventories for Counties
Projected to be Above the Annual PM2.5 NAAQS in 2040
County Name
Population in 2040a
Change in Annual
PM2 5 Design Value
(Hg/m3)
Madera, California
173,045
-0.03
Imperial, California
228,454
-0.02
Kings, California
173,643
-0.02
Fresno, California
1,350,320
-0.02
Kern, California
1,100,054
-0.02
Stanislaus, California
732,713
-0.02
Tulare, California
509,803
-0.02
Merced, California
323,734
-0.01
Riverside, California
3,984,181
-0.02
Notes:
a Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. (2011).
2012 Complete Economic and Demographic Data Source (CEDDS).

-------
6A2.2 24-hour PM2.5 Results
The air quality modeling indicates that for the majority of the country, 24-hour PM2.5
design values (DV) will decrease due to these standards. The decreases in 24-hour PM2.5 DV,
less than 0.6 |ig/m3, are likely due to the projected reductions in upstream primary PM2.5
emissions, and reductions in both upstream and downstream NOx, SOx and VOCs. As
described in Chapter 5.5.2.3, the air quality modeling used inventories that do not reflect the new
requirements for controlling PM2.5 emissions from APUs installed in new tractors and therefore
show increases in downstream PM2.5 emissions. Although in most areas this direct PM2.5
increase is outweighed by reductions in secondary PM2.5, the air quality modeling does predict
ambient PM2.5 increases in a few places. EPA is adopting Phase 1 and Phase 2 requirements to
control PM2.5 emissions from APUs installed in new tractors, therefore we do not expect to
actually see increases in PM2.5 DV from the Phase 2 program. In addition, assumptions about
the usage of diesel-powered APUs also differs between the air quality inventories and the final
rule inventories. The air quality inventories assumed more widespread usage of diesel-powered
APUs than was assumed for the final rule. The APU assumptions mean that the NOx reductions
assumed in the air quality inventories are larger than we expect to occur and reductions in
ambient PM2.5 due to secondary nitrate formation are over-estimated in the air quality modeling.
The magnitude of the reductions in PM2.5 DV from the final rule inventories is difficult
to estimate due to the differences in the air quality inventories, namely overestimation of nitrate
reductions and underestimation of direct PM2.5 reductions. However, EPA does expect
reductions in ambient concentrations of PM2.5 due to these final standards. Maps and summary
tables of the projected impacts of the air quality inventories on PM2.5 DV are presented below.
Figure 6A-3 presents the changes in 24-hour PM2.5 design values in 2040.00
00 An annual PM2 5 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.

-------
_| > -0.05 to < 0.05	267
>=0.05 to <0.15	11
^ >= 0.15 to < 0.25	5
Difference in Daily PM2.5 DV
2040ei_ldghgp2_ctl minus 2040ei_hdghgp2_ref
Figure 6A-3 Projected Change in 2040 Annual PMis Design Values Using Air Quality Inventories
Table 6A-3 presents the average change in 2040 24-hour PM2.5 design values for: (1) all
counties with 2011 baseline design values, (2) counties with 2011 baseline design values that
exceeded the 2012 24-hour PM2.5 standard, (3) counties with 2011 baseline design values that
did not exceed the 2012 standard, but were within 10 percent of it, (4) counties with 2040 design
values that exceeded the 2012 24-hour PM2.5 standard, and (5) counties with 2040 design values
that did not exceed the standard, but were within 10 percent of it. Counties within 10 percent 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.

-------
Table 6A-3 Average Change in 2040 24-hour PM2.5 Design Values using Air Quality Inventories

Number
of US
Counties
2040
Population
Change in
2040 design
value
(Hg/m3)
All
513

-0.06
All, population-weighted
247,723,536
-0.05
Counties whose 2011 base year is violating the 2012 24-hour PM25 standard
23

-0.15
Counties whose 2011 base year is violating the 2012 24-hour PM25
standard, population-weighted
14,226,741
-0.18
Counties whose 2011 base year is within 10 percent of the 2012 24-hour
PM2.5 standard
13

-0.08
Counties whose 2011 base year is within 10 percent of the 2012 24-hour
PM2.5 standard, population-weighted
6,249,037
-0.10
Counties whose 2040 control case is violating the 2012 24-hour PM25
standard
11

-0.05
Counties whose 2040 control case is violating the 2012 24-hour PM25
standard, population-weighted
4,475,471
-0.09
Counties whose 2040 control case is within 10% of the 2012 24-hour PM2 5
standard
11

-0.14
Counties whose 2040 control case is within 10% of the 2012 24-hour PM2 5
standard, population-weighted
6,241,043
-0.23
Notes:
a Averages are over counties with 2011 modeled design values
b Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. (2011). 2012 Complete Economic and
Demographic Data Source (CEDDS).
There are 11 counties, mainly in California, that are projected to have 24-hour PM2.5
design values above the NAAQS in 2040 without the Phase 2 standards or any other additional
standards in place.

-------
Table 6A-4 below presents the changes in design values for these counties.
Table 6A-4 Change in 24-hour PM2.5 Design Values (jig/m3) using Air Quality Inventories for Counties
Projected to be Above the 24-hour PM2.5 NAAQS in 2040
County Name
Population
in 2040a
Change in 24-
hour PM2 5
Design Value
(Hg/m3)
Ravalli, Montana
53,253
0.0
Fresno, California
1,350,320
-0.1
Kings, California
173,643
-0.1
Kern, California
1,100,054
-0.1
Madera, California
173,045
0.0
Stanislaus, California
732,713
-0.1
Lake, Oregon
9,349
0.0
Tulare, California
509,803
-0.1
Shoshone, Idaho
10,981
0.0
Silver Bow, Montana
38,576
-0.1
Merced, California
323,734
0.0
Notes:
a Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. (2011).
2012 Complete Economic and Demographic Data Source (CEDDS).
6A2.3 Ozone Results
Air quality modeling results indicate that 8-hour ozone DV will be reduced across the
country. The decreases in 8-hour ozone DV, max reduction of 1.7 ppb, are likely due to the
projected reductions in both upstream and downstream NOx and VOC emissions. As described
in Chapter 5.5.2.3, assumptions about the usage of diesel-powered APUs differs between the air
quality inventories and the final rule inventories. The air quality inventories assumed more
widespread usage of diesel-powered APUs than was assumed for the final rule. The APU
assumptions mean that the NOx reductions assumed in the air quality inventories are larger than
we expect to occur and reductions in 8-hour ozone are over-estimated in the air quality modeling.
The magnitude of the reductions in 8-hour ozone DV from the final rule inventories is
difficult to estimate due to the complex, non-linear chemistry governing ozone formation.
However, EPA does expect reductions in ambient ozone concentrations due to these final
standards. Maps and summary tables of the projected impacts of the air quality inventories on 8-
hour ozone DV are presented below. Figure 6A-4 presents the changes in 8-hour ozone design
values in 2040.pp
pp An 8-hour ozone design value is the concentration that determines whether a monitoring site meets the NAAQS
for ozone. The Ml details involved in calculating an 8-hour ozone design value are given in appendix I of 40 CFR
part 50.

-------
>= -0.5 to < -0.25
>= -0.75 lo < -0.5
271
267
] >= -0.25 to < -0.1
55
Difference in Ozone DV
35
2Q40ei_ldghgp2_ctl minus 2040ei_hdghgp2_ref
Figure 6A-4 Projected Change in 2040 8-hour Ozone Design Values Using Air Quality Inventories
Table 6 A-5 presents the average change in 2040 8-hour ozone design values for: (1) all
counties with 2011 baseline design values, (2) counties with 2011 baseline design values that
exceeded the 2015 8-hour ozone standard, (3) counties with 2011 baseline design values that did
not exceed the 2015 standard, but were within 10 percent of it, (4) counties with 2040 design
values that exceeded the 2015 8-hour ozone standard, and (5) counties with 2040 design values
that did not exceed the standard, but were within 10 percent of it. Counties within 10 percent 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 8-hour ozone
NAAQS.

-------
Table 6A-5 Average Change in 2040 8-hour Ozone Design Values using Air Quality Inventories

Number
of US
Counties
2040
Population
Change in
2040 design
value (ppb)
All
706
286,828,135
-0.50
All, population-weighted

-0.46
Counties whose 2011 base year is violating the 2015 8-hour ozone standard
372

-0.54
Counties whose 2011 base year is violating the 2040 8-hour ozone standard,
population-weighted
207,493,027
-0.47
Counties whose 2011 base year is within 10 percent of the 2015 8-hour
ozone standard
239

-0.50
Counties whose 2011 base year is within 10 percent of the 2015 8-hour
ozone standard, population-weighted
56,116,399
-0.48
Counties whose 2040 control case is violating the 2015 8-hour ozone
standard
14

-0.19
Counties whose 2040 control case is violating the 2015 8-hour ozone
standard, population-weighted
29,944,552
-0.14
Counties whose 2040 control case is within 10% of the 2015 8-hour ozone
standard
37

-0.37
Counties whose 2040 control case is within 10% of the 2015 8-hour ozone
standard, population-weighted
32,176,523
-0.45
Notes:
a Averages are over counties with 2011 modeled design values
b Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. (2011). 2012 Complete
Economic and Demographic Data Source (CEDDS).
There are 16 counties that are projected to have 8-hour Ozone design values above the
NAAQS in 2040 without the Phase 2 standards or any other additional standards in place. Table
6A-6 below presents the changes in design values for these counties.

-------
Table 6A-6 Change in 8-hour Ozone Design Values (jig/m3) using Air Quality Inventories for Counties
Projected to be Above the 8-hour Ozone NAAQS in 2040
County Name
Population in
2040a
Change in 8-
hour Ozone
Design Value
(ppb)
San Bernardino, California
3,273,894
-0.11
Los Angeles, California
10,765,068
-0.08
Riverside, California
3,984,181
-0.10
Fairfield, Connecticut
1,019,651
-0.46
Queens, New York
2,462,190
-0.20
Fresno, California
1,350,320
-0.08
Westchester, New York
1,026,461
-0.23
Tulare, California
509,803
-0.07
Kern, California
1,100,054
-0.13
Richmond, New York
672,799
-0.38
Bronx, New York
1,643,295
-0.17
Suffolk, New York
1,889,102
-0.33
Imperial, California
228,455
-0.13
Sublette, Wyoming
19,279
-0.17
Larimer, Colorado
597,906
-0.39
Rio Blanco, Colorado
7,422
-0.20
Note:
" Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. (2011).
2012 Complete Economic and Demographic Data Source (CEDDS).
6A2.4 NO2 Results
Air quality modeling results indicate that annual average NO2 concentrations will be
reduced across the country, see Figure 6A-5. However, the magnitude of the reductions that will
actually result from the final standards is difficult to estimate because the air quality modeling
inventories included larger NOx emission reductions than we now expect to occur. As
described in Chapter 5.5.2.3, the air quality inventories and the final rule inventories make
different assumptions about the usage of diesel-powered APUs. The air quality inventories
assumed more widespread usage of diesel-powered APUs than was assumed for the final rule,
and as a result the reductions in ambient NO2 concentrations are overestimated in the air quality
modeling.

-------
%

*•* 7
,'
-------
Legend
Figure 6A-6 Annual Changes in Acetaldehyde Ambient Concentrations between the Reference Case and the Control Case in 2040
Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in fig/m3 (right)
Legend
Legend
Figure 6A-7 Winter Changes in Acetaldehyde Ambient Concentrations between the Reference Case and the Control Case in 2040
Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in fig/m3 (right)

-------
Legend
Legend
Figure 6A-8 Summer Changes in Acetaldehyde Ambient Concentrations between the Reference Case and the Control Case in
2040 Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in jug/m1 (right)
Legend
Figure 6A-9 Annual Changes in Acrolein Ambient Concentrations between the Reference Case and the Control Case in 2040 Using
Air Quality Inventories: Percent Changes (left) and Absolute Changes in jug/m 5 (right)

-------
Legend
Percent Change tot Acrolein - Winter Season
204M_hdghgp2_ctl minus 2040ol_hdghgp2_ref
Absolute Difference for Acrolein - Winter Season
2M0ol_l>dghgp2_ctl minus 2040al_hdghgp2_ref

Figure 6A-10 Winter Changes in Acrolein Ambient Concentrations between the Reference Case and the Control Case in 2040
Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in fig/m3 (right)
Legend
Figure 6A-11 Summer Changes in Acrolein Ambient Concentrations between the Reference Case and the Control Case in 2040
Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in fig/m3 (right)

-------
Legend
Figure 6A-12 Annual Changes in Benzene Ambient Concentrations between the Reference Case and the Control
Case in 2040 Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in jig/m3 (right)
Legend
Figure 6A-13 Winter Changes in Benzene Ambient Concentrations between the Reference Case and the Control Case in 2040
Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in ug/rn1 (right)

-------
Legend
Legend
Figure 6A-14 Summer Changes in Benzene Ambient Concentrations between the Reference Case and the Control Case in 2040
Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in fig/m3 (right)
Legend
1 to< 1%
1 to < 2.5%
2.5 to < 5%
5 to < 10%
10 to < 50%
50 to <100%
100%
= 0.002 to <0.003
nparable hetwee
Percent Change for 1,3 Butadiene ¦¦ Annual
204dghgp2_ref
Absolute Difference for 1,3 Butadiene ¦¦ Annual
204dghgp2_ref
Figure 6A-15 Changes in 1,3-Butadiene Ambient Concentrations between the Reference Case and the Control Case in 2040 Using
Air Quality Inventories: Percent Changes (left) and Absolute Changes in jug/m1 (right)

-------
V	|^| «=-0.005 Ua;m3	1 \_
\	H > O00S;c<- 0004 J \ \
I j	H » -0.001 to <= -0.003 \ J
Percent Change for 1,3 Butadiene - Winter Season
204M_hdghgp2_ctl minus 2040ol_hdghgp2_ref
Absolute Difference for 1,3 Butadiene - Winter Season
2MtM_hdghgp2_ctl minus 2040al_hdghgp2_ref

Figure 6A-16 Winter Changes in 1,3-Butadiene Ambient Concentrations between the Reference Case and the Control Case in
2040 Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in jug/m1 (right)
Legend
Legend
Percent Change for 1,3 Butadiene - Summer Season
204M_hdghgp2_ctl minus 2040el_lidghgp2_ref
Absolute Difference for 1,3 Butadiene ¦¦ Summer Season
2CU0ol_hdghgp2_ctl minus 2040al_hdghgp2_raf

Figure 6A-17 Summer Changes in 1,3-Butadiene Ambient Concentrations between the Reference Case and the Control Case in
2040 Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in jug/m1 (right)

-------
Legend
Figure 6A-18 Changes in Formaldehyde Ambient Concentrations between the Reference Case and the Control
Case in 2040 Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in jug/m1 (right)
Legend
Legend
Figure 6A-19 Winter Changes in Formaldehyde Ambient Concentrations between the Reference Case and the Control Case in
2040 Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in jug/m1 (right)

-------
m
Legend
Legend
Figure 6A-20 Summer Changes in Formaldehyde Ambient Concentrations between the Reference Case and the Control Case in
2040 Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in jug/nr1 (right)
Legend
¦	100%
-100 tn <= -50%
-50 ta <= -10%
-1010 <= -5%
-5 to <= -2.5%
-2.5 to ¦== -1%
1 to<1%
¦	1 to < 2.5%
= 2.5 to < 5%
= 5 to < 10%
= 10 to <50%
= 50 to < 100%
= 100%
Percent Change for Naphthalene —Annual
!040el_hdghgp2_cll minus 2040el_hdghgp2_rgf
Absolute Difference for Naphthalene "Annual
2040el_hdghgp2_ctl minus 204M_hdghgp2_r6f
Figure 6A-21 Changes in Naphthalene Ambient Concentrations between the Reference Case and the Control Case in 2040 Using
Air Quality Inventories: Percent Changes (left) and Absolute Changes in jag/m3 (right)

-------
Legend
Figure 6A-22 Winter Changes in Naphthalene Ambient Concentrations between the Reference Case and the Control Case in 2040
Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in fig/m3 (right)
Legend
Legend
Percent Change for Naphthalene -- Summer Season
2040ol_hdghgp2_ctl minus 2040el_lidghgp2_ref
te Difference for Naphthalene » Summer Season
2M0el_hdghgp2_ctl minus 2M0al_hdghgp2_raf

Figure 6A-23 Summer Changes in Naphthalene Ambient Concentrations between the Reference Case and the Control Case in
2040 Using Air Quality Inventories: Percent Changes (left) and Absolute Changes in jug/m1 (right)

-------
Table 6A-7 Percent of Total Population Experiencing Changes in Annual Ambient Concentrations of Toxic
Pollutants in 2040 as a Result of the Standards
Percent
Change
Acetaldehyde
Acrolein
Benzene
1,3-
Butadiene
Ethanol
Formaldehyde
Naphthalene
<-50

0%




0%
> -50 to < -25

1%




4%
> -25 to < -10

8%



1%
20%
> -10 to < -5
0%
15%
0%


2%
24%
> -5 to < -2.5
0%
25%
1%


5%
21%
> -2.5 to < -1
3%
28%
5%
1%

18%
15%
> -1 to < 1
97%
23%
94%
99%
100%
74%
15%
> 1 to < 2.5



0%



>2.5 to <5







> 5 to < 10







> 10 to < 25







> 25 to < 50







>50







6A2.6 Visibility Results
Table 6A-8 Visibility Levels (in Deciviews) for Mandatory Class I Federal Areas on the 20 Percent Worst
Days Using Air Quality Inventories
Class 1 Area
(20% worst days)
State
2011
Baseline
Visibility
2040
Reference
2040
HDGHGP2
Control
Natural
Background
Sipsey Wilderness
Alabama
22.93
18.16
18.07
10.99
Mazatzal Wilderness
Arizona
12.03
11.40
11.38
6.68
Pine Mountain Wilderness
Arizona
12.03
11.40
11.38
6.68
Superstition Wilderness
Arizona
12.72
11.82
11.80
6.54
Chiricahua NM
Arizona
12.08
11.54
11.53
7.20
Chiricahua Wilderness
Arizona
12.08
11.54
11.53
7.20
Galiuro Wilderness
Arizona
12.08
11.54
11.53
7.20
Grand Canyon NP
Arizona
10.92
10.53
10.52
7.04
Petrified Forest NP
Arizona
11.92
11.64
11.63
6.49
Sycamore Canyon Wilderness
Arizona
14.62
14.00
14.01
6.65
Caney Creek Wilderness
Arkansas
22.23
19.01
18.96
11.58
Upper Buffalo Wilderness
Arkansas
22.12
19.00
18.95
11.57
Joshua Tree NM
California
15.07
13.49
13.47
7.19
Kings Canyon NP
California
20.82
17.93
17.91
7.70
San Rafael Wilderness
California
16.46
14.51
14.49
7.57
San Gorgonio Wilderness
California
16.85
14.11
14.09
7.30

-------
San Jacinto Wilderness
California
16.85
14.11
14.09
7.30
Sequoia NP
California
20.82
17.93
17.91
7.70
Agua Tibia Wilderness
California
18.44
15.66
15.65
7.64
Ansel Adams Wilderness (Minarets)
California
14.27
13.01
13.00
7.12
Desolation Wilderness
California
11.82
11.02
11.01
6.05
Dome Land Wilderness
California
17.23
15.93
15.92
7.46
Emigrant Wilderness
California
14.75
14.16
14.15
7.64
Hoover Wilderness
California
10.78
10.31
10.30
7.71
John Muir Wilderness
California
14.27
13.01
13.00
7.12
Kaiser Wilderness
California
14.27
13.01
13.00
7.12
Marble Mountain Wilderness
California
14.10
13.34
13.33
7.90
Mokelumne Wilderness
California
11.82
11.02
11.01
6.05
Pinnacles NM
California
16.15
14.42
14.41
7.99
Ventana Wilderness
California
16.15
14.42
14.41
7.99
Yolla Bolly Middle Eel Wilderness
California
14.10
13.34
13.33
7.90
Yosemite NP
California
14.75
14.16
14.15
7.64
Caribou Wilderness
California
13.49
12.83
12.83
7.31
Lava Beds NM
California
13.38
12.93
12.93
7.85
Lassen Volcanic NP
California
13.49
12.83
12.83
7.31
Point Reyes NS
California
20.98
19.93
19.93
15.77
Redwood NP
California
17.38
16.82
16.82
13.91
South Warner Wilderness
California
13.38
12.93
12.93
7.85
Thousand Lakes Wilderness
California
13.49
12.83
12.83
7.31
Rocky Mountain NP
Colorado
11.84
10.93
10.91
7.15
Black Canyon of the Gunnison NM
Colorado
9.88
9.71
9.70
6.21
La Garita Wilderness
Colorado
9.88
9.71
9.70
6.21
Weminuche Wilderness
Colorado
9.88
9.71
9.70
6.21
Eagles Nest Wilderness
Colorado
8.48
8.04
8.03
6.06
Flat Tops Wilderness
Colorado
8.48
8.04
8.03
6.06
Great Sand Dunes NM
Colorado
11.57
11.50
11.49
6.66
Maroon Bells-Snowmass Wilderness
Colorado
8.48
8.04
8.03
6.06
Mount Zirkel Wilderness
Colorado
9.11
8.70
8.69
6.08
Rawah Wilderness
Colorado
9.11
8.70
8.69
6.08
West Elk Wilderness
Colorado
8.48
8.04
8.03
6.06
Mesa Verde NP
Colorado
11.22
11.37
11.37
6.81
Chassahowitzka
Florida
21.34
18.21
18.17
11.03
St. Marks
Florida
22.23
18.74
18.70
11.67
Everglades NP
Florida
18.15
17.65
17.62
12.15
Cohutta Wilderness
Georgia
22.71
17.47
17.43
10.78
Okefenokee
Georgia
22.68
18.82
18.78
11.44
Wolf Island
Georgia
22.68
18.82
18.78
11.44
Craters of the Moon NM
Idaho
14.05
12.93
12.80
7.53
Sawtooth Wilderness
Idaho
15.64
15.44
15.44
6.42

-------
Selway-Bitterroot Wilderness
Idaho
14.89
14.77
14.77
7.43
Mammoth Cave NP
Kentucky
25.09
19.83
19.75
11.08
Acadia NP
Maine
17.93
15.81
15.80
12.43
Moosehorn
Maine
16.83
15.27
15.26
12.01
Roosevelt Campobello International Park
Maine
16.83
15.27
15.26
12.01
Seney
Michigan
20.56
17.15
17.08
12.65
Isle Royale NP
Michigan
18.92
16.06
16.01
12.37
Boundary Waters Canoe Area
Minnesota
18.82
16.66
16.60
11.61
Hercules-Glades Wilderness
Missouri
22.89
19.57
19.51
11.30
Mingo
Missouri
24.31
20.91
20.86
11.62
Medicine Lake
Montana
17.98
17.07
17.06
7.89
Bob Marshall Wilderness
Montana
14.43
14.33
14.32
7.73
Cabinet Mountains Wilderness
Montana
12.73
12.24
12.23
7.52
Glacier NP
Montana
16.03
15.82
15.81
9.18
Mission Mountains Wilderness
Montana
14.43
14.33
14.32
7.73
Red Rock Lakes
Montana
11.98
11.73
11.72
6.44
Scapegoat Wilderness
Montana
14.43
14.33
14.32
7.73
UL Bend
Montana
14.11
13.77
13.76
8.16
Anaconda-Pintler Wilderness
Montana
14.89
14.77
14.77
7.43
Jarbidge Wilderness
Nevada
11.97
11.90
11.90
7.87
Great Gulf Wilderness
New Hampshire
16.66
13.61
13.60
11.99
Presidential Range-Dry River Wilderness
New Hampshire
16.66
13.61
13.60
11.99
Brigantine
New Jersey
23.75
19.64
19.61
12.24
Bosque del Apache
New Mexico
14.02
14.37
14.34
6.73
Salt Creek
New Mexico
17.42
18.32
18.30
6.81
Bandelier NM
New Mexico
11.92
12.22
12.21
6.26
Carlsbad Caverns NP
New Mexico
15.32
15.09
15.08
6.65
Pecos Wilderness
New Mexico
9.93
9.84
9.83
6.08
San Pedro Parks Wilderness
New Mexico
10.02
10.02
10.01
5.72
Wheeler Peak Wilderness
New Mexico
9.93
9.84
9.83
6.08
White Mountain Wilderness
New Mexico
14.19
14.56
14.56
6.80
Linville Gorge Wilderness
North Carolina
21.60
15.94
15.91
11.22
Swanquarter
North Carolina
21.77
16.75
16.73
11.55
Theodore Roosevelt NP
North Dakota
16.96
15.96
15.95
7.80
Wichita Mountains
Oklahoma
21.24
18.83
18.76
7.53
Hells Canyon Wilderness
Oregon
16.58
15.10
14.94
8.32
Eagle Cap Wilderness
Oregon
14.87
14.20
14.17
8.92
Strawberry Mountain Wilderness
Oregon
14.87
14.20
14.17
8.92
Kalmiopsis Wilderness
Oregon
15.01
14.52
14.51
9.44
Mount Hood Wilderness
Oregon
13.35
12.72
12.71
8.43
Mount Jefferson Wilderness
Oregon
15.77
15.52
15.51
8.79
Mount Washington Wilderness
Oregon
15.77
15.52
15.51
8.79
Three Sisters Wilderness
Oregon
15.77
15.52
15.51
8.79

-------
Crater Lake NP
Oregon
11.64
11.33
11.33
7.62
Diamond Peak Wilderness
Oregon
11.64
11.33
11.33
7.62
Gearhart Mountain Wilderness
Oregon
11.64
11.33
11.33
7.62
Mountain Lakes Wilderness
Oregon
11.64
11.33
11.33
7.62
Cape Romain
South Carolina
23.17
19.02
18.99
12.12
Wind Cave NP
South Dakota
14.04
12.85
12.82
7.71
Badlands NP
South Dakota
15.67
14.32
14.30
8.06
Great Smoky Mountains NP
Tennessee
22.50
16.99
16.95
11.24
Joyce-Kilmer-Slickrock Wilderness
Tennessee
22.50
16.99
16.95
11.24
Guadalupe Mountains NP
Texas
15.32
15.09
15.08
6.65
Big Bend NP
Texas
16.30
16.54
16.54
7.16
Arches NP
Utah
10.83
10.53
10.50
6.43
Canyonlands NP
Utah
10.83
10.53
10.50
6.43
Capitol Reef NP
Utah
10.18
9.69
9.66
6.03
Bryce Canyon NP
Utah
10.61
10.21
10.19
6.80
Lye Brook Wilderness
Vermont
19.26
14.94
14.92
11.73
James River Face Wilderness
Virginia
22.55
17.28
17.24
11.13
Shenandoah NP
Virginia
21.82
15.20
15.16
11.35
Alpine Lake Wilderness
Washington
16.14
14.86
14.80
8.43
Mount Rainier NP
Washington
15.50
14.43
14.41
8.54
Olympic NP
Washington
14.10
13.50
13.48
8.44
Pasayten Wilderness
Washington
12.44
11.83
11.81
8.25
Glacier Peak Wilderness
Washington
13.51
12.82
12.81
8.39
Goat Rocks Wilderness
Washington
12.37
11.77
11.76
8.35
North Cascades NP
Washington
13.51
12.82
12.81
8.01
Mount Adams Wilderness
Washington
12.37
11.77
11.76
8.35
Dolly Sods Wilderness
West Virginia
22.40
16.06
16.03
10.39
Otter Creek Wilderness
West Virginia
22.40
16.06
16.03
10.39
Bridger Wilderness
Wyoming
10.25
9.91
9.90
6.45
Fitzpatrick Wilderness
Wyoming
10.25
9.91
9.90
6.45
Grand Teton NP
Wyoming
11.98
11.73
11.72
6.44
Teton Wilderness
Wyoming
11.98
11.73
11.72
6.44
Yellowstone NP
Wyoming
11.98
11.73
11.72
6.44
6A2.7 Deposition Results
Figure 6A-24 presents changes in projected nitrogen deposition in 2040 due to the
standards and Figure 6A-25 presents changes in projected sulfur deposition due to the standards.

-------
Legend
Legend
Figure 6A-24 Changes in Nitrogen Deposition between the Reference Case and the Control Case in 2040 using Air Quality
Inventories: Percent Changes (left) and Absolute Changes in kg/ha (right)
Legend
¦ Deposition
dghgp2_ref
Figure 6A-25 Changes in Sulfur Deposition between the Reference Case and the Control Case in 2040 using Air Quality
Inventories: Percent Changes (left) and Absolute Changes in kg/ha (right)

-------
References
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.
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.
3	78 FR 3086 (January 15, 2013) pages 3103-3104.
4	77 FR 38890 (June 29, 2012) pages 38906-38911.
5	78 FR 3086 (January 15, 2013) pages 3103-3104.
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, Chapter 6 (Section 6.5) and Chapter 7 (Section 7.6).
7	78 FR 3103 (January 15, 2013).
8	78 FR 3103 (January 15, 2013).
9	78 FR 3103-3104 (January 15,2013).
10	78 FR 3104 (January 15, 2013).
11	U.S. EPA. (2009). Integrated Science Assessment for Particulate Matter (Final Report).
Protection Agency, Washington, DC, EPA/600/R-08/139F. pg 2-13.
12	U.S. EPA. (2009). Integrated Science Assessment for Particulate Matter (Final Report).
Protection Agency, Washington, DC, EPA/600/R-08/139F. pg 2-26.
13	U.S. EPA. (2009). Integrated Science Assessment for Particulate Matter (Final Report).
Protection Agency, Washington, DC, EPA/600/R-08/139F. Section 2.3.4 and Table 2-6.
14	78 FR 3167-8 (January 15, 2013).
15	77 FR 38947-51 (June 29, 2012).
16	U.S. EPA. (2009). Integrated Science Assessment for Particulate Matter (Final Report).
Protection Agency, Washington, DC, EPA/600/R-08/139F. Section 2.3.5 and Table 2-6.
17	78 FR 3121 (January 15,2013).
18	U.S. EPA. (2009). Integrated Science Assessment for Particulate Matter (Final Report).
Protection Agency, Washington, DC, EPA/600/R-08/139F. Chapter 8 and Chapter 2.
19	77 FR 38890, (June 29, 2012) page 38910.
20	78 FR 3086 (January 15, 2013) page 3104.
21	U.S. EPA. (2011). Policy Assessment for the Review of the PM NAAQS. U.S. Environmental Protection Agency,
Washington, DC, EPA/452/R-11-003. Section2.2.1.
22	U.S. EPA. (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-08/139F. Chapter 8 and Chapter 2 (Section 2.4.1).
23	U.S. EPA. Integrated Science Assessment of Ozone and Related Photochemical Oxidants (Final Report). U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R-10/076F, 2013. The ISA is available at
http://clpub.epa. gov/ncea/isa/recordisplav.cfm?deid=247492#Download.
24	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/recordisplav.cfm?deid=198843.
25	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/recordisplav.cfm?deid=218686. See Section 2.1.
26	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://cljpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=218686.
27	U.S. EPA. (1999). Guidelines for Carcinogen Risk Assessment. Review Draft. NCEA-F-0644, July. Washington,
DC: U.S. EPA. Retrieved on March 19, 2009 from http://cljpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=54932.
28	U.S. EPA (2002). Health Assessment Document for Diesel Engine Exhaust. EPA/600/8-90/057F Office of
research and Development, Washington DC. Retrieved on March 17, 2009 from
http://cljpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=29060. pp. 1-1 1-2.
U.S. Environmental
U.S. Environmental
U.S. Environmental
U.S. Environmental
U.S. Environmental

-------
29	Garshick, Eric, Francine Laden, Jaime E. Hart, Mary E. Davis, Ellen A. Eisen, and Thomas J. Smith. 2012. Lung
cancer and elemental carbon exposure in trucking industry workers. Environmental Health Perspectives 120(9):
1301-1306.
30	Silverman, D. T., Samanic, C. M., Lubin, J. H., Blair, A. E., Stewart, P. A., Vermeulen, R., & Attfield, M. D.
(2012). The diesel exhaust in miners study: a nested case-control study of lung cancer and diesel exhaust. Journal of
the National Cancer Institute.
31	Olsson, Ann C., et al. "Exposure to diesel motor exhaust and lung cancer risk in a pooled analysis from case-
control studies in Europe and Canada." American journal of respiratory and critical care medicine 183.7 (2011):
941-948.
32IARC [International Agency for Research on Cancer], (2013). Diesel and gasoline engine exhausts and some
nitroarenes. IARC Monographs Volume 105. [Online at
http://monographs.iarc.fr/ENG/Monographs/voll05/index.php]
33	U.S. EPA. (2015) Summary of Results for the 2011 National-Scale Assessment.
http ://www3. epa. gov/sites/production/files/2015-12/documents/2011 -nata-summary-results.pdf.
34	U.S. EPA (2015) 2011 National Air Toxics Assessment, http://www3.epa.gov/national-air-toxics-
assessment/2011 -national-air-toxics-assessment.
35	U.S. EPA. (2000). Integrated Risk Information System File for Benzene. This material is available electronically
at: http://www3.epa.gov/iris/subst/0276.htm.
36	International Agency for Research on Cancer. (1982). 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 1982.
37	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.
38	U.S. EPA. (2000). Integrated Risk Information System File for Benzene. This material is available electronically
at: http://www3.epa.gov/iris/subst/0276.htm.
39	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.
40	NTP. (2014). 13th Report on Carcinogens. Research Triangle Park, NC: U.S. Department of Health and Human
Services, Public Health Service, National Toxicology Program.
41Aksoy, M. (1989). Hematotoxicity and carcinogenicity of benzene. Environ. Health Perspect. 82: 193-197. EPA-
HQ-OAR-2011-0135.
42 Goldstein, B.D. (1988). Benzene toxicity. Occupational medicine. State of the Art Reviews. 3:541-554.
43Rothman, 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, andR.B. Hayes. (1996). Hematotoxicity among Chinese
workers heavily exposed to benzene. Am. J. Ind. Med. 29: 236-246.
44 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://www3.epa.gov/iris/subst/0276.htm.
45Qu, 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.
46	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.
47	Lan, Qing, Zhang, L., Li, G., Vermeulen, R., et al. (2004). Hematotoxically in Workers Exposed to Low Levels of
Benzene. Science 306: 1774-1776.
48	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.
49	U.S. Agency for Toxic Substances and Disease Registry (ATSDR). (2007). Toxicological profile for benzene.
Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service.
http://www.atsdr.cdc. gov/ToxProfiles/tp3 .pdf.
50U.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://www3.epa.gov/iris/supdocs/buta-sup.pdf.

-------
51	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://www3.epa. gov/iris/subst/0139.htm.
52	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.
53	International Agency for Research on Cancer (IARC). (2008). Monographs on the evaluation of carcinogenic risk
of chemicals to humans, 1,3-Butadiene, Ethylene Oxide and Vinyl Halides (Vinyl Fluoride, Vinyl Chloride and
Vinyl Bromide) Volume 97, World Health Organization, Lyon, France.
54	NTP. (2014). 13th Report on Carcinogens. Research Triangle Park, NC: U.S. Department of Health and Human
Services, Public Health Service, National Toxicology Program.
55	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://www3.epa.gov/iris/subst/0139.htm.
56Bevan, 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.
57	EPA. Integrated Risk Information System. Formaldehyde (CASRN 50-00-0)
http://www3.epa.gov/iris/subst/0419/htm.
58	NTP. (2014). 13th Report on Carcinogens. Research Triangle Park, NC: U.S. Department of Health and Human
Services, Public Health Service, National Toxicology Program.
59	IARC Monographs on the Evaluation of Carcinogenic Risks to Humans Volume 88 (2006): Formaldehyde, 2-
Butoxyethanol and l-tert-Butoxypropan-2-ol.
60	IARC Monographs on the Evaluation of Carcinogenic Risks to Humans Volume 100F (2012): Formaldehyde.
61	Hauptmann, M.; Lubin, J. H.; Stewart, P. A.; Hayes, R. B.; Blair, A. 2003. Mortality from lymphohematopoetic
malignancies among workers in formaldehyde industries. Journal of the National Cancer Institute 95: 1615-1623.
62Hauptmann, 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.
63Beane 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.
64	Pinkerton, L. E. 2004. Mortality among a cohort of garment workers exposed to formaldehyde: an update.
Occup. Environ. Med. 61: 193-200.
65	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.
66	Hauptmann, M,; Stewart P. A.; Lubin J. H.; Beane Freeman, L. E.; Hornung, R. W.; Herrick, R. F.; Hoover, R. N.;
Fraumeni, J. F.; Hayes, R. B. 2009. Mortality from lymphohematopoietic malignancies and brain cancer among
embalmers exposed to formaldehyde. Journal of the National Cancer Institute 101:1696-1708.
67	ATSDR. 1999. Toxicological Profile for Formaldehyde, U.S. Department of Health and Human Services (HHS),
July 1999.
68	ATSDR. 2010. Addendum to the Toxicological Profile for Formaldehyde. U.S. Department of Health and Human
Services (HHS), October 2010.
69	IPCS. 2002. Concise International Chemical Assessment Document 40. Formaldehyde. World Health
Organization.
70	EPA (U.S. Environmental Protection Agency). 2010. Toxicological Review of Formaldehyde (CAS No. 50-00-0)
- Inhalation Assessment: In Support of Summary Information on the Integrated Risk Information System (IRIS).
External Review Draft. EPA/635/R-10/002A. U.S. Environmental Protection Agency, Washington DC [online].
Available: http://clpub.epa.gov/ncea/irs_drats/recordisplay.cfm?deid=223614.
71	NRC (National Research Council). 2011. Review of the Environmental Protection Agency's Draft IRIS
Assessment of Formaldehyde. Washington DC: National Academies Press.
http ://books. nap. edu/openbook.php?record_id= 13142.
72U.S. EPA (1991). Integrated Risk Information System File of Acetaldehyde. Research and Development,
National Center for Environmental Assessment, Washington, DC. This material is available electronically at
http ://www3 .epa. gov/iris/sub st/0290. htm.
73 U.S. EPA (1991). Integrated Risk Information System File of Acetaldehyde. This material is available
electronically at http://www3.epa.gov/iris/subst/0290.htm.

-------
74	NTP. (2014). 13th Report on Carcinogens. Research Triangle Park, NC: U.S. Department of Health and Human
Services, Public Health Service, National Toxicology Program.
75	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.
76	U.S. EPA (1991). Integrated Risk Information System File of Acetaldehyde. This material is available
electronically at http://www3.epa.gov/iris/subst/0290.htm.
77	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://www3.epa. gov/iris/subst/0364.htm.
78	Appleman, L.M., R. A. Woutersen, and V.J. Feron. (1982). Inhalation toxicity of acetaldehyde in rats. I. Acute and
subacute studies. Toxicology. 23: 293-297.
79	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.
80	U.S. EPA. (2003). Integrated Risk Information System File of Acrolein. Research and Development, National
Center for Environmental Assessment, Washington, DC. This material is available at
http://www3.epa. gov/iris/subst/0364.htm.
81	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.
82	U.S. EPA. (2003). 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: //www3. epa. go v/iris/sub st/0364. htm.
83	U.S. EPA. (2003). 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: //www3 .epa. go v/iris/sub st/0364. htm.
84	U.S. EPA. (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. p.
10. Available online at: http://www3.epa.gov/ncea/iris/toxreviews/0364tr.pdf.
85	U.S. EPA. (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://www3.epa.gov/ncea/iris/toxreviews/0364tr.pdf.
86	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.
87	U.S. EPA. (2009). Graphical Arrays of Chemical-Specific Health Effect Reference Values for Inhalation
Exposures (Final Report). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-09/061, 2009.
http://clpub.epa. gov/ncea/cfm/recordisplay.cfm?deid=211003.
88	Agency for Toxic Substances and Disease Registry (ATSDR). (1995). Toxicological profile for Polycyclic
Aromatic Hydrocarbons (PAHs). Atlanta, GA: U.S. Department of Health and Human Services, Public Health
Service. Available electronically at http://www.atsdr.cdc.gov/ToxProfiles/TP.asp?id=122&tid=25.
89	U.S. EPA (2002). Health Assessment Document for Diesel Engine Exhaust. EPA/600/8-90/057F Office of
Research and Development, Washington DC. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=29060.
90	International Agency for Research on Cancer (IARC). (2012). Monographs on the Evaluation of the Carcinogenic
Risk of Chemicals for Humans, Chemical Agents and Related Occupations. Vol. 100F. Lyon, France.
91	U.S. EPA (1997). Integrated Risk Information System File of indeno (l,2,3-cd)pyrene. Research and
Development, National Center for Environmental Assessment, Washington, DC. This material is available
electronically at http://www3.epa.gov/ncea/iris/subst/0457.htm.
92	Perera, F.P.; Rauh, V.; Tsai, W-Y.; et al. (2002). Effect of transplacental exposure to environmental pollutants on
birth outcomes in a multiethnic population. Environ Health Perspect. 111: 201 -205.
93	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.
94	U. S. EPA. 1998. 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://www3 .epa. gov/iris/subst/0436.htm.

-------
95 U. S. EPA. 1998. 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://www3.epa. gov/iris/subst/0436.htm.
96U. S. EPA. (1998). 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://www3.epa.gov/iris/subst/0436.htm.
97	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/recordisplav.cfm?deid=84403.
98	NTP. (2014). 13th Report on Carcinogens. Research Triangle Park, NC: U.S. Department of Health and Human
Services, Public Health Service, National Toxicology Program.
99	International Agency for Research on Cancer (IARC). (2002). Monographs on the Evaluation of the Carcinogenic
Risk of Chemicals for Humans. Vol. 82. Lyon, France.
100U. 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://www3.epa.gov/iris/subst/0436.htm.
101	U.S. EPA. (1998). Toxicological Review of Naphthalene. Environmental Protection Agency, Integrated Risk
Information System (IRIS), Research and Development, National Center for Environmental Assessment,
Washington, DC http://www3.epa. gov/iris/subst/0436.htm.
102	U.S. EPA Integrated Risk Information System (IRIS) database is available at: www3.epa.gov/iris.
103	Health Effects Institute Panel on the Health Effects of Traffic-Related Air Pollution. (2010). Traffic-related air
pollution: a critical review of the literature on emissions, exposure, and health effects. HEI Special Report 17.
[Online at http://www.healtheffects.org],
104	Bureau of Labor Statistics. (2009). American Time Use Survey. [Online at http://www.bls.gov/tus].
105	Bureau of Transportation Statistics. (2003). Highlights of the 2001 National Household Travel Survey. Report
BTS03-05. [Online at http://www.bts.gov].
106	Baja, E.S.; Schwartz, J.D.; Wellenius, G.A.; Coull, B.A.; Zanobetti, A.; Vokonas, P.S.; Suh, H.H. (2010).
Traffic-related air pollution and QT interval: modification by diabetes, obesity, and oxidate stress gene
polymorphisms in the Normative Aging Study. Environ Health Perspect 118: 840-846. doi:10.1289/ehp.0901396.
107	Zanobettia, A.; Stone, P.H.; Speizer, F.E.; Schwarz, J.D.; Coull, B.A.; Suh, H.H.; Nearing, B.D.; Mittleman,
M.A.; Verrier, R.L.; Gold, D.R. (2009). T-wave alternans, air pollution and traffic in high-risk subjects. Am J
Cardiol 104: 665-670. doi:10.1016/j.amjcard.2009.04.046.
108	Brook, R.D.; Rajagopalan, S.; Pope, C.A.; Brook, J.R.; Bhatnagar, A.; Diez-Rouz, A.V.; Holguin, F.; Hong, Y.;
Luepker, R.V.; et la. (2010). Particulate matter air pollution and cardiovascular disease: An Update To The
Scientific Statement From The American Heart Association. Circulation 121: 2331-2378.
doi: 10.1161/CIR. ObO 13e3181dbecel.
109	Bastain, T.M.; Gilliland, F.D.; Li, Y.; Saxon, A.; Diaz-Sanchez, D. (2003) Intraindividual reproducibility of nazal
allergic responses to diesel exhaust particles indicates a susceptible phenotype. Clinical Immunol 109: 130-136.
110	Gilliland, F.D.; Li, Y.; Diaz-Sanchez, D. (2004) Effect of glutathione-S-transferase Ml and PI genotypes on
xenobiotic enhancement of allergic responses: randomized, placebo-controlled crossover study. Lancet 363:119-
125.
111	Svartengren, M., Strand, V.; Bylin, G. Jarup, L.; Pershagen, G. (2000) Short-term exposure to air pollution in a
road tunnel enhances the asthmatic response to allergen. EurRespirJ 15: 716-724.
112	Vrijheid, M.; Martinez, D.; Manzanares, S.; Dadvand, P.; Schembari, A.; Rankin, F.; Nieuwenhuijsen, M. (2011).
Ambient air pollution and risk of congenital anomalies: a systematic review and meta-analysis. Environ Health
Perspect 119: 598-606. doi: 10.1289/ehp. 1002946.
113	Boothe, VL.; Boehmer, T.K.; Wendel, A.M.; Yip, F.Y. (2014) Residential traffic exposure and childhood
leukemia: a systematic review and meta-analysis. AmJPrevMed46: 413-422.
114	Sun, X.; Zhang, S.; Ma, X. (2014) No association between traffic density and risk of childhood leukemia: a
meta-analysis. Asia Pac J Cancer Prev 15: 5229-5232.
115	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-2010-0162.
116	National Research Council, (1993). Protecting Visibility in National Parks and Wilderness Areas. National
Academy of Sciences Committee on Haze in National Parks and Wilderness Areas. National Academy Press,

-------
Washington, DC. This book can be viewed on the National Academy Press Website at
http://www.nap.edu/books/0309048443/html/.
117	Sisler, J.F. 1996. Spatial and seasonal patterns and long-term variability of the composition of the haze in the
United States: an analysis of data from the IMPROVE network. CIRA Report, ISSN 0737-5352-32, Colorado State
University.
118	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.
119	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.
120	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. pg 9-19 through 9-23.
121	73 FR 16486 (March 27, 2008).
122	Chapter 9, Section 9.3.4 of U.S. EPA, 2013 Integrated Science Assessment for Ozone and Related Photochemical
Oxidants. Office of Research and Development/National Center for Environmental Assessment. U.S.
Environmental Protection Agency. EPA 600/R-10/076F.
123	73 FR 16492 (March 27, 2008).
124	73 FR 16493-16494 (March 27, 2008).
125	73 FR 16490/ 16497 (March 27, 2008).
126	U.S. EPA. Integrated Science Assessment of Ozone and Related Photochemical Oxidants (Final Report). U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R-10/076F, 2013. The ISA is available at
http://cljpub.epa. gov/ncea/isa/recordisplay.cfm?deid=247492#Download.
127	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.
128	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.
129	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.
130	U.S. Environmental Protection Agency (U.S. EPA). 2008. Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria National (Final Report). National Center for Environmental Assessment, Research
Triangle Park, NC. EPA/600/R-08/139. December. Available on the Internet at
.
131	U.S. Environmental Protection Agency (U.S. EPA). 2008. Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria National (Final Report). National Center for Environmental Assessment, Research
Triangle Park, NC. EPA/600/R-08/139. December. Available on the Internet at
.
132	U.S. Environmental Protection Agency (U.S. EPA). 2008. Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria National (Final Report). National Center for Environmental Assessment, Research
Triangle Park, NC. EPA/600/R-08/139. December. Available on the Internet at
.
133	U.S. Environmental Protection Agency (U.S. EPA). 2008. Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria National (Final Report). National Center for Environmental Assessment, Research
Triangle Park, NC. EPA/600/R-08/139. December. Available on the Internet at
.
134	U.S. Environmental Protection Agency (U.S. EPA). 2008. Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria National (Final Report). National Center for Environmental Assessment, Research
Triangle Park, NC. EPA/600/R-08/139. December. Available on the Internet at
.
135	Hutchison, R., and C.E. Kraft. 1994. "Hmong Fishing Activity and Fish Consumption." Journal of Great Lakes
Research 20(2):471-487.
136	Peterson, D.E., M.S. Kanarek, M.A. Kuykendall, J.M. Diedrich, H.A. Anderson, P.L. Remington, and T.B.
Sheffy. 1994. "Fish Consumption Patterns and Blood Mercury Levels in Wisconsin Chippewa Indians." Archives of
Environmental Health 49(l):53-58.
137	U.S. Environmental Protection Agency (U.S. EPA). 2008. Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria National (Final Report). National Center for Environmental Assessment, Research
Triangle Park, NC. EPA/600/R-08/139. December. Available on the Internet at
.

-------
138	Joslin, J.D., Kelly, J.M., and van Miegroet, H. 1992. "Soil chemistry and nutrition of North American spruce-fir
stands: evidence for recent change. " Journal of Environmental Quality, 21, 12-30.
139	U.S. Environmental Protection Agency (U.S. EPA). 2008. Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria National (Final Report). National Center for Environmental Assessment, Research
Triangle Park, NC. EPA/600/R-08/139. December. Available on the Internet at
.
140	DeHayes, D.H., P.G. Schaberg, G.J. Hawley, and G.R. Strimbeck. 1999. "Acid rain impacts on calcium nutrition
and forest health." Bioscience 49(10):789-800.
141	Bricker, S., B. Longstaff, W. Dennison, A. Jones, K. Boicourt, C. Wicks, and J. Woerner. 2007. Effects of
Nutrient Enrichment In the Nation's Estuaries: A Decade of Change. NOAA Coastal Ocean Program Decision
Analysis Series No. 26. National Centers for Coastal Ocean Science, Silver Spring, MD. 328 pp.
142	Valigura, R.A., R.B. Alexander, M.S. Castro, T.P. Meyers, H.W. Paerl, P.E. Stacy, and R.E. Turner. 2001.
Nitrogen Loading in Coastal Water Bodies: An Atmospheric Perspective. Washington, DC: American Geophysical
Union.
143	U.S. Environmental Protection Agency (U.S. EPA). 2008. Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria National (Final Report). National Center for Environmental Assessment, Research
Triangle Park, NC. EPA/600/R-08/139. December. Available on the Internet at
.
144	U.S. Environmental Protection Agency (U.S. EPA). 2008. Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria National (Final Report). National Center for Environmental Assessment, Research
Triangle Park, NC. EPA/600/R-08/139. December. Available on the Internet at
.
145	U.S. Environmental Protection Agency (U.S. EPA). 2008. Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria National (Final Report). National Center for Environmental Assessment, Research
Triangle Park, NC. EPA/600/R-08/139. December. Available on the Internet at
.
146	Hutchinson J; Maynard D; Geiser L. (1996). Air quality and lichens - a literature review emphasizing the Pacific
Northwest, USA. Pacific Northwest Region Air Resource Management Program; U.S. Forest Service; U.S.
Department of Agriculture (USD A).
147	U.S. Environmental Protection Agency (U.S. EPA). 2008. Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria National (Final Report). National Center for Environmental Assessment, Research
Triangle Park, NC. EPA/600/R-08/139. December. Available on the Internet at
.
148	U.S. Environmental Protection Agency (U.S. EPA). 2008. Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria National (Final Report). National Center for Environmental Assessment, Research
Triangle Park, NC. EPA/600/R-08/139. December. Available on the Internet at
.
149	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .
150	Grantz DA; Garner JHB; Johnson DW. 2003. "Ecological effects of particulate matter." Environ Int, 29: 213-
239.
151	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .
152	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .
153	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .
154	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .

-------
155	Chameides. W.L., Yu, H., Liu, S.C., Bergin, M., Zhou, X., Mearns, L., Wang, G., Kiang, C.S., Saylor, R.D., Luo,
C. Huang, Y., Steiner, A., and Giorgi, F. 1999. "Case study of the effects of atmospheric aerosols and regional haze
on agriculture: an opportunity to enhance crop yields in China through emission controls?" PNAS, 96: 13626-13633.
156	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .
157	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .
158	Parrish ZD; White JC; Isleyen M; Gent MPN; Iannucci-Berger W; Eitzer BD; Kelsey JW; Mattina MI. 2006.
"Accumulation of weathered polycyclic aromatic hydrocarbons (PAHs) by plant and earthworm species."
Chemosphere, 64: 609-618.
159	Simcik M.F., Eisenreich S.J., Lioy P.J. 1999. "Source apportionment and source/sink relationship of PAHs in the
coastal atmosphere of Chicago and Lake Michigan." Atmospheric Environment, 33, 5071-5079.
160	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.
161	Arzavus 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.
162	Cotrufo, M.F., De Santo A.V., Alfani A., Bartoli G., De Cristofaro A. 1995. "Effects of urban heavy metal
pollution on organic matter decomposition in Quercus ilex L. Woods." Environmental Pollution, 89(1), 81-87.
163	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.
164	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .
165	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .
166	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .
167	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .
168	Hageman KJ; Simonich SL; Campbell DH; Wilson GR; Landers DH. 2006. "Atmospheric deposition of current-
use and historic-use pesticides in snow at national parks in the western United States." Environ Sci Technol, 40:
3174-3180.
169	Landers DH; Simonich SL; Jaffe DA; Geiser LH; Campbell DH; Schwindt AR; Schreck CB; Kent ML; Hafner
WD; Taylor HE; Hageman KJ; Usenko S; Ackerman LK; Schrlau JE; Rose NL; Blett TF; Erway MM. (2008). The
Fate, Transport and Ecological Impacts of Airborne Contaminants in Western National Parks (USA). EPA/600/R-
07/138. U.S. Environmental Protection Agency, Office of Research and Development, NHEERL, Western Ecology
Division. Corvallis, Oregon.
170	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .
171	Irving, P.M., e.d. 1991. Acid Deposition: State of Science and Technology, Volume III, Terrestrial, Materials,
Health, and Visibility Effects, The U.S. National Acid Precipitation Assessment Program, Chapter 24, page 24-76.
172U.S. EPA. (1991). Effects of organic chemicals in the atmosphere on terrestrial plants. EPA/600/3-91/001.
173	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.
174	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.
175	Viskari E-L. (2000). Epicuticular wax of Norway spruce needles as indicator of traffic pollutant deposition.
Water, Air, and Soil Pollut. 121:327-337.

-------
176Ugrekhelidze D, F Korte, G Kvesitadze. (1997). Uptake and transformation of benzene and toluene by plant
leaves. Ecotox. Environ. Safety 37:24-29.
177	Kammerbauer H, H Selinger, R Rommelt, A Ziegler-Jons, D Knoppik, B Hock. (1987). Toxic components of
motor vehicle emissions for the spruce Picea abies. Environ. Pollut. 48:235-243.
178	U.S. EPA, 2011. Our Nation's Air: Status and Trends through 2010. EPA-454/R-12-001. February 2012.
Available at: http://www3.epa.gov/airtrends/2011/.
179	U.S. EPA. (2015) Summary of Results for the 2011 National-Scale Assessment.
https://www3.epa.gov/sites/production/files/2015-12/documents/2011-nata-summary-results.pdf.
180	Health Effects Institute Panel on the Health Effects of Traffic-Related Air Pollution. (2010) Traffic-related air
pollution: a critical review of the literature on emissions, exposure, and health effects. HEI Special Report 17.
Available at http://www.healtheffects.org],
181	70 FR 19844 (April 14, 2005).
182	EPA 2014. Fact Sheet: Final Area Designations for the Annual Fine Particle Standard.
https://www3 .epa.gov/pmdesignations/2012standards/final/20141218fs.pdf.
183	74 FR 58688 (November 13, 2009) and 76 FR 6056 (February 3, 2011).
184	77FR 30088 (May 21, 2012) and 77 FR 34221 (June 11, 2012).
185	https://www3.epa.gov/ozone-pollution/2015-ozone-naaqs-timelines.
186	U.S. EPA. (2012). Fact Sheet - Air Quality Designations for the 2010 Primary Nitrogen Dioxide (NO2) National
Ambient Air Quality Standards. http://www3.epa.gov/airquality/nitrogenoxides/designations/pdfs/20120120FS.pdf.
187	U.S. Environmental Protection Agency (2013). Revision to Ambient Nitrogen Dioxide Monitoring
Requirements. March 7, 2013. http://www3.epa.gov/airquality/nitrogenoxides/pdfs/20130307fr.pdf.
188	U.S. EPA (2015) 2011 National-Scale Air Toxics Assessment, https://www3.epa.gov/national-air-toxics-
assessment/2011 -nata-assessment-results#emissions.
189	U.S. EPA (2015) 2011 National-Scale Air Toxics Assessment, https://www3.epa.gov/national-air-toxics-
assessment/2011 -nata-assessment-results#emissions.
190	U.S. Environmental Protection Agency (2007). Control of Hazardous Air Pollutants from Mobile Sources; Final
Rule. 72 FR 8434, February 26, 2007.
191	U.S. EPA. (2015) 2011 NATA: Assessment Results https://www3.epa.gov/national-air-toxics-
assessment/2011 -nata-assessment-results.
192	U.S. EPA. (2015) 2011 NATA: Assessment Results, https://www3.epa.gov/national-air-toxics-assessment/2011-
nata-assessment-results.
193	U.S. Environmental Protection Agency (EPA). 2009. Integrated Science Assessment for Particulate Matter. U.S.
Environmental Protection Agency. Research Triangle Park. EPA/600/R-08/139F.
194	EPA's Report on the Environment: Regional Haze, https://clipub.epa.gov/roe/indicator.cfm?i=21
195	Trijonis, J.C. et al. 1987. Preliminary extinction budget results from the RESOLVE program, pp. 872-883. In:
P.J. Bhardwaja, et.. al.. Visibility Protection Research and Policy Aspects. Air Pollution Control Assoc., Pittsburgh,
PA.
196	Trijonis, J.C. et al. 1988. RESOLVE Project Final Report: Visibility conditions and Causes of Visibility
Degradation in the Mojave Desert of California. NWC TP #6869. Naval Weapons Center, China Lake, CA.
197	Irving, Patricia M., e.d., 1991. Acid Deposition: State of Science and Technology, Volume III, Terrestrial,
Materials, Health, and Visibility Effects, The U.S. National Acid Precipitation Assessment Program, Chapter 24,
page 24-76.
198	EPA Report on the Environment. Accessed in 2016, https://c!pub.epa.gov/roe/indicator.cfm?i=l. Based on data
from the NADP/National Trends Network (accessed 2014, http://nadp.sws.uiuc.edu) and the Clean Air Status and
Trends Network (accessed 2014, https://www3.epa.gov/castnet.).
199	"EPA's Denial of the Petitions to Reconsider the Endangerment and Cause or Contribute Findings for
Greenhouse Gases under Section 202(a) of the Clean Air Act," 75 Fed. Reg. 49,556 (Aug. 13, 2010)
("Reconsideration Denial").
200	Intergovernmental Panel on Climate Change (IPCC). 2012: Managing the Risks of Extreme
Events and Disasters to Advance Climate Change Adaptation. A Special Report of Working
Groups I and II of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK,
and New York, NY, USA.
201	Intergovernmental Panel on Climate Change (IPCC). 2013. Climate Change 2013: The Physical Science Basis.
Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate
Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, Intergovernmental
Panel on Climate Change (IPCC). 2014. Climate Change 2014: Impacts, Adaptation, and Vulnerability.

-------
Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate
Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, Intergovernmental
Panel on Climate Change (IPCC). 2014. Climate Change 2014: Mitigation of Climate Change. Contribution of
Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA.
202	Melillo, Jerry M., Terese (T.C.) Richmond, and Gary W. Yohe, Eds. 2014. Climate Change Impacts in the United
States: The Third National Climate Assessment. U.S. Global Change Research Program. Available at
http ://nca2014 .globalchange. gov.
203	National Research Council (NRC). 2010. Ocean Acidification: A National Strategy to Meet the Challenges of a
Changing Ocean. National Academies Press. Washington, DC.
204	National Research Council (NRC). 2011. Climate Stabilization Targets: Emissions, Concentrations, and Impacts
over Decades to Millennia. National Academies Press, Washington, DC.
205	National Research Council (NRC) 2011. National Security Implications of Climate Change for U.S. Naval
Forces. National Academies Press. Washington, DC.
206	National Research Council (NRC). 2012. Sea-Level Rise for the Coasts of California, Oregon, and Washington:
Past, Present, and Future. National Academies Press. Washington, DC.
207	National Research Council (NRC). 2012. Sea-Level Rise for the Coasts of California, Oregon, and Washington:
Past, Present, and Future. National Academies Press. Washington, DC.
208	National Research Council (NRC). 2013. Climate and Social Stress: Implications for Security Analysis. National
Academies Press. Washington, DC.
209	National Research Council (NRC). 2013. Abrupt Impacts of Climate Change: Anticipating Surprises. National
Academies Press. Washington, DC.
210	Brenkert A, S. Smith, S. Kim, and H. Pitcher, 2003: Model Documentation for the MiniCAM. PNNL-14337,
Pacific Northwest National Laboratory, Richland, Washington.
211	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.
212	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.
213	Thompson AM, KV Calvin, SJ Smith, GP Kyle, A Volke, P Patel, S Delgado-Arias, B Bond-Lamberty, MA
Wise, LE Clarke and JA Edmonds. 2010. "RCP4.5: A Pathway for Stabilization of Radiative Forcing by 2100."
Climatic Change (in review).
214	Clarke, L., J. Edmonds, H. Jacoby, H. Pitcher, J. Reilly, R. Richels, (2007) Scenarios of Greenhouse Gas
Emissions and Atmospheric Concentrations. Sub-report 2.1A 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.).
215	Wigley, T.M.L. 2008. MAGICC 5.3.v2 User Manual. UCAR - Climate and Global Dynamics Division, Boulder,
Colorado, http://www.cgd.ucar.edu/cas/wigley/magicc/.
216	Meehl, G. A. et al. (2007) Global Climate Projections. In: Climate Change 2007: The Physical Science Basis.
Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate
Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)].
Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.
217	National Research Council, 2011. Climate Stabilization Targets: Emissions, Concentrations, and Impacts over
Decades to Millenia. Washington, DC: National Academies Press.
218	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.
219	Pierrot, D. E. Lewis, and D. W. R. Wallace. 2006. MS Excel Program Developed for CO2 System Calculations.
ORNL/CDIAC-105a. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S.
Department of Energy, Oak Ridge, Tennessee, doi: 10.3334/CDIAC/otg.CO2SYS_XLS_CDIAC105a
220	Mehrbach, C., C. H. Culberson, J. E. Hawley, and R. N. Pytkowicz. 1973. Measurement of the apparent
dissociation constants of carbonic acid in seawater at atmospheric pressure. Limnology and Oceanography 18:897-
907.
221	Dickson, A. G. and F. J. Millero. 1987. A comparison of the equilibrium constants for the dissociation of
carbonic acid in seawater media. Deep-Sea Res. 34, 1733-1743. (Corrigenda. Deep-Sea Res. 36, 983).

-------
222	A. G. Dickson. 1990. Thermodynamics of the dissociation of boric acid in synthetic sea water from 273.15 to
318.15 K. Deep-Sea Res. 37, 755-766.
223	A. G. Dickson. 1990. Thermodynamics of the dissociation of boric acid in synthetic sea water from 273.15 to
318.15 K. Deep-Sea Res. 37, 755-766.
224	Dickson, A. G. 2003. Certificate of Analysis - Reference material for oceanic CO2 measurements (Batch #62,
bottled on August 21, 2003). Certified by Andrew Dickson, Scripps Institution of Oceanography. November 21,
2003.
Dickson, A. G. 2005. Certificate of Analysis - Reference material for oceanic CO2 measurements (Batch #69,
bottled on January 4, 2005). Certified by Andrew Dickson, Scripps Institution of Oceanography. July 12, 2005.
Dickson, A. G. 2009. Certificate of Analysis - Reference material for oceanic CO2 measurements (Batch #100,
bottled on November 13, 2009). Certified by Andrew Dickson, Scripps Institution of Oceanography. February 10,
2010.
225	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.
226	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.
227	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.
228	Brown, N., Allen, D., Amar, P., Kallos, G., McNider, R., Russell,, A., Stockwell, W. (September 2011). Final
Report: Fourth Peer Review of the CMAQ Model, NERL/ORD/EPA. U.S. EPA, Research Triangle Park, NC.,
http://www3.epa.gov/asmdnerl/Reviews/201 l_CMAQ_Review_FinalReport.pdf.
229	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.
230	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.
231	U.S. EPA, 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.
232	Community Modeling and Analysis System (CMAS) website: http://www.cmascenter.org.. RELEASENOTES
for CMAQv5.0 - February 2012.
233	Community Modeling and Analysis System (CMAS) website: http://www.cmascenter.org.. RELEASE NOTES
for CMAQv5.0.1 - July 2012.
234	Community Modeling and Analysis System (CMAS) website: http://www.cmascenter.org. CMAQ version 5.0.2
(April 2014 release) Technical Documentation. - May 2014.
235	Skamarock, W.C., Klemp, J.B., Dudhia, J., Gill, D.O., Barker, D.M., Duda, M.G., Huang, X., Wang, W., Powers,
J.G., 2008. A Description of the Advanced Research WRF Version 3.
236	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/.
237	Skamarock, W., Klemp, J., Dudhia, J., Gill, D., Barker, D., Dudha, M., Huang, X-Y., Wang, W., Powers, J., June
2008. A Description of the Advanced Research WRF Version 3, NCAR Technical Note, NCAR/TN-475+STR,
Mesoscale and Microscale Meteorology Division, NCAR, Boulder, Colorado.
238	Yantosca, R.M., Long, M.S., Payer, M., and Cooper, M. 2012: GEOS-Chemv9-01-03 Online User's Guide,
available at: http://acmg.seas.harvard.edu/geos/doc/man/.
239	Henderson, B.H., Akhtar, F., Pye, O.T., Napelenok, S.L., Hutzell, W.T. 2014. A database and tool for boundary
conditions for regional air quality modeling: description and evaluation. Geosci. Model Dev., 7, 339-360.
240	Karl, T. R. and Koss, W. J., 1984: "Regional and National Monthly, Seasonal, and Annual Temperature
Weighted by Area, 1895-1983." Historical Climatology Series 4-3, National Climatic Data Center, Asheville, NC,
38 pp.
241	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.
242	U.S. EPA, 2011, Final Cross State Air Pollution Rule Air Quality Modeling TSD.

-------
243	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.
244	Yarwood G, Rao S, Yocke M, Whitten GZ (2005) Updates to the Carbon Bond Chemical Mechanism: CB05.
Final Report to the U.S. EPA, RT-0400675, December 8, 2005,
http://www.camx.com/publ/pdfs/CB05_Final_Report_120805.pdf.
245	U.S. EPA. 2014. Memorandum to Docket: Updates to MOVES for the Tier 3 FRM Analysis.
246	Dodge, M.C., 2000. Chemical oxidant mechanisms for air quality modeling: critical review. Atmospheric
Environment 34, 2103-2130.
247	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.iupac-kinetic.ch.cam.ac.uk/index.html.
248	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.iupac-kinetic.ch.cam.ac.uk/index.html.
249	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.iupac-kinetic.ch.cam.ac.uk/index.html.
250	Sander, S.P., Friedl, R.R., Golden, D.M., Kurylo, M.J., Huie, R.E., Orkin, V.L., Moortgat, G.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.
251	J.G. Calvert, A. Mellouki, J.J. Orlando, M.J. Pilling, and T.J. Wallington, 2011. The mechanisms of atmospheric
oxidation of the oxygenates. Oxford University Press, New York/Oxford.
252	Sander, S.P., Friedl, R.R., Golden, D.M., Kurylo, M.J., Huie, R.E., Orkin, V.L., Moortgat, G.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.
253	Yarwood, G., Rao, S., Yocke, M., Whitten, G.Z., 2005. Updates to the Carbon Bond Mechanism: CB05. Final
Report to the U.S. EPA, RT-0400675. Yocke and Company, Novato, CA.
254	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.
255	Sander, S.P., Friedl, R.R., Golden, D.M., Kurylo, M.J., Huie, R.E., Orkin, V.L., Moortgat, G.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.
256	Sander, S.P., Friedl, R.R., Golden, D.M., Kurylo, M.J., Huie, R.E., Orkin, V.L., Moortgat, G.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.
257	Luecken, D.J., Hutzell, W.T., Strum, M., Pouliot, G., 2012. Regional sources of atmospheric formaldehyde and
acetaldehyde, and implications for atmospheric modeling. Atmospheric Environment, 47, 477-490.
258	Atkinson R, Arey J (2003) Atmospheric Degradation of Volatile Organic Compounds. ChemRev 103: 4605-
4638.
259	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.iupac-kinetic.ch.cam.ac.uk/index.html.
260	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.
261	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.
262	Offenberg JH, CW Lewis, M Lewandowski, M Jaoui, TE Kleindienst, EO Edney (2007) Contributions of

-------
Toluene and ©nene to SOA Formed in an Irradiated Toluene/ Ej3nene.NO,/Air Mixture: Comparison of Results
Using 14C Content and SOA Organic Tracer Methods, Environ Sci Technol 41: 3972-3976.
263	Pandis, S.N., Harley, R.A., Cass, G.R., Seinfeld, J.H. (1992) Secondary organic aerosol formation and transport.
Atmos Environ 26, 2269-2282.
264	Takekawa, H. Minoura, H. Yamazaki, S. (2003) Temperature dependence of secondary organic aerosol formation
by photo-oxidation of hydrocarbons. Atmos Environ 37: 3413-3424.
265	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.
266	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.
267	Griffin, R. J.; Cocker, D. R.; Seinfeld, J. H.; Dabdub, D. (1999) Estimate of global atmospheric organic aerosol
from oxidation of biogenic hydrocarbons. Geophys Res Lett 26 (17) 2721- 2724.
268	Lewis, C. W.; Klouda, G. A.; Ellenson, W. D. (2004) Radiocarbon measurement of the biogenic contribution to
summertime PM-2.5 ambient aerosol in Nashville, TN. Atmos Environ 38 ( 35) 6053- 6061.
269	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.
270	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.
271	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.
272	Carlton, A. G., P. V. Bhave, S. L. Napelenok, E. O. Edney, G. Sarwar, R. W. Pinder, G. A. Pouliot, and M.
Houyoux. Model Representation of Secondary Organic Aerosol in CMAQv4.7. Env. Sci. & Techno. 2010, 44 (22),
8553-8560.
273	Simulating the Degree of Oxidation in Atmospheric Organic Particles. Simon, H and Bhave, Environ. Sci.
Technol., 2012, 46 (1), pp 331-339.
274	U.S. EPA, 2010. Our Nations Air, Status and Trends through 2008. EPA 454/R-09-002, February 2010.
http://www3.epa.gov/airtrends/2010.
275	Hildebrandt et al. 2009 ACP http://www.atmos-chem-phys.net/9/2973/2009/acp-9-2973-2009.html.
276	Ng et al. 2007 ACP http://www.atmos-chem-phys.net/7/3909/2007/acp-7-3909-2007.html.
277	U.S. EPA. 2015. 2011 NATA: Assessment Results, https://www3.epa.gov/national-air-toxics-assessment/2011-
nata-assessment-results#modeled.
278	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://www3.epa.gov/otaq/regs/toxics/420r07002.pdf.
279	Ng et al. 2007 ACP http://www.atmos-chem-phys.net/7/3909/2007/acp-7-3909-2007.html.
280	Lim, Y.B., Ziemann, P.J. (2009) Effects of Molecular Structure on Aerosol Yields from OH Radical-Initiated
Reactions of Linear, Branched, and Cyclic Alkanes in the Presence of NOx. Environ Sci Technol 43 (7): 2328-2334.
281	Kleindienst, T.E. (2008) Hypothetical SOA Production from Ethanol Photooxidation. Memo to the Docket EPA-
HQ-0 AR-2005-0161.
282	Turpin, B.J., Huntzicker, J. J., Larson, S.M., Cass, G.R. (1991) Los Angeles Summer Midday Particulate Carbon:
Primary and Secondary Aerosol. Environ Sci Technol 25: 1788-1793.
283	Turpin, B.J., Huntzicker, J.J. (1995) Identification of Secondary Organic Aerosol Episodes and Quantitation of
Primary and Secondary Organic Aerosol Concentrations During SCAQS. Atmos Environ 29(23): 3527-3544.
284	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.
285	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.
286	Offenberg JH, CW Lewis, M Lewandowski, M Jaoui, TE Kleindienst, EO Edney (2007) Contributions of
Toluene and ©nene to SOA Formed in an Irradiated Toluene/ Ej3ncnc.NO./Air Mixture: Comparison of Results
Using 14C Content and SOA Organic Tracer Methods, Environ Sci Technol 41: 3972-3976.

-------
287	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 Ejjfncnc. Environ Sci Technol 41(5): 1628-1634.
288	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/S02/air mixtures and their detection in ambient PM2 5 samples collected in the Eastern United States.
Atmos Environ 39: 5281-5289.
289	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.
290	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.
291	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.
292	Pye et al. 2013 ES&T Epoxide pathways improve model predictions of isoprene markers and reveal key role of
acidity in aerosol formation, http://pubs.acs.org/doi/abs/10.1021/es402106h.
293	Carlton etal. 2010 ES&T Can biogenic SOAbe controlled? http://pubs.acs.org/doi/full/10.1021/es903506b
294	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.
295	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.
296	Henze DK, JH Seinfeld (2006) Global secondary organic aerosol from isoprene oxidation. Geophys Res Lett 33:
L09812. doi: 10.1029/2006GL025976.
297	Hildebrandt, 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-2011-0135.
298	Ng, N. L., Kroll, J. H., Chan, A. W. H., Chabra, P. S., Flagan, R. C., Seinfield, J. H., Secondary organic aerosol
formation from ///-xylene, toluene, and benzene, Atmospheric Chemistry and Physics Discussion, 7, 3909-3922,
2007. Docket EPA-HQ-OAR-2011-0135.
299	Henze, D. K., Seinfeld, J. H., Ng, N. L., Kroll, J. H., Fu, T.-M., Jacob, D. J., and Heald, C. L. (2008) Global
modeling of secondary organic aerosol formation from aromatic hydrocarbons: high- vs. low-yield pathways,
Atmos. Chem. Phys., 8, 2405-2420, doi: 10.5194/acp-8-2405-2008.
300	Lane, T. E., Donahue, N.M. and Pandis, S.N. (2008) Simulating secondary organic aerosol formation using the
volatility basis-set approach in a chemical transport model, Atmos. Environ., 42, 7439-7451, doi:
10.1016/j.atmosenv.2008.06.026.
301	Carlton, A.G., Bhave, P.V., Napelenok, S.L., Edney, E.O., Sarwar, g., Pinder, R.W., Pouliot, G.A., Houyoux, M.,
(2010). Model Representation of Secondary Organic Aerosol in CMAQv4.7. Environ Sci Technol 44(22), 8553-
8560.
302	Parikh, H.M., Carlton, A.G., Vizuete, W., and Kamen, R.M. (2011) Modeling secondary organic aerosol using a
dynamic partitioning approach incorporating particle aqueous-phase chemistry, Atmospheric Environment, 45,
1126-1137.
303	Volkamer, R., J.L. Jimenez, F. SanMartini,K.Dzepina,Q. Zhang,D. Salcedo,L. T. Molina, D. R.Worsnop, andM.
J. Molina (2006), Secondary organic aerosol formation from anthropogenic air pollution: Rapid and higher than
expected, Geophys. Res. Lett., 33, L17811, doi:10.1029/2006GL026899.
304	Baker, K. R., Carlton, A. G., Kleindienst, T. E., Offenberg, J. H., Beaver, M. R., Gentner, D. R., Goldstein, A.
H., Hayes, P. L., Jimenez, J. L., Gilman, J. B., de Gouw, J. A., Woody, M. C., Pye, H. O. T., Kelly, J. T.,
Lewandowski, M., Jaoui, M., Stevens, P. S., Brune, W. H., Lin, Y.-H., Rubitschun, C. L., and Surratt, J. D.: Gas and
aerosol carbon in California: comparison of measurements and model predictions in Pasadena and Bakersfield,
Atmos. Chem. Phys., 15, 5243-5258, doi:10.5194/acp-15-5243-2015, 2015.
305	Napelenok, S. L., H. Simon, P. Bhave, H. O. T. Pye, G. A. Pouliot, R. J. Sheesley, J. J. Schauer, Diagnostic air
quality model evaluation of source specific primary and secondary fine particulate carbon, Environ. Sci. Technol.,
doi: 10.1021/es4033024w, 2014.
306	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.iupac-kinetic.ch.cam.ac.uk/index.html. Docket EPA-HQ-OAR-2011-0135.
307	Sander, S.P., Friedl, R.R., Golden, D.M., Kurylo, M.J., Huie, R.E., Orkin, V.L., Moortgat, G.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-2011-0135.
308	Finlayson-Pitts BJ, Pitts JN Jr. (1986) Atmospheric Chemistry: Fundamentals and Experimental Techniques,
Wiley, New York.
309	Yarwood G, Rao S, Yocke M, Whitten GZ (2005) Updates to the Carbon Bond Chemical Mechanism: CB05.
Final Report to the U.S. EPA, RT-0400675, December 8, 2005.
http://www.camx.eom/publ/pdfs/CB05_Final_Report_120805.pdf. Docket EPA-HQ-OAR-2011-0135,
310	http://www.cmascenter.Org/help/model_docs/cmaq/4.7/RELEASE_NOTES.txt.

-------
Chapter 7: Vehicle-Related Costs, Fuel Savings &
Maintenance Costs
In this chapter, the agencies present estimates of the vehicle -related costs associated with
the standards along with corresponding fuel savings and maintenance costs. For this final rule,
the agencies used two analytical methods for the heavy-duty pick up and van segment by
employing both DOT's CAFE model and EPA's MOVES model. The agencies used EPA's
MOVES model to estimate fuel consumption and emissions impacts for tractor-trailers
(including the engine that powers the tractor), and vocational vehicles (including the engine that
powers the vehicle). Additional calculations were performed to determine corresponding
monetized program costs and benefits. For heavy-duty pickups and vans, the agencies
performed separate analyses, which we refer to as "Method A" and "Method B." In Method A,
the CAFE model was used to project a pathway the industry could use to comply with each
regulatory alternative and the estimated effects on fuel consumption, emissions, benefits and
costs. In Method B, the CAFE model was used to project a pathway the industry could use to
comply with each regulatory alternative, along with resultant impacts on per-vehicle costs, and
the MOVES model was used to calculate corresponding changes in total fuel consumption and
annual emissions. Additional calculations were performed to determine corresponding
monetized program costs and benefits. NHTSA considered Method A as its central analysis and
Method B as a supplemental analysis. EPA considered the results of Method B. The agencies
concluded that both methods led the agencies to the same conclusions and the same selection of
the standards. Throughout this chapter and in later chapters presenting program-related costs and
benefits, engine costs are included along with vehicle-related costs.
7.1 Vehicle Costs, Fuel Savings and Maintenance Costs vs. the Dynamic
Baseline and Using Method A
NHTSA's analysis of the potential costs of the standards combines DOT CAFE model
calculations of HD pickup and van costs with EPA MOVES modeling of vocational vehicle,
tractor and trailer fuel consumption along with EPA analysis of vocational vehicle, tractor and
trailer costs. The analysis includes costs for fuel-saving technology that manufacturers could add
in response to the standards, EPA estimates of the additional compliance and R&D costs for
vocational vehicles and combination tractor trailers, and some additional maintenance costs.
7.1.1 Vehicle Program Costs
In this section, NHTSA presents its estimate of the vehicle- -related costs associated with
the program versus Alternative lb using the CAFE model analysis of HD pickups and vans. The
presentation here summarizes the costs associated with new technology the agencies estimate
manufacturers could add to meet the GHG and fuel consumption standards. The analysis
summarized here provides our estimate of incremental costs on a per vehicle basis and on a MY
lifetime basis. In Chapter 7.2, where EPA presents the Method B analysis, the analogous
information is presented along with costs on an annual, or calendar year, basis for all segments.
For details behind the cost estimates associated with individual technologies, the reader is
directed to Sections III through VI of the Preamble and to Chapter 2 of the RIA.

-------
Note that all discounted costs presented in this chapter, whether in the Calendar Year (or
annual) analysis or the Model Year Lifetime analysis, are discounted back to 2015 at the
discount rate shown in the table(s).
7.1.1.1 Technology Costs
For the HD pickup trucks and vans, the Method A analysis used technology costs
consistent with that used for our recent 2017-2025 light-duty joint rulemaking since most of the
technologies expected for HD pickup trucks and vans are consistent with those expected for the
larger light-duty trucks. The cost estimates presented in the recent light-duty joint rulemaking
were then scaled upward to account for the larger weight, towing capacity, and work demands of
the trucks in these heavier classes. For details on that scaling process and the resultant costs for
individual technologies, the reader is directed to Chapter 2.6 and 2.12 of this RIA. Note also that
all cost estimates have been updated to 2013 dollars for this analysis while the 2017-2025 light-
duty joint rulemaking was presented in 2010 dollars.1 To mark-up the technology costs to
consider indirect costs the agencies use two different methodologies: NHTSA uses the retail
price equivalent (RPE) multiplier, and EPA uses the indirect cost multiplier (ICM). For more
details on these two methodologies see Chapter 2.11.1.2 and Chapter 10 in the RIA and Section
VI.C in the Preamble.
For vocational vehicles, tractors and trailers, consistent with the Phase 1 rule, the
agencies have estimated costs using a different methodology than that employed in the recent
light-duty joint rulemaking establishing fuel economy and GHG standards. In the recent light-
duty joint rulemaking, all fixed costs were included in the hardware costs via ICM's. As such,
the hardware costs presented in that analysis included both the actual hardware and the
associated fixed costs. For the vocational, tractor and trailer segments in this analysis, some of
the fixed costs are estimated separately and are presented separately from the technology costs.
As noted above, all costs are presented in 2013 dollars.
The estimates of vehicle costs are generated relative to two unique "no action" baselines.
The first of these (alternative la, presented below in Chapter 7.2) representing generally flat fuel
consumption improvements, or a fleet of vehicles meeting the Phase 1 heavy-duty requirements.
The second of these (alternative lb and presented here) representing dynamic fuel consumption
improvements, or a fleet of vehicles with some improvement in fuel consumption even without
additional regulatory action. See Section X of the Preamble and Chapter 11 of this RIA for more
detail on these two baselines. As such, costs to comply with the Phase 1 standards are not
included in the estimates here. In fact, in the methodology used for vocational vehicles, tractors
and trailers, there are cases where Phase 1 technologies are being removed in favor of Phase 2
technologies - that is, the technology basis for the Phase 2 standards involves removing certain
of the Phase 1 technologies. In those cases, savings are associated with the removal of the Phase
1 technology. The details of which technologies and where such savings occur are presented in
Chapter 2.11 of the RIA.
For HD pickups and vans, as described in Chapter 2 of this RIA, the agencies used
NHTSA's CAFE model to estimate the cost per vehicle associated with the standards (and

-------
possible alternative).A That model has the capability to look ahead at future standards when
making determinations of how vehicles should be changed to comply. It does this because
redesign cycles do not always line up well with regulatory implementation schedules, so a
manufacturer may choose to redesign a vehicle in MY2018 in preparation for upcoming
MY2021-2027 standards if that particular vehicle is not scheduled for another redesign until, say,
MY2026. The result being new technology costs in years prior to implementation of the
standards. The CAFE model's output would show such costs occurring in years prior to
MY2021. On the other hand, the CAFE model also estimates the potential that credits generated
in earlier model years might be carried forward (i.e., "banked") and then used in later model
years, potentially reducing costs in some model years covered by the analysis.
Table 7-1 presents the average incremental technology costs per vehicle for the program
relative to alternative lb. These tables include both engine and vehicle technologies. For HD
pickups and vans, costs begin with new standards in MY2018, as technology is utilized in
vehicles with early redesign cycles. The costs jump in MY2021 as more complex technologies
are utilized, then generally increase through the remainder of the analysis period. For vocational
vehicles, the costs begin in MY2021, then decrease slightly through MY2023, with an increase
in MY2024, decreasing slightly through MY2026, and followed by a large increase in costs from
MY2027 until the end of the analysis period. The decreasing costs from MY2021 through MY
2023 and MY2024 through MY2026 are due to technology learning, whereby manufacturers can
produce the same technologies at a lower cost. For tractor/trailers, the costs begin in MY2018 as
trailers begin adding new technology to meet the 2018 trailer standards. Costs then increase in
MY2021 as the tractor standards begin through 2027. After 2027, costs begin to decrease due to
learning effects. All costs shown in the table represent the weighted average cost of all vehicles
within the category shown in the heading.
A The CAFE model also provides a full benefit-cost analysis associated with standards, and NHTSA has used this
analysis as part of Method A to provide estimates of the costs and benefits of today's standards. The full benefit-cost
analysis for Method A is presented in Chapters 9 and 10 of this RIA. The full benefit-cost analysis for Method B is
presented in Chapter 8 of this RIA.

-------
Table 7-1 Estimated Technology Costs per Vehicle for the Final Program versus the Dynamic
Baseline and using Method A (2013$) a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
2018
$235
$0
$639
2019
$468
$0
$573
2020
$441
$0
$482
2021
$752
$1,110
$7,248
2022
$774
$1,088
$7,120
2023
$779
$1,027
$6,624
2024
$762
$2,022
$10,925
2025
$950
$1,986
$10,660
2026
$1,347
$1,927
$10,447
2027
$1,335
$2,662
$13,226
2028
$1,468
$2,616
$12,906
2029
$1,486
$2,586
$12,768
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section
I.D; for an explanation of the less dynamic baseline, la, and more dynamic
baseline, lb, please see Preamble Section X.A.I
As noted in the text, MYs 2018-2020 include costs for trailers only, and in MYs
2021 and later the costs include both tractor and trailer costs. Detailed
technology and package costs for all segments can be found in Chapter 2 of this
RIA (notably, see Sections 2.12 and 2.13).
Table 7-2 presents the model year lifetime costs for new technology discounted at 3
percent using Method A. And Table 7-3 presents the model year lifetime costs for new
technology discounted at 7 percent using Method A.

-------
Table 7-2 Discounted MY Lifetime New Technology Costs of the Final Program
Vs. the Dynamic Baseline and using Method A
(3% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$174
$0
$91
$265
2019
$335
$0
$79
$414
2020
$308
$0
$72
$380
2021
$503
$471
$897
$1,871
2022
$498
$450
$862
$1,810
2023
$486
$414
$781
$1,681
2024
$464
$801
$1,283
$2,548
2025
$570
$778
$1,233
$2,581
2026
$795
$744
$1,183
$2,722
2027
$775
$1,009
$1,465
$3,249
2028
$837
$977
$1,403
$3,217
2029
$833
$952
$1,372
$3,157
Sum
$6,578
$6,597
$10,722
$23,897
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an
explanation of the less dynamic baseline, la, and more dynamic baseline, lb, please see Preamble
Section X.A.I
Table 7-3 Discounted MY Lifetime New Technology Costs of the Final Program
Vs. the Dynamic Baseline and using Method A
(7% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$162
$0
$81
$243
2019
$299
$0
$68
$367
2020
$264
$0
$59
$323
2021
$416
$375
$714
$1,505
2022
$397
$345
$660
$1,402
2023
$373
$305
$576
$1,254
2024
$342
$568
$910
$1,820
2025
$404
$532
$842
$1,778
2026
$543
$489
$778
$1,810
2027
$510
$639
$928
$2,077
2028
$530
$595
$855
$1,980
2029
$507
$558
$805
$1,870
Sum
$4,746
$4,407
$7,277
$16,430
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the less dynamic baseline, la, and more dynamic baseline, lb, please see Preamble Section X.A.I

-------
7.1.1.2 Compliance Costs
As noted above, for vocational vehicles and tractor trailers, some fixed costs were
estimated separately from the hardware costs. As such, not all fixed costs are included in the
tables presented in Chapter 7.1.1.1. The agencies have estimated additional and/or new
compliance costs associated with the standards. Normally, compliance program costs would be
considered part of the indirect costs and, therefore, would be accounted for via the markup
applied to direct manufacturing costs. However, since the agencies are proposing new
compliance elements that were not present during development of the indirect cost markups used
in this analysis, additional compliance program costs are being accounted for via a separate
"line-item" here. Note, for HD pickups and vans, the RPE methodology used for Method A
already accounts for these costs. Again, see Chapter 10 of this RIA or Section VI.C of the
Preamble for more on NHTSA's decision to use RPE in Method A.
There are three elements to the compliance costs estimated in this analysis. The first is
for construction of new, or upgrades to existing, test facilities for conducting powertrain testing.
The second costs are for conducting the powertrain tests themselves. And the third is for
reporting of compliance data to EPA and NHTSA. We estimated these latter costs in the Phase 1
rule as $0.24 million, $0.9 million and $1.1 million for HD pickups and vans, vocational and
tractors, respectively, for a total of $2.3 million per year (2009$).2 All of these are industry-
wide, annual costs.
We have estimated reporting costs in this Phase 2 final rule associated with new
powertrain testing within the vocational vehicle program, the increased level of reporting in the
tractor program and an all new compliance program where none has existed to date within the
trailer program. We have estimated those costs, inclusive of the Phase 1 costs, such that the new
GHG program reporting costs are estimated at $1.1 million and $1.2 million for vocational and
tractor programs both in 2013$. All of these are industry-wide, annual costs.
For powertrain testing facility upgrades and construction, we have estimated that 6
manufacturers would upgrade and 5 would construct new facilities at an upgrade cost of $1.2
million and a new construction cost of $1.9 million, all in 2013$. The result being an industry-
wide (but vocational program only) cost of $16.8 million (2013$). This cost would occur once
which we have attributed to CY2020, one year prior to the first year of the Phase 2 vocational
standards.
Lastly, the vocational program is also estimated to incur costs associated with conducting
powertrain testing. We have estimated the cost of testing at $40,000 per test (2013$) and expect
10 tests/year for a total of $400,000/year. We have also estimated that the vocational program
will incur costs associated with conducting transmission efficiency testing at a cost of $24,600
per test (2013$). We have estimated 11 tests per year for a total annual cost of $270,600. We
have also estimated that the vocational program will incur costs associated with conducting axle
efficiency testing at a cost of $12,600 per test (2013$). We have estimated that 9 tests would be
done per year for a total annual cost of $113,400. We have also estimated an annual cost of
$8,700 in tire testing will be incurring by the vocational program.

-------
In the tractor program, we have used the same per test costs noted above for vocational
and have estimated one transmission efficiency test per year for a total annual cost of $24,600
(2013$) and 15 axle efficiency tests per year for a total annual cost of $189,000 (2013$). To
those costs, we have also added $300,000 (2013$) per year in aero-related testing and $5,400
(2013$) per year in tire testing. For the trailer program, we have estimated an annual compliance
program cost of $7 million (2013$) to cover reporting, testing and capital costs.
Table 7-4 and Table 7-5 present the MY lifetime costs for new compliance program
elements at 3 percent and 7 percent, respectively.
Table 7-4 Discounted MY Lifetime Compliance Costs of the Final Program
Vs. The Dynamic Baseline and using Method A
(3% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
0
0
0
2019
0
0
0
2020
14.5
0
14.5
2021
1.6
7.3
8.9
2022
1.5
7.1
8.6
2023
1.5
6.9
8.4
2024
1.4
6.7
8.1
2025
1.4
6.5
7.9
2026
1.3
6.3
7.6
2027
1.3
6.1
7.4
2028
1.3
5.9
7.2
2029
1.2
5.8
7.0
Sum
27.0
58.6
85.6
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D;
for an explanation of the less dynamic baseline, la, and more dynamic baseline, lb,
please see Preamble Section X. A. 1

-------
Table 7-5 Discounted MY Lifetime Compliance Costs of the Final Program
Vs. The Dynamic Baseline and using Method A
(7% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
0
0
0.0
2019
0
0
0.0
2020
12.0
0
12.0
2021
1.2
5.8
7.0
2022
1.2
5.4
6.6
2023
1.1
5.1
6.2
2024
1.0
4.7
5.7
2025
0.9
4.4
5.3
2026
0.9
4.1
5.0
2027
0.8
3.9
4.7
2028
0.8
3.6
4.4
2029
0.7
3.4
4.1
Sum
20.6
40.4
62.0
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D;
for an explanation of the less dynamic baseline, la, and more dynamic baseline, lb,
please see Preamble Section X. A. 1
7.1.1.3 Research & Development Costs
Much like the compliance program costs described above, Method A estimates additional
engine, vocational vehicle and tractor R&D associated with the standards that is not accounted
for via the indirect cost markups used in this analysis for those segments. The necessary R&D
for HD pickups and vans is covered by the indirect costs included as part of the technology costs
described above. In the Phase 1 rule, the agencies estimated the engine R&D costs at $6.8
million (2009$) per engine class per manufacturer per year for five years. In this Phase 2
analysis, both the Method A and Method B analyses estimate this same level of R&D and has
assumed 12 heavy-heavy and 12 medium-heavy HD engine R&D programs would be conducted
for a total of $218 million/year (2013$). In both methods, the agencies assume those costs would
occur annually for 4 years, MYs 2021-2024. The total being $873 million (2013$) over 4 years
(by comparison, the Phase 1 rule estimated a total of $852 million (2009$) over 5 years). To
this, the agencies have estimated an additional $20 million/year spent by vocational vehicle
manufacturers and $20 million/year spent by tractor manufacturers. In the end, the agencies are
estimating a total of over $1 billion in R&D spending above and beyond the level included in the
markups used to estimate indirect costs for these segments. The agencies have not included any
additional R&D would be spent by trailer manufacturers since our trailer technology cost
estimates include R&D conducted by trailer parts suppliers which are subsequently included in
the prices charged by those suppliers to the trailer manufacturer. Additionally, the markups we
have applied to cover indirect costs (see Chapter 2.12 of this RIA) include costs associated with
R&D incurred by the trailer manufacturer.

-------
Table 7-6 and Table 7-7 present the annual costs for R&D spending along with net
present values at 3 percent and 7 percent, and the model year lifetime R&D costs discounted at 3
percent and 7 percent, respectively.
Table 7-6 Discounted MY Lifetime R&D Costs of the Final Program
Vs. The Dynamic Baseline and using Method A
(3% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
0
0
0
2019
0
0
0
2020
0
0
0
2021
108
108
216
2022
105
105
210
2023
102
102
204
2024
99
99
198
2025
0
0
0
2026
0
0
0
2027
0
0
0
2028
0
0
0
2029
0
0
0
Sum
415
415
830
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D;
for an explanation of the less dynamic baseline, la, and more dynamic baseline, lb,
please see Preamble Section X. A. 1

-------
Table 7-7 Discounted MY Lifetime R&D Costs of the Final Program
Vs. The Dynamic Baseline and using Method A
(7% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
0
0
0
2019
0
0
0
2020
0
0
0
2021
86
86
172
2022
81
81
161
2023
75
75
151
2024
70
70
141
2025
0
0
0
2026
0
0
0
2027
0
0
0
2028
0
0
0
2029
0
0
0
Sum
313
313
625
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D;
for an explanation of the less dynamic baseline, la, and more dynamic baseline, lb,
please see Preamble Section X. A. 1
7.1.1.4 Summary of Vehicle-Related Costs of the Program using Method A
Table 7-8 presents the model year lifetime costs associated with the final program
discounted at 3 percent relative to the dynamic baseline and using Method A. Table 7-9 presents
the model year lifetime costs associated with the final program discounted at 7 percent relative to
the dynamic baseline and using Method A.

-------
Table 7-8 Discounted MY Lifetime Vehicle-Related Costs of the Final Program
Vs. The Dynamic Baseline and using Method A
(3% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$174
$0
$91
$265
2019
$335
$0
$79
$414
2020
$308
$15
$72
$395
2021
$503
$581
$1,012
$2,096
2022
$498
$557
$974
$2,029
2023
$486
$518
$890
$1,893
2024
$464
$901
$1,389
$2,754
2025
$570
$779
$1,240
$2,589
2026
$795
$745
$1,189
$2,730
2027
$775
$1,010
$1,471
$3,256
2028
$837
$978
$1,409
$3,224
2029
$833
$953
$1,378
$3,164
Sum
$6,578
$7,039
$11,196
$24,813
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the less dynamic baseline, la, and more dynamic baseline, lb, please see Preamble Section X.A.I
Table 7-9 Discounted MY Lifetime Vehicle-Related Costs of the Final Program
Vs. The Dynamic Baseline and using Method A
(7% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$162
$0
$81
$243
2019
$299
$0
$68
$367
2020
$264
$12
$59
$335
2021
$416
$484
$828
$1,728
2022
$397
$451
$770
$1,619
2023
$373
$408
$683
$1,464
2024
$342
$668
$1,014
$2,024
2025
$404
$533
$846
$1,783
2026
$543
$490
$782
$1,815
2027
$510
$640
$932
$2,082
2028
$530
$596
$859
$1,984
2029
$507
$559
$808
$1,874
Sum
$4,746
$4,843
$7,732
$17,322
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the less dynamic baseline, la, and more dynamic baseline, lb, please see Preamble Section X.A.I

-------
7.1.2 Changes in Fuel Consumption and Savings
7.1.2.1 Changes in Fuel Consumption
The standards would result in significant improvements in the fuel efficiency of affected
vehicles. Drivers of those vehicles would see corresponding savings associated with reduced
fuel expenditures. The agencies have estimated the impacts on fuel consumption for the
standards. More detail behind these changes in fuel consumption is presented in Chapter 5 and
Chapter 10 of this RIA. The expected impacts on fuel consumption are shown in Table 7-10 as
reductions from the dynamic baseline reference case (i.e., positive values represent fewer gallons
consumed) and using Method A. The gallons shown in this table include any increased
consumption resulting from the rebound effect.
Table 7-10 MY Lifetime Fuel Consumption Reductions due to the Final Program
Vs. The Dynamic Baseline and using Method A
(Million Gallons)a

GASOLINE REDUCTIONS b
DIESEL REDUCTIONS
MODEL
YEAR
HD
PICKUPS
& VANS
voc
TRACTOR/
TRAILERS
SUM
HD
PICKUPS
& VANS
VOC
TRACTOR/
TRAILERS
SUM
2018
162
0
0
162
137
0
302
439
2019
262
0
0
262
217
0
191
408
2020
251
0
0
251
211
0
114
325
2021
432
186
0
618
208
701
3622
4531
2022
451
185
0
636
205
697
3509
4411
2023
464
184
0
648
200
693
3409
4302
2024
463
263
0
726
204
1097
5564
6865
2025
559
265
0
824
247
1108
5524
6879
2026
696
267
0
963
303
1114
5483
6900
2027
724
368
0
1092
319
1481
7384
9184
2028
756
371
0
1127
341
1492
7260
9093
2029
771
374
0
1145
353
1504
7337
9194
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the
less dynamic baseline, la, and more dynamic baseline, lb, please see Preamble Section X. A. 1
b Gasoline reductions include reductions in Ethanol85.
7.1.2.2 Changes in Fuel Expenditures
Using the fuel consumption reductions presented above, NHTSA has calculated the fuel
expenditure changes associated with the standards, subcategory by subcategory. To do this,
reduced fuel consumption is multiplied in each year by the corresponding estimated average fuel
price in that year, using the reference case fuel prices from AEO 2015 final release. As the AEO
fuel price projections go through 2040 and not beyond, fuel prices beyond 2040 were set equal to
the 2040 values. These estimates do not account for the significant uncertainty in future fuel

-------
prices; the monetized fuel savings would be understated if actual fuel prices are higher (or
overstated if fuel prices are lower) than estimated. The Annual Energy Outlook (AEO) is a
standard reference used by NHTSA and EPA and many other government agencies to estimate
the projected price of fuel. This has been done using both the pre-tax and post-tax fuel prices.
Since the post-tax fuel prices are the prices paid at fuel pumps, the fuel expenditure changes
calculated using these prices represent the changes fuel purchasers would see. The pre-tax fuel
savings are those that society would see. Assuming no change in fuel tax rates, the difference
between these two columns represents the reduction in fuel tax revenues that would be received
by state and federal governments. The MY lifetime fuel savings for the final program relative to
the dynamic baseline and using Method A are shown in Table 7-11 using a 3 percent discount
rate and in Table 7-12 using a 7 percent discount rate. Note that in Chapters 8 and 11 of this
RIA, the overall benefits and costs of the rulemaking are presented and only the pre-tax fuel
expenditure impacts are presented there.
Table 7-11 Discounted MY Lifetime Reductions in Fuel Expenditures of the Final Program
Vs. The Dynamic Baseline and using Method A
(3% Discount Rate, Billions of 2013$)a

REDUCED FUEL EXPENDITURES -
RETAIL
REDUCED FUEL EXPENDITURES -
UNTAXED
MODEL
YEAR
HD PICKUPS
& VANS
voc
TRACTOR/
TRAILERS
SUM
HD PICKUPS
& VANS
VOC
TRACTOR/
TRAILERS
SUM
2018
0.9
0.0
0.8
1.7
0.8
0.0
0.7
1.5
2019
1.4
0.0
0.5
1.9
1.2
0.0
0.4
1.6
2020
1.3
0.0
0.3
1.6
1.2
0.0
0.3
1.5
2021
1.8
2.2
9.1
13.1
1.6
1.9
8.0
11.5
2022
1.8
2.1
8.7
12.6
1.6
1.9
7.7
11.2
2023
1.8
2.1
8.4
12.3
1.6
1.9
7.5
11.0
2024
1.8
3.2
13.5
18.5
1.6
2.9
12.1
16.6
2025
2.1
3.2
13.3
18.6
1.9
2.9
11.9
16.7
2026
2.6
3.2
13.0
18.8
2.3
2.8
11.7
16.8
2027
2.7
4.2
17.3
24.2
2.4
3.8
15.6
21.8
2028
2.8
4.2
16.8
23.8
2.5
3.8
15.2
21.5
2029
2.8
4.2
16.8
23.8
2.5
3.8
15.2
21.5
Sum
23.9
28.5
118.6
171.0
21.2
25.5
106.2
152.9
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the less
dynamic baseline, la, and more dynamic baseline, lb, please see Preamble Section X.A.I

-------
Table 7-12 Discounted MY Lifetime Reductions in Fuel Expenditures of the Final Program
Vs. The Dynamic Baseline and using Method A
(7% Discount Rate, Billions of 2013$)a

REDUCED FUEL EXPENDITURES -
REDUCED FUEL EXPENDITURES -


RETAIL


UNTAXED

MODEL
HD PICKUPS
\/np
TRACTOR/

HD PICKUPS
\/np
TRACTOR/

YEAR
& VANS
VUL
TRAILERS
oUlvl
& VANS
VUL
TRAILERS
oUlvl
2018
0.8
0.0
0.6
1.4
0.7
0.0
0.5
1.2
2019
1.3
0.0
0.4
1.7
1.1
0.0
0.3
1.4
2020
1.2
0.0
0.2
1.4
1.0
0.0
0.2
1.2
2021
1.5
1.4
5.7
8.6
1.3
1.2
5.0
7.5
2022
1.4
1.3
5.2
7.9
1.3
1.1
4.6
7.0
2023
1.4
1.2
4.9
7.5
1.2
1.1
4.3
6.6
2024
1.3
1.8
7.5
10.6
1.2
1.6
6.7
9.5
2025
1.5
1.7
7.1
10.3
1.3
1.5
6.4
9.2
2026
1.8
1.7
6.8
10.3
1.6
1.5
6.0
9.1
2027
1.8
2.1
8.7
12.6
1.6
1.9
7.8
11.3
2028
1.8
2.0
8.1
11.9
1.6
1.8
7.3
10.7
2029
1.7
1.9
7.8
11.4
1.5
1.7
7.0
10.2
Sum
17.4
15.1
62.9
95.4
15.4
13.4
56.1
84.9
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the less
dynamic baseline, la, and more dynamic baseline, lb, please see Preamble Section X.A.I
7.1.3 Maintenance Costs
The agencies have estimated increased maintenance costs associated with installation of
lower rolling resistance tires. We expect that, when replaced, the lower rolling resistance tires
would be replaced by equivalent performing tires throughout the vehicle lifetime. As such, the
incremental increases in costs for lower rolling resistance tires would be incurred throughout the
vehicle lifetime at intervals consistent with current tire replacement intervals. Those intervals
are difficult to quantify given the variety of vehicles and operating modes within the HD
industry. For HD pickups and vans, we have chosen a tire replacement interval of 40,000 miles.
We have done the same for all vocational vehicles which is probably overly conservative as
more frequent intervals results in higher maintenance costs. For tractors and trailers, we have
used a maintenance interval of 200,000 miles. The presence of tire inflation management
systems, and the increased use of those systems expected due to this final rule, should serve to
improve tire maintenance intervals.
In evaluating maintenance costs associated with the rule relative to Alternative lb,
NHTSA has used, for HD pickups and vans, the integrated analysis performed using the CAFE
modeling system, which includes additional miles from an estimated rebound effect of 10
percent (the rebound effect is the demand response of VMT when the cost-per-mile travel
becomes less expensive). For vocational vehicles, tractors and trailers, NHTSA has used the
MOVES-based approach outlined above. The results of NHTSA's analysis are reported as
"Method A."
Table 7-13 presents the model year lifetime in-use maintenance costs—versus the
dynamic baseline and using Method A— discounted at 3 percent. Table 7-14 presents the model

-------
year lifetime in-use maintenance costs—versus the dynamic baseline and using Method A—
discounted at 7 percent.
Table 7-13 Discounted MY Lifetime Maintenance Costs of the Final Program
Vs. The Dynamic Baseline and using Method A
(3% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
0
6.7
6.7
2019
0
6.5
6.5
2020
0
6.5
6.5
2021
19.0
130.6
149.6
2022
17.8
126.2
144
2023
20.3
121.9
142.2
2024
53.6
97.6
151.2
2025
52.3
95.1
147.4
2026
42.5
93.4
135.9
2027
89.9
186.8
276.7
2028
86.1
181.4
267.5
2029
83.7
176.7
260.4
Sum
465.1
1229.5
1694.6
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the less dynamic baseline, la, and more dynamic baseline, lb, please see Preamble Section X.A.I
Table 7-14 Discounted MY Lifetime Maintenance Costs of the Final Program
Vs. The Dynamic Baseline and using Method A
(7% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
0.0
4.7
4.7
2019
0.0
4.4
4.4
2020
0.0
4.2
4.2
2021
12.3
82.9
95.2
2022
11.1
77.1
88.2
2023
12.1
71.8
83.9
2024
30.8
55.6
86.4
2025
28.9
52.1
81
2026
22.7
49.3
72
2027
46.2
94.9
141.1
2028
42.6
88.8
131.4
2029
39.9
83.3
123.2
Sum
246.7
669.1
915.8
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the less dynamic baseline, la, and more dynamic baseline, lb, please see Preamble Section X.A.I

-------
7.1.4 Analysis of Payback Periods
An important metric to vehicle purchasers is the payback period that can be expected on
any new purchase. In other words, there is greater willingness to pay for new technology if that
new technology "pays back" within an acceptable period of time. We make no effort to define
the acceptable period of time here, but seek to estimate the payback period for others to make the
decision themselves. We define the payback period as the point at which reduced fuel
expenditures outpace increased vehicle costs. For example, a new MY2027 HD pickup truck is
estimated to cost roughly $1,300 more (on average, in 2013$, and relative to the reference case
vehicle) due to the addition of new fuel consumption improving and GHG reducing technology.
This new technology would result in lower fuel consumption and, therefore, reduced fuel
expenditures. But how many months or years would pass before the reduced fuel expenditures
would surpass the increased costs?
To estimate the costs, we have considered not only the cost of the new technology, but
also the taxes paid on the incrementally higher purchase expense, the slightly higher insurance
expenses on the slightly higher value vehicle, the increased finance cost, and the increased
maintenance costs associated with the new technology. Taxes and fees paid were estimated as
5.46 percent of the final MSRP. Financing was estimated to be 15.32 percent of final MSRP,
and for insurance costs, the model uses an estimate of 19.23 percent of the final MSRP of a
vehicle as the cost of insurance. These calculations do not represent specific vehicle classes or
specific use cases so should not be seen as being applicable to any particular individual's
situation. However, the payback periods do provide a general sense, on average, of what sort of
payback periods are likely at a national, societal perspective.
Table 7-15 presents the discounted annual increased vehicle costs and fuel expenditure
impacts associated with owning a new MY2027 HD pickup or van using both 3 percent and 7
percent discount rates. The results in this table use Method A. As shown in the table, the
payback for HD pickups and vans occurs late in the 3rd year of ownership (the year in which
cumulative expenditures become negative) using a 3 percent discount rate and in the early part of
the 4th year using a 7 percent discount rate. For other classes of vehicles, including vehicle
types such as refuse trucks and transit buses, refer to the Method B analysis of payback periods
presented in Chapter 7.2.4.

-------
Table 7-15 Discounted Owner Expenditures & Payback Period for MY2027 HD Pickups & Vans under the
Final Program Vs. The Dynamic Baseline and using Method A
3% and 7% Discount Rates (2013$)a

3% Discount Rate
7% Discount Rate
Age
Technology cost,
Fuel
Cumulative
Technology cost,
Fuel
Cumulative

taxes, insurance b
expenditures 0
expenditures
taxes, insurance b
expenditures0
expenditures
1
1296
-554
742
1248
-534
714
2
0
-494
248
0
-457
257
3
0
-424
-176
0
-378
-121
4
0
-357
-533
0
-306
-427
5
0
-284
-817
0
-235
-662
6
0
-214
-1031
0
-170
-832
7
0
-208
-1239
0
-160
-992
8
0
-175
-1414
0
-129
-1121
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the less
dynamic baseline, la, and more dynamic baseline, lb, please see Preamble Section X.A.I
b 6% sales tax; insurance estimates are described in text.
0 Fuel expenditures calculated using retail fuel prices according to AEO2015 final release, reference case estimates.
7.2 Vehicle Costs, Fuel Savings and Maintenance Costs vs. the Flat
Baseline and using Method B
As noted in the introduction to Chapter 7.1, the Method B analysis of the potential costs
of the standards combines EPA MOVES modeling of vocational vehicle, tractor and trailer fuel
consumption, EPA analysis of vocational vehicle, tractor and trailer costs, along with DOT
CAFE model calculations of HD pickup and van costs per vehicle. The analysis includes costs
for fuel-saving technology that manufacturers could add in response to the standards, EPA
estimates of the additional compliance and R&D costs for vocational vehicles and combination
tractor trailers, and some additional maintenance costs.3
7.2.1 Vehicle Program Costs
This section presents the Method B estimate of the vehicle-related costs associated with
the final program versus the flat baseline (Alternative la) using the MOVES analysis of HD
pickups and vans as well as vocational vehicle, tractors and trailers. The presentation here
summarizes the costs associated with new technology the agencies estimate manufacturers could
add to meet the GHG and fuel consumption standards. The analysis summarized here provides
our estimate of incremental costs on a per vehicle basis, on a MY lifetime basis and on an annual
basis. For details behind the cost estimates associated with individual technologies, the reader is
directed to Sections III through VI of the Preamble and to Chapter 2 of the RIA. The analysis
here also includes a look at payback periods—the time at which cumulative fuel savings
outweigh increased costs associated with new, more fuel efficient vehicles. And finally, the
analysis here includes a look at the cost per ton of GHG emissions reduced by the addition of
new technology.

-------
Note that all discounted costs presented in this chapter, whether in the Calendar Year (or
annual) analysis or the Model Year Lifetime analysis, are discounted back to 2015 at the
discount rate shown in the table(s).
7.2.1.1 Technology Costs
For the HD pickups and vans, Method B used the same technology costs as used in the
proposal, except that those costs have been updated to 2013 dollars using a factor of 1.016
applied to the 2012$-based NPRM costs. As in the proposal, we have used a methodology
consistent with that used for our recent 2017-2025 light-duty joint rulemaking since most of the
technologies expected for HD pickups and vans are consistent with those expected for the larger
light-duty trucks. The cost estimates presented in the recent light-duty joint rulemaking were
then scaled upward to account for the larger weight, towing capacity, and work demands of the
trucks in these heavier classes. For details on that scaling process and the resultant costs for
individual technologies, the reader is directed to Chapter 2.12 of this RIA.
For vocational vehicles, tractors and trailers, consistent with the Phase 1 rule, the
agencies have estimated costs using a different methodology than that employed in the recent
light-duty joint rulemaking establishing fuel economy and GHG standards. In the recent light-
duty joint rulemaking, all fixed costs were included in the hardware costs via an indirect cost
multiplier. As such, the hardware costs presented in that analysis included both the actual
hardware and the associated fixed costs. For the vocational, tractor and trailer segments in this
analysis, some of the fixed costs are estimated separately and are presented separately from the
technology costs. As noted above, all costs are presented in 2013 dollars.
The estimates of vehicle costs are generated relative to two unique "no action" baselines.
The first of these (alternative la, presented here) representing generally flat or flat fuel
consumption improvements, or a fleet of vehicles meeting the Phase 1 heavy-duty requirements.
The second of these (alternative lb and presented in detail in Chapter 7.1) representing dynamic
fuel consumption improvements, or a fleet of vehicles with improving fuel consumption despite
the lack of regulatory drivers. See Section X of the Preamble and Chapter 11 of this RIA for
more detail on these two baselines. As such, costs to comply with the Phase 1 standards are not
included in the estimates here. In fact, in the methodology used for vocational vehicles, tractors
and trailers, there are cases where Phase 1 technologies are being removed in favor of Phase 2
technologies - that is, the technology basis for the Phase 2 standards involves removing certain
of the Phase 1 technologies. In those cases, savings are associated with the removal of the Phase
1 technology. The details of which technologies and where such savings occur are presented in
Chapter 2.12 of the RIA.
For HD pickups and vans, as described in Chapter 2 of this RIA, Method B uses
NHTSA's CAFE model to estimate the cost per vehicle associated with the preferred and
possible alternative standards.6 That model has the capability to look ahead at future standards
B The CAFE model also provides a full benefit-cost analysis associated with the HD pickup and van portion of the
standards. The full benefit-cost analysis for Method A is presented in Chapters 9 and 10 of this RIA. The full
benefit-cost analysis for Method B is presented in Chapter 8 of this RIA.

-------
when making determinations of how vehicles should be changed to comply. It does this because
redesign cycles do not always line up well with regulatory implementation schedules, so a
manufacturer may choose to redesign a vehicle in MY2018 in preparation for upcoming
MY2021 and later standards if that particular vehicle is not scheduled for another redesign until
after the timeframe covered by the upcoming standards. The result being new technology costs
in years prior to implementation of the standards. The CAFE model's output would show such
costs occurring in years prior to MY2021. On the other hand, the CAFE model also estimates
the potential that credits generated in earlier model years might be carried forward (i.e.,
"banked") and then used in later model years, potentially reducing costs in some model years
covered by the analysis.
Table 7-16 presents the average incremental technology costs per vehicle for the final
program relative to the flat baseline and using Method B (the MOVES analysis for all vehicle
categories). These tables include both engine and vehicle technologies. For HD pickups and
vans, costs begin before the MY2021 implementation as the CAFE model projects that
manufacturers will start adding technology in anticipation of the standards. For vocational
vehicles, costs begin in MY2021, then decrease slightly due to learning effects, then increase
again in MY2024 and 2027 as the more stringent standards take effect. The story is similar for
tractor-trailers where costs begin in MY2018 on trailers then follow a pattern similar to
vocational vehicles as the MY2021, 2024 and 2027 standards take effect on tractors. Costs then
decrease beyond MY2027 for each category due to learning effects. All costs shown in the table
represent the weighted average cost of all vehicles within the category shown in the heading.

-------
Table 7-16 Estimated Technology Costs per Vehicle for the Final Program versus the Flat Baseline and using
Method B (2013$) a
MODEL
HD PICKUPS

TRACTOR/
YEAR
& VANS

TRAILERS
2018
$114
$0
$568
2019
$105
$0
$548
2020
$108
$0
$535
2021
$524
$1,110
$7,352
2022
$516
$1,088
$7,269
2023
$804
$1,027
$6,799
2024
$963
$2,022
$11,134
2025
$1,180
$1,986
$10,901
2026
$1,244
$1,927
$10,712
2027
$1,364
$2,662
$13,550
2028
$1,354
$2,616
$13,229
2029
$1,355
$2,586
$13,089
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section
I.D; for an explanation of the flat baseline, la, and dynamic baseline, lb, please see
Preamble Section X.A.I
As noted in the text, MYs 2018-2020 include costs for trailers only, and in MYs
2021 and later the costs include both tractor and trailer costs, inclusive of
engine-related costs. Detailed technology and package costs for all segments can
be found in Chapter 2 of this RIA (notably, see Sections 2.12 and 2.13). Also,
for HD pickups and vans, EPA has taken early costs and spread them over the
years 2021 through 2026 so that those costs can be fully realized while showing
them occurring during the expected years of implementation.
Table 7-17 presents the annual costs—versus the flat baseline and using Method B—for
new engine- and vehicle-related technology along with net present values at 3 percent and 7
percent. Table 7-18 presents the model year lifetime costs—versus the flat baseline and using
Method B—for new technology discounted at 3 percent. Table 7-19 presents the model year
lifetime costs—versus the flat baseline and using Method B—for new technology discounted at 7
percent.

-------
Table 7-17 Annual Technology Costs and Net Present Values Associated with the Final Program vs. the Flat
Baseline and using Method B (SMillions of 2013$)a
CALENDAR
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$137
$0
$90
$227
2019
$126
$0
$89
$215
2020
$129
$0
$91
$220
2021
$621
$563
$1,087
$2,270
2022
$607
$553
$1,082
$2,243
2023
$944
$525
$1,016
$2,485
2024
$1,140
$1,045
$1,706
$3,890
2025
$1,406
$1,046
$1,695
$4,146
2026
$1,494
$1,030
$1,679
$4,203
2027
$1,639
$1,439
$2,141
$5,219
2028
$1,628
$1,435
$2,113
$5,176
2029
$1,627
$1,440
$2,128
$5,195
2030
$1,610
$1,449
$2,159
$5,219
2035
$1,625
$1,585
$2,432
$5,642
2040
$1,671
$1,776
$2,798
$6,245
2050
$1,755
$2,145
$3,369
$7,270
NPV, 3%
$25,007
$23,932
$37,841
$86,780
NPV, 7%
$12,239
$11,120
$17,789
$41,148
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1

-------
Table 7-18 Discounted MY Lifetime New Technology Costs of the Final Program
Vs. the Flat Baseline and using Method B (3% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$124
$0
$81
$205
2019
$110
$0
$78
$188
2020
$110
$0
$77
$187
2021
$512
$465
$897
$1,873
2022
$487
$443
$867
$1,797
2023
$735
$408
$790
$1,933
2024
$861
$789
$1,288
$2,938
2025
$1,031
$767
$1,242
$3,040
2026
$1,064
$733
$1,195
$2,992
2027
$1,133
$995
$1,479
$3,607
2028
$1,092
$963
$1,418
$3,473
2029
$1,060
$938
$1,386
$3,384
Sum
$8,316
$6,500
$10,800
$25,617
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1
Table 7-19 Discounted MY Lifetime New Technology Costs of the Final Program
Vs. the Flat Baseline and using Method B (7% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$108
$0
$71
$179
2019
$93
$0
$66
$159
2020
$89
$0
$63
$152
2021
$400
$363
$700
$1,462
2022
$366
$333
$652
$1,350
2023
$531
$295
$572
$1,398
2024
$599
$549
$897
$2,046
2025
$691
$514
$833
$2,038
2026
$686
$473
$771
$1,930
2027
$704
$618
$919
$2,240
2028
$653
$576
$848
$2,076
2029
$610
$540
$798
$1,948
Sum
$5,530
$4,260
$7,188
$16,978
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of
the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
7.2.1.2 Compliance Costs
As noted above, some fixed costs were estimated separately from the hardware costs. As
such, not all fixed costs are included in the tables presented in Chapter 7.2.1.1. The agencies
have estimated additional and/or new compliance costs associated with the standards. Normally,
compliance program costs would be considered part of the indirect costs and, therefore, would be
accounted for via the markup applied to direct manufacturing costs. However, since the agencies
are requiring new compliance elements that were not present during development of the indirect
cost markups used in this analysis, additional compliance program costs are being accounted for
via a separate "line-item" here. Note that, for HD pickups and vans, compliance elements were
present during development of the indirect cost markups used; as such, these costs are already
included as part of the technology costs described above.
There are three elements to the compliance costs estimated in this analysis. The first is
for construction of new, or upgrades to existing, test facilities for conducting powertrain testing.
The second costs are for conducting the powertrain tests themselves. And the third is for
reporting of compliance data to EPA and NHTSA. We estimated these latter costs in the Phase 1
rule as $0.24 million, $0.9 million and $1.1 million for HD pickups and vans, vocational and
tractors, respectively, for a total of $2.3 million per year (2009$).4 All of these are industry-
wide, annual costs.
We have estimated reporting costs in this Phase 2 final rule associated with new
powertrain testing within the vocational vehicle program, the increased level of reporting in the
tractor program and an all new compliance program where none has existed to date within the
trailer program. We have estimated those costs, inclusive of the Phase 1 costs, such that the new
GHG program reporting costs are estimated at $1.1 million and $1.2 million for vocational and
tractor programs both in 2013$. All of these are industry-wide, annual costs.
For powertrain testing facility upgrades and construction, we have estimated that 6
manufacturers would upgrade and 5 would construct new facilities at an upgrade cost of $1.2
million and a new construction cost of $1.9 million, all in 2013$. The result being an industry-
wide (but vocational program only) cost of $16.8 million (2013$). This cost would occur once
which we have attributed to CY2020, one year prior to the first year of the Phase 2 vocational
standards.
Lastly, the vocational program is also estimated to incur costs associated with conducting
powertrain testing. We have estimated the cost of testing at $40,000 per test (2013$) and expect
10 tests/year for a total of $400,000/year. We have also estimated that the vocational program
will incur costs associated with conducting transmission efficiency testing at a cost of $24,600
per test (2013$). We have estimated 11 tests per year for a total annual cost of $270,600. We
have also estimated that the vocational program will incur costs associated with conducting axle
efficiency testing at a cost of $12,600 per test (2013$). We have estimated that 9 tests would be
done per year for a total annual cost of $113,400. We have also estimated an annual cost of
$8,700 in tire testing will be incurring by the vocational program.
In the tractor program, we have used the same per test costs noted above for vocational
and have estimated one transmission efficiency test per year for a total annual cost of $24,600

-------
(2013$) and 15 axle efficiency tests per year for a total annual cost of $189,000 (2013$). To
those costs, we have also added $300,000 (2013$) per year in aero-related testing and $5,400
(2013$) per year in tire testing.
For the trailer program, we have estimated an annual compliance program cost of $7
million (2013$) to cover reporting, testing and capital costs.
Table 7-20 through Table 7-22 present the annual costs for new compliance program
elements along with net present values at 3 percent and 7 percent, and the model year lifetime
compliance costs discounted at 3 percent and 7 percent, respectively.
Table 7-20 Annual Compliance Costs and Net Present Values Associated with the Final Program
Vs. The Flat Baseline and using Method B (SMillions of 2013$)a
CALENDAR
HD PICKUPS
VOCATIONAL
TRACTOR/
SUM
YEAR
& VANS
TRAILERS
2018
$0
$0
$0
$0
2019
$0
$0
$0
$0
2020
$0
$17
$0
$17
2021
$0
$1.9
$8.7
$11
2022
$0
$1.9
$8.7
$11
2023
$0
$1.9
$8.7
$11
2024
$0
$1.9
$8.7
$11
2025
$0
$1.9
$8.7
$11
2026
$0
$1.9
$8.7
$11
2027
$0
$1.9
$8.7
$11
2028
$0
$1.9
$8.7
$11
2029
$0
$1.9
$8.7
$11
2030
$0
$1.9
$8.7
$11
2035
$0
$1.9
$8.7
$11
2040
$0
$1.9
$8.7
$11
2050
$0
$1.9
$8.7
$11
NPV, 3%
$0
$45
$145
$191
NPV, 7%
$0
$27
$75
$102
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1

-------
Table 7-21 Discounted MY Lifetime Compliance Costs of the Final Program
Vs. The Flat Baseline and using Method B (3% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$0
$0
$0
$0
2019
$0
$0
$0
$0
2020
$0
$14
$0
$14
2021
$0
$1.5
$7.2
$8.7
2022
$0
$1.5
$7.0
$8.5
2023
$0
$1.4
$6.8
$8.2
2024
$0
$1.4
$6.6
$8.0
2025
$0
$1.4
$6.4
$7.8
2026
$0
$1.3
$6.2
$7.5
2027
$0
$1.3
$6.0
$7.3
2028
$0
$1.2
$5.9
$7.1
2029
$0
$1.2
$5.7
$6.9
Sum
$0
$27
$58
$84
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of
the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 7-22 Discounted MY Lifetime Compliance Costs of the Final Program
Vs. The Flat Baseline and using Method B (7% Discount Rate, SMillions of 2013$)a
MODEL
HD PICKUPS
VOCATIONAL
TRACTOR/
SUM
YEAR
& VANS
TRAILERS
2018
$0
$0
$0
$0
2019
$0
$0
$0
$0
2020
$0
$12
$0
$12
2021
$0
$1.2
$5.6
$6.8
2022
$0
$1.1
$5.3
$6.4
2023
$0
$1.0
$4.9
$6.0
2024
$0
$1.0
$4.6
$5.6
2025
$0
$0.9
$4.3
$5.2
2026
$0
$0.9
$4.0
$4.9
2027
$0
$0.8
$3.7
$4.5
2028
$0
$0.7
$3.5
$4.2
2029
$0
$0.7
$3.3
$4.0
Sum
$0
$20
$39
$59
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1

-------
7.2.1.3 Research & Development Costs
Much like the compliance program costs described above, EPA has estimated additional
engine, vocational vehicle and tractor R&D associated with the standards that is not accounted
for via the indirect cost markups used in this analysis for those segments. The necessary R&D
for HD pickups and vans is covered by the indirect costs included as part of the technology costs
described above. In the Phase 1 rule, the agencies estimated the engine R&D costs at $6.8
million (2009$) per engine class per manufacturer per year for five years. In this Phase 2
analysis, EPA has estimated this same level of R&D and has assumed 12 heavy-heavy and 12
medium-heavy HD engine R&D programs would be conducted for a total of $218 million/year
(2013$). In this analysis, EPA has assumed those costs would occur annually for 4 years, MYs
2021-2024. The total being $873 million (2013$) over 4 years (by comparison, the Phase 1 rule
estimated a total of $852 million (2009$) over 5 years). To this, EPA has estimated an additional
$20 million/year spent by vocational vehicle manufacturers and $20 million/year spent by tractor
manufacturers. In the end, EPA is estimating a total of over $1 billion in R&D spending above
and beyond the level included in the markups used to estimate indirect costs for these segments.
EPA has not included any additional R&D would be spent by trailer manufacturers since our
trailer technology cost estimates include R&D conducted by trailer parts suppliers which are
subsequently included in the prices charged by those suppliers to the trailer manufacturer.
Additionally, the markups we have applied to cover indirect costs (see Chapter 2.12 of this RIA)
include costs associated with R&D incurred by the trailer manufacturer.
Table 7-23 through Table 7-25 present the annual costs for R&D spending along with net
present values at 3 percent and 7 percent, and the model year lifetime R&D costs discounted at 3
percent and 7 percent, respectively.

-------
Table 7-23 Additional Annual R&D Costs, Not Covered by Indirect Cost Markups), and Net Present Values
Associated with the Final Program Vs. The Flat Baseline and using Method B
(SMillions of 2013$)a
CALENDAR
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$0
$0
$0
$0
2019
$0
$0
$0
$0
2020
$0
$0
$0
$0
2021
$0
$129
$129
$259
2022
$0
$129
$129
$259
2023
$0
$129
$129
$259
2024
$0
$129
$129
$259
2025
$0
$0
$0
$0
2026
$0
$0
$0
$0
2027
$0
$0
$0
$0
2028
$0
$0
$0
$0
2029
$0
$0
$0
$0
2030
$0
$0
$0
$0
2035
$0
$0
$0
$0
2040
$0
$0
$0
$0
2050
$0
$0
$0
$0
NPV, 3%
$0
$409
$409
$818
NPV, 7%
$0
$302
$302
$604
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1

-------
Table 7-24 Discounted MY Lifetime R&D Costs, Not Covered by Indirect Cost Markups), of the Final
Program Vs. The Flat Baseline and using Method B (3% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$0
$0
$0
$0
2019
$0
$0
$0
$0
2020
$0
$0
$0
$0
2021
$0
$107
$107
$214
2022
$0
$104
$104
$207
2023
$0
$101
$101
$201
2024
$0
$98
$98
$195
2025
$0
$0
$0
$0
2026
$0
$0
$0
$0
2027
$0
$0
$0
$0
2028
$0
$0
$0
$0
2029
$0
$0
$0
$0
Sum
$0
$409
$409
$818
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1
Table 7-25 Discounted MY Lifetime R&D Costs, Not Covered by Indirect Cost Markups), of the Final
Program Vs. The Flat Baseline and using Method B (7% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$0
$0
$0
$0
2019
$0
$0
$0
$0
2020
$0
$0
$0
$0
2021
$0
$83
$83
$167
2022
$0
$78
$78
$156
2023
$0
$73
$73
$146
2024
$0
$68
$68
$136
2025
$0
$0
$0
$0
2026
$0
$0
$0
$0
2027
$0
$0
$0
$0
2028
$0
$0
$0
$0
2029
$0
$0
$0
$0
Sum
$0
$302
$302
$604
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1

-------
7.2.1.4 Summary of Vehicle-Related Costs of the Program using Method B
Table 7-26 presents the annual new vehicle costs (including engine-related costs)
associated with the final program for HD pickups and vans, vocational vehicles, and tractor and
trailer programs along with net present values at 3 percent and 7 percent. This table presents
costs relative to the flat baseline and using the MOVES analysis of all vehicle categories
(Method B). Table 7-27 presents the model year lifetime costs associated with the final program
discounted at 3 percent relative to the flat baseline and using Method B. Table 7-28 presents the
model year lifetime costs associated with the final program discounted at 7 percent relative to the
flat baseline and using Method B.
Table 7-26 Annual Vehicle-Related Costs and Net Present Values Associated with the Final Program
Vs. The Flat Baseline and using Method B (SMillions of 2013$)a
CALENDAR
HD PICKUPS
VOCATIONAL
TRACTOR/
SUM
YEAR
& VANS
TRAILERS
2018
$137
$0
$90
$227
2019
$126
$0
$89
$215
2020
$129
$17
$91
$237
2021
$621
$694
$1,225
$2,540
2022
$607
$685
$1,220
$2,512
2023
$944
$656
$1,154
$2,755
2024
$1,140
$1,176
$1,844
$4,160
2025
$1,406
$1,048
$1,703
$4,157
2026
$1,494
$1,032
$1,687
$4,213
2027
$1,639
$1,441
$2,149
$5,230
2028
$1,628
$1,437
$2,122
$5,186
2029
$1,627
$1,442
$2,137
$5,206
2030
$1,610
$1,451
$2,168
$5,229
2035
$1,625
$1,587
$2,441
$5,653
2040
$1,671
$1,778
$2,807
$6,255
2050
$1,755
$2,147
$3,378
$7,280
NPV, 3%
$25,007
$24,386
$38,395
$87,788
NPV, 7%
$12,239
$11,449
$18,166
$41,854
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1

-------
Table 7-27 Discounted MY Lifetime Vehicle-Related Costs of the Final Program
Vs. The Flat Baseline and using Method B (3% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$124
$0
$81
$205
2019
$110
$0
$78
$188
2020
$110
$14
$77
$201
2021
$512
$573
$1,011
$2,096
2022
$487
$549
$978
$2,013
2023
$735
$510
$898
$2,143
2024
$861
$888
$1,393
$3,141
2025
$1,031
$768
$1,249
$3,048
2026
$1,064
$734
$1,201
$2,999
2027
$1,133
$996
$1,485
$3,614
2028
$1,092
$964
$1,424
$3,480
2029
$1,060
$939
$1,392
$3,391
Sum
$8,316
$6,935
$11,267
$26,519
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1
Table 7-28 Discounted MY Lifetime Vehicle-Related Costs of the Final Program
Vs. The Flat Baseline and using Method B (7% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$108
$0
$71
$179
2019
$93
$0
$66
$159
2020
$89
$12
$63
$163
2021
$400
$447
$789
$1,636
2022
$366
$412
$735
$1,513
2023
$531
$369
$649
$1,550
2024
$599
$618
$970
$2,187
2025
$691
$515
$837
$2,043
2026
$686
$474
$775
$1,935
2027
$704
$619
$923
$2,245
2028
$653
$576
$851
$2,080
2029
$610
$541
$801
$1,952
Sum
$5,530
$4,583
$7,530
$17,642
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1

-------
7.2.2 Changes in Fuel Consumption and Savings
7.2.2.1 Changes in Fuel Consumption
The standards will result in significant improvements in the fuel efficiency of affected
vehicles. Drivers of those vehicles will see corresponding savings associated with reduced fuel
expenditures. The agencies have estimated the impacts on fuel consumption for the standards.
More detail behind these changes in fuel consumption is presented in Chapter 5 of this RIA. The
expected impacts on fuel consumption are shown in Table 7-29 as reductions from the flat
baseline reference case (i.e., positive values represent fewer gallons consumed) and using the
MOVES analysis of all vehicle categories (Method B). The gallons shown in this table include
any increased consumption resulting from the rebound effect.
Table 7-29 Annual Fuel Consumption Reductions due to the Final Program
Vs. The Flat Baseline and using Method B (Million Gallons)a

GASOLINE REDUCTIONS
DIESEL REDUCTIONS
CALENDAR
YEAR
HD
PICKUPS
& VANS
voc
TRACTOR/
TRAILERS
SUM
HD
PICKUPS
& VANS
VOC
TRACTOR/
TRAILERS
SUM
2018
0
0
0
0
0
0
37
37
2019
0
0
0
0
0
0
76
76
2020
0
0
0
0
0
0
117
117
2021
11
17
0
28
8
57
363
428
2022
41
33
0
74
29
113
670
812
2023
89
50
0
138
62
169
980
1,211
2024
153
73
0
226
107
258
1,470
1,835
2025
235
95
0
330
164
344
1,949
2,457
2026
331
116
0
448
232
426
2,405
3,063
2027
442
146
0
588
310
536
3,007
3,853
2028
549
174
0
723
385
641
3,584
4,610
2029
651
201
0
852
457
742
4,136
5,335
2030
748
226
0
974
525
839
4,667
6,031
2035
1,131
323
0
1,454
792
1,223
6,867
8,883
2040
1,346
377
0
1,724
940
1,464
8,374
10,778
2050
1,476
428
0
1,904
1,059
1,729
10,198
12,986
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
Table 7-30 MY Lifetime Fuel Consumption Reductions due to the Final Program
Vs. The Flat Baseline and using Method B (Million Gallons)a

GASOLINE REDUCTIONS
DIESEL REDUCTIONS
MODEL
YEAR
HD
PICKUPS
& VANS
voc
TRACTOR/
TRAILERS
SUM
HD
PICKUPS
& VANS
VOC
TRACTOR/
TRAILERS
SUM
2018
0
0
0
0
0
0
302
302
2019
0
0
0
0
0
0
293
293
2020
0
0
0
0
0
0
286
286
2021
136
186
0
322
91
701
3,852
4,643
2022
365
185
0
550
243
697
3,867
4,807
2023
588
184
0
772
391
693
3,862
4,947
2024
813
263
0
1,075
542
1,097
6,104
7,742
2025
1,036
265
0
1,301
691
1,108
6,154
7,954
2026
1,258
267
0
1,525
838
1,114
6,159
8,111
2027
1,467
368
0
1,836
980
1,481
8,184
10,646
2028
1,469
371
0
1,840
984
1,492
8,222
10,698
2029
1,468
374
0
1,841
987
1,504
8,309
10,800
Sum
8,598
2,464
0
11,062
5,748
9,887
55,593
71,229
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and more dynamic baseline, lb, please see Preamble Section X. A. 1
7.2.2.2 Changes in Fuel Expenditures
Using the fuel consumption reductions presented above, the agencies have calculated the
fuel expenditure changes associated with the standards, subcategory by subcategory. To do this,
reduced fuel consumption is multiplied in each year by the corresponding estimated average fuel
price in that year, using the reference case fuel prices from AEO 2015. As the AEO fuel price
projections go through 2040 and not beyond, fuel prices beyond 2040 were set equal to the 2040
values. These estimates do not account for the significant uncertainty in future fuel prices; the
monetized fuel savings would be understated if actual fuel prices are higher (or overstated if fuel
prices are lower) than estimated. The Annual Energy Outlook (AEO) is a standard reference
used by NHTSA and EPA and many other government agencies to estimate the projected price
of fuel. This has been done using both the pre-tax and post-tax fuel prices. Since the post-tax
fuel prices are the prices paid at fuel pumps, the fuel expenditure changes calculated using these
prices represent the changes fuel purchasers would see. The pre-tax fuel savings are those that
society would see. Assuming no change in fuel tax rates, the difference between these two
columns represents the reduction in fuel tax revenues that would be received by state and federal
governments, or about $200 million in 2021 and $5.8 billion by 2050 as shown in Table 7-31.
Table 7-32 presents the model year lifetime fuel savings—versus the flat baseline and using
Method B—discounted at 3 percent. Table 7-33 presents the model year lifetime costs fuel
savings—versus the flat baseline and using Method B—discounted at 7 percent. Note that in

-------
Chapters 8 and 11 of this RIA, the overall benefits and costs of the rulemaking are presented and
only the pre-tax fuel expenditure impacts are presented there.
Table 7-31 Annual Reductions in Fuel Expenditures and Net Present Values due to the Final Program Vs.
The Flat Baseline and using Method B (Millions of 2013$)a

REDUCED FUEL EXPENDITURES -
RETAIL
REDUCED FUEL EXPENDITURES -
UNTAXED
CALENDAR
YEAR
HD
PICKUPS
& VANS
voc
TRACTOR/
TRAILERS
SUM
HD
PICKUPS
& VANS
VOC
TRACTOR/
TRAILERS
SUM
2018
$0
$0
$114
$114
$0
$0
$97
$97
2019
$0
$0
$237
$237
$0
$0
$202
$202
2020
$0
$0
$371
$371
$0
$0
$319
$319
2021
$56
$232
$1,174
$1,462
$48
$199
$1,010
$1,258
2022
$210
$470
$2,219
$2,899
$181
$406
$1,917
$2,504
2023
$461
$713
$3,302
$4,476
$399
$619
$2,871
$3,889
2024
$812
$1,097
$5,043
$6,952
$707
$956
$4,396
$6,059
2025
$1,265
$1,482
$6,803
$9,550
$1,104
$1,295
$5,945
$8,343
2026
$1,819
$1,866
$8,561
$12,246
$1,593
$1,639
$7,527
$10,759
2027
$2,468
$2,388
$10,915
$15,772
$2,167
$2,102
$9,622
$13,892
2028
$3,121
$2,910
$13,259
$19,290
$2,747
$2,568
$11,718
$17,033
2029
$3,768
$3,429
$15,591
$22,789
$3,329
$3,041
$13,854
$20,224
2030
$4,410
$3,944
$17,921
$26,276
$3,905
$3,506
$15,961
$23,373
2035
$7,367
$6,350
$29,254
$42,971
$6,632
$5,741
$26,507
$38,880
2040
$9,717
$8,423
$39,777
$57,916
$8,865
$7,716
$36,511
$53,093
2050
$10,787
$9,881
$48,442
$69,109
$9,843
$9,052
$44,464
$63,359
NPV, 3%
$94,080
$84,437
$398,245
$576,763
$85,014
$76,542
$361,745
$523,301
NPV, 7%
$38,342
$34,811
$163,449
$236,602
$34,530
$31,433
$147,870
$213,833
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and more dynamic baseline, lb, please see Preamble Section X. A. 1

-------
Table 7-32 Discounted MY Lifetime Reductions in Fuel Expenditures of the Final Program
Vs. The Flat Baseline and using Method B (3% Discount Rate, Millions of 2013$)a

REDUCED FUEL EXPENDITURES -
RETAIL
REDUCED FUEL EXPENDITURES -
UNTAXED
MODEL
YEAR
HD
PICKUPS
& VANS
voc
TRACTOR/
TRAILERS
SUM
HD
PICKUPS
& VANS
VOC
TRACTOR/
TRAILERS
SUM
2018
$0
$0
$781
$781
$0
$0
$680
$680
2019
$0
$0
$747
$747
$0
$0
$653
$653
2020
$0
$0
$719
$719
$0
$0
$631
$631
2021
$507
$2,127
$9,538
$12,171
$446
$1,875
$8,425
$10,746
2022
$1,346
$2,090
$9,477
$12,912
$1,187
$1,849
$8,399
$11,435
2023
$2,142
$2,055
$9,360
$13,557
$1,895
$1,824
$8,322
$12,041
2024
$2,927
$3,155
$14,627
$20,709
$2,597
$2,809
$13,045
$18,451
2025
$3,686
$3,152
$14,582
$21,420
$3,280
$2,814
$13,044
$19,137
2026
$4,418
$3,131
$14,427
$21,976
$3,941
$2,803
$12,943
$19,688
2027
$5,096
$4,141
$18,943
$28,180
$4,557
$3,717
$17,041
$25,315
2028
$5,043
$4,118
$18,791
$27,953
$4,521
$3,707
$16,949
$25,176
2029
$4,981
$4,098
$18,749
$27,828
$4,477
$3,697
$16,954
$25,128
Sum
$30,147
$28,066
$130,741
$188,954
$26,900
$25,094
$117,087
$169,081
Note:
aFor an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of
the flat baseline, la, and more dynamic baseline, lb, please see Preamble Section X.A.I
Table 7-33 Discounted MY Lifetime Reductions in Fuel Expenditures of the Final Program
Vs. The Flat Baseline and using Method B (7% Discount Rate, Millions of 2013$)a

REDUCED FUEL EXPENDITURES -
RETAIL
REDUCED FUEL EXPENDITURES -
UNTAXED
MODEL
YEAR
HD
PICKUPS
& VANS
VOC
TRACTOR/
TRAILERS
SUM
HD
PICKUPS
& VANS
VOC
TRACTOR/
TRAILERS
SUM
2018
$0
$0
$558
$558
$0
$0
$483
$483
2019
$0
$0
$510
$510
$0
$0
$444
$444
2020
$0
$0
$466
$466
$0
$0
$408
$408
2021
$312
$1,308
$5,831
$7,451
$274
$1,149
$5,132
$6,554
2022
$798
$1,238
$5,584
$7,620
$701
$1,091
$4,932
$6,725
2023
$1,222
$1,173
$5,315
$7,710
$1,078
$1,037
$4,711
$6,826
2024
$1,608
$1,735
$8,004
$11,347
$1,423
$1,539
$7,116
$10,078
2025
$1,951
$1,669
$7,689
$11,309
$1,731
$1,486
$6,858
$10,074
2026
$2,253
$1,598
$7,332
$11,182
$2,005
$1,427
$6,560
$9,991
2027
$2,504
$2,038
$9,276
$13,818
$2,234
$1,824
$8,323
$12,381
2028
$2,388
$1,952
$8,866
$13,206
$2,136
$1,753
$7,978
$11,866
2029
$2,274
$1,872
$8,526
$12,672
$2,039
$1,686
$7,693
$11,419
Sum
$15,311
$14,582
$67,957
$97,849
$13,621
$12,992
$60,636
$87,249
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and more dynamic baseline, lb, please see Preamble Section X. A. 1

-------
7.2.3 Maintenance Costs
The agencies have estimated increased maintenance costs associated with installation of
new technologies. The technologies for which we have estimated increased costs are shown in
Table 7-34 along with the estimated maintenance intervals and costs per event. We expect that,
when replaced, the lower rolling resistance tires would be replaced by equivalent performing
tires throughout the vehicle lifetime. As such, the incremental increases in costs for lower
rolling resistance tires would be incurred throughout the vehicle lifetime at intervals consistent
with current tire replacement intervals. Those intervals are difficult to quantify given the variety
of vehicles and operating modes within the HD industry. For HD pickups and vans, we have
chosen a tire replacement interval of 40,000 miles. We have done the same for all vocational
vehicles which is probably overly conservative as more frequent intervals results in higher
maintenance costs. For tractors and trailers, we have used a maintenance interval of 200,000
miles. The presence of tire inflation management systems, and the increased use of those
systems expected due to this final rule, should serve to improve tire maintenance intervals and
perhaps reduce vehicle downtime due to tire issues; they may also carry with them some
increased maintenance costs to ensure that the tire inflation systems themselves remain in proper
operation. For the analysis, we have considered these two competing factors to cancel each other
out. Similarly, the agencies considered the maintenance impact of 6x2 axles. As noted in the
NACFE Confidence Report on 6x2 axles, the industry expects an overall reduction in
maintenance costs and labor for vehicles with a 6x2 configuration as compared to a 6x4
configuration.5 The reduction in number of parts, such as the interaxle drive shaft, will reduce
the number of lubrication procedures needed and reduce the overall quantity of differential fluid
needed at change intervals. The agencies have taken a conservative approach to the maintenance
costs for the 6x2 technology and considered the incremental maintenance cost to be zero. The
other technologies shown carry with them the indicated costs per maintenance event conducted
at the indicated interval. These costs will be incurred according to the technology penetration
rates estimated and presented in Chapter 2 of this RIA. In other words, not all vehicles will incur
these costs, only those vehicles with the technologies will incur these costs.

-------
Table 7-34 Maintenance Costs and Miles per Event (2013$)
SEGMENT
TECHNOLOGY/SYSTEM
COST/EVENT
MILES/EVENT
Engines
Waste Heat Recovery
$300
100,000
2b/3 Pickups & Vans
Lower rolling resistance
tires level 2
Dependent on package
costs of the technology
40,000
Vocational vehicles
Lower rolling resistance
tires
Dependent on package
costs of the LRR
technology
40,000
Stop-start & automatic
engine shutdown system
$10 savings on oil
changes
10,000
Axle lubrication, tied to
high efficiency axles
$100
100,000
Transmission fluids, tied to
automated transmissions
$100
100,000
Hybrid systems
$3500
250,000
Tractors
Lower rolling resistance
tires
Dependent on package
costs of the LRR
technology
200,000
Auxiliary Power Unit
$300
100,000
Auxiliary Power Unit with
DPF
$400
100,000
Auxiliary Power Unit,
battery powered
$310
100,000
Axle lubrication, tied to
high efficiency axles
$100
500,000
Transmission fluids, tied to
powershift automatic
transmissions
$100
100,000
Fuel Operated Heaters
$110
100,000
Trailers
Lower rolling resistance
tires
Dependent on package
costs of the LRR
technology
200,000
In evaluating maintenance costs associated with the rule relative to the flat baseline, EPA
has used the maintenance intervals noted above, MOVES VMT, and the MOVES population of
specific MY vehicles in future calendar years to estimate the increased maintenance costs
associated with the final rule, again for each subcategory. Note that, in the context of the
benefit-cost analysis, EPA has estimated policy case maintenance costs using the policy case
VMT which, by definition, includes rebound VMT (see Section IX of the Preamble and Chapter
8 of this RIA for a discussion of rebound VMT).
Table 7-35 presents the annual in-use maintenance costs associated with the final
program along with net present values at 3 percent and 7 percent. This table presents costs
relative to the flat baseline and using the MOVES analysis for all vehicle categories (Method B).
Table 7-36 presents the model year lifetime in-use maintenance costs—versus the flat baseline
and using Method B— discounted at 3 percent. Table 7-37 presents the model year lifetime in-
use maintenance costs—versus the flat baseline and using Method B—discounted at 7 percent.

-------
Table 7-35 Annual Increased Maintenance Costs and Net Present Values Associated with the Final Program
Vs. The Flat Baseline and using Method B (SMillions of 2013$)a
CALENDAR
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$0.0
$0.0
$0.5
$0.5
2019
$0.0
$0.0
$1.1
$1.1
2020
$0.0
$0.0
$1.6
$1.6
2021
$0.9
$0.2
$17
$18
2022
$2.6
$0.3
$32
$35
2023
$5.2
$0.8
$47
$53
2024
$8.6
$7.6
$60
$76
2025
$13
$14
$72
$99
2026
$17
$20
$83
$119
2027
$21
$24
$106
$151
2028
$25
$28
$129
$182
2029
$28
$32
$151
$211
2030
$28
$32
$151
$211
2035
$28
$32
$151
$211
2040
$28
$32
$151
$211
2050
$28
$32
$151
$211
NPV, 3%
$367
$408
$2,014
$2,788
NPV, 7%
$167
$184
$933
$1,284
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1
Table 7-36 Discounted MY Lifetime Maintenance Costs of the Final Program
Vs. The Flat Baseline and using Method B (3% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$0.0
$0.0
$6.6
$6.6
2019
$0.0
$0.0
$6.4
$6.4
2020
$0.0
$0.0
$6.4
$6.4
2021
$7.1
$1.8
$129
$138
2022
$14
$1.2
$124
$139
2023
$20
$4.2
$120
$144
2024
$26
$50
$96
$172
2025
$32
$48
$94
$174
2026
$31
$39
$92
$162
2027
$31
$31
$184
$245
2028
$30
$28
$179
$237
2029
$29
$27
$174
$230
Sum
$220
$229
$1,211
$1,660
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1

-------
Table 7-37 Discounted MY Lifetime Maintenance Costs of the Final Program
Vs. The Flat Baseline and using Method B (7% Discount Rate, SMillions of 2013$)a
MODEL
YEAR
HD PICKUPS
& VANS
VOCATIONAL
TRACTOR/
TRAILERS
SUM
2018
$0.0
$0.0
$4.5
$4.5
2019
$0.0
$0.0
$4.3
$4.3
2020
$0.0
$0.0
$4.1
$4.1
2021
$4.5
$1.1
$80
$86
2022
$8.3
$0.7
$75
$84
2023
$12
$2.4
$69
$83
2024
$15
$28
$54
$96
2025
$17
$26
$50
$94
2026
$16
$20
$48
$84
2027
$15
$15
$92
$122
2028
$14
$14
$86
$114
2029
$13
$13
$80
$106
Sum
$115
$120
$647
$882
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation
of the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1
7.2.4 Analysis of Payback Periods
An important metric to vehicle purchasers is the payback period that can be expected on
any new purchase. In other words, there is greater willingness to pay for new technology if that
new technology "pays back" within an acceptable period of time. We make no effort to define
the acceptable period of time here, but seek to estimate the payback period for others to make the
decision themselves. We define the payback period as the point at which reduced fuel
expenditures outpace increased vehicle costs. For example, a new MY2027 tractor with trailer is
estimated to cost roughly $13,550 more (on average, including an "average" trailer, in 2013$,
and relative to the reference case vehicle) due to the addition of new GHG reducing/fuel
consumption improving technology. This new technology would result in lower fuel
consumption and, therefore, reduced fuel expenditures. But how many months or years would
pass before the reduced fuel expenditures would surpass the increased costs?
To estimate the costs, we have considered not only the cost of the new technology, but
also the taxes paid on the incrementally higher purchase expense, the slightly higher insurance
expenses on the slightly higher value vehicle, and the increased maintenance costs associated
with the new technology. Taxes paid were estimated as 6 percent sales tax in all regulated
sectors and a 12 percent excise tax applicable in the tractor/trailer and vocational sectors. As
such, the vehicle costs presented here are slightly higher than those presented elsewhere in this
RIA. For insurance costs, we have estimated the collision insurance to be 2 percent of the
purchase price of a vehicle consistent with the approach taken in our 2017-2025 light-duty
GHG/CAFE rule.6 Therefore, increased insurance costs would equal 2 percent of the increased
technology costs, and would be incurred every year going forward. But, since collision

-------
insurance is tied to vehicle value, we have also included a depreciation rate consisting of
straight-line depreciation of 3 percent each year through the 25th year of ownership at which time
we have flat-lined the depreciation and held vehicle value constant (see Table 7-58 in Chapter
7.2.6, below). For maintenance costs, we have used the same method described above. Also,
here we use retail fuel prices since those are the prices paid by owners of these vehicles.
We have conducted this payback analysis for HD pickups and vans, vocational vehicles
and for tractor/trailers (including the engines used in each of these subcategories). All
calculations are for the average vehicle, or average tractor/trailer combination, that drives the
average number of miles each year. The calculations do not represent specific vehicle classes or
specific use cases so should not be seen as being applicable to any particular individual's
situation. However, the payback periods do provide a general sense, on average, of what sort of
payback periods are likely at a national, societal perspective.
Table 7-38 presents the discounted annual increased vehicle costs and fuel expenditure
impacts associated with owning a new MY2027 HD pickup or van using both 3 percent and 7
percent discount rates. The results in this table use Method B. As shown in the table, the
payback for HD pickups and vans occurs in the 3rd year of ownership (the year in which
cumulative expenditures become positive) using a 3 percent and 7 percent discount rate.
Table 7-38 Discounted Owner Expenditures & Payback Period for MY2027 HD Pickups & Vans under the
Final Program Vs. The Flat Baseline and using Method B
3% and 7% Discount Rates (2013$)a
Age
3% Discount Rate
7% Discount Rate
Technology
cost,
taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
Technology
cost,
taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
1
-$1,451
-$4
$550
-$905
-$1,424
-$4
$540
-$888
2
-$25
-$4
$539
-$395
-$24
-$3
$509
-$406
3
-$24
-$3
$527
$105
-$21
-$3
$479
$49
4
-$22
-$3
$515
$595
-$19
-$3
$451
$477
5
-$21
-$3
$492
$1,064
-$17
-$3
$415
$872
6
-$19
-$3
$469
$1,511
-$16
-$2
$381
$1,235
7
-$18
-$3
$446
$1,936
-$14
-$2
$348
$1,567
8
-$17
-$2
$423
$2,340
-$13
-$2
$318
$1,870
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b 6% sales tax; insurance estimates are described in text.
0 Fuel expenditures calculated using retail fuel prices according to AEO2015 reference fuel price case.
Table 7-39 and Table 7-40 show the same information for a MY2027 vocational vehicle
and a tractor/trailer, respectively. As shown, payback for vocational vehicles occurs in the 4th
year of ownership while payback for tractor/trailers occurs early in the 2nd year of ownership.

-------
Table 7-39 Discounted Owner Expenditures & Payback Period for MY2027 Vocational Vehicles under the
Final Program Vs. The Flat Baseline and using Method B
3% and 7% Discount Rates (2013$)a
Age
3% Discount Rate
7% Discount Rate
Technology
cost,
taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
Technology
cost,
taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
1
-$3,147
-$25
$1,022
-$2,151
-$3,088
-$25
$1,003
-$2,110
2
-$49
-$24
$1,004
-$1,220
-$46
-$23
$948
-$1,231
3
-$46
-$24
$987
-$303
-$42
-$21
$898
-$397
4
-$43
-$23
$970
$602
-$38
-$20
$849
$394
5
-$40
-$21
$909
$1,450
-$34
-$18
$766
$1,109
6
-$38
-$19
$850
$2,243
-$31
-$15
$689
$1,752
7
-$35
-$17
$796
$2,987
-$27
-$14
$622
$2,333
8
-$33
-$16
$743
$3,681
-$25
-$12
$558
$2,854
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b 6% sales tax and 12% excise tax; insurance estimates are described in text.
0 Fuel expenditures calculated using retail fuel prices according to AEO2015 reference fuel price case.
Table 7-40 Discounted Owner Expenditures & Payback Period for MY2027Tractor/Trailers under the Final
Program Vs. The Flat Baseline and using Method B
3% and 7% Discount Rates (2013$)a
Age
3% Discount Rate
7% Discount Rate
Technology
cost,
taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
Technology
cost,
taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
1
-$16,022
-$169
$15,310
-$880
-$15,719
-$166
$15,021
-$864
2
-$251
-$163
$15,095
$13,801
-$237
-$154
$14,256
$13,002
3
-$235
-$158
$14,872
$28,280
-$214
-$144
$13,521
$26,166
4
-$220
-$153
$14,637
$42,545
-$192
-$134
$12,809
$38,649
5
-$206
-$140
$13,683
$55,882
-$173
-$118
$11,527
$49,885
6
-$192
-$127
$12,730
$68,292
-$156
-$103
$10,323
$59,950
7
-$179
-$116
$11,880
$79,878
-$140
-$90
$9,274
$68,993
8
-$166
-$105
$11,025
$90,630
-$125
-$79
$8,285
$77,074
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b 6% sales tax and 12% excise tax; insurance estimates are described in text.
0 Fuel expenditures calculated using retail fuel prices according to AEO2015 reference fuel price case.
The fuel expenditure column uses retail fuel prices specific to gasoline and diesel fuel as
projected in AEO2015. This payback analysis does not include other private impacts, such as
reduced refueling events, or other societal impacts, such as noise, congestion and crashes. It also
does not include societal impacts such as co-pollutant environmental benefits or benefits
associated with reduced GHG emissions. We use retail fuel prices and exclude these other
private and social impacts because the focus is meant to be on those factors that buyers think

-------
about most while considering a new vehicle purchase and those factors that result in more or
fewer dollars in their pockets.
In an effort to provide further information on payback, we have also looked at the
payback periods for more specific vehicle subcategories. For example, while the tractor/trailer
payback shown in Table 7-40 occurs early in the 2nd year, the payback for a Class 8 sleeper cab
would occur within the first year of ownership as shown in Table 7-41.
Table 7-41 Discounted Owner Expenditures & Payback Period for MY2027 Sleeper Cab with Trailer under
the Final Program Vs. The Flat Baseline and using Method B
3% and 7% Discount Rates (2013$)a
Age
3% Discount Rate
7% Discount Rate
Technology
cost, taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
Technology
cost, taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
1
-$17,523
-$538
$19,926
$1,866
-$17,192
-$528
$19,550
$1,830
2
-$274
-$521
$19,646
$20,717
-$259
-$492
$18,555
$19,635
3
-$257
-$503
$19,356
$39,314
-$234
-$458
$17,598
$36,542
4
-$241
-$486
$19,050
$57,637
-$211
-$426
$16,672
$52,577
5
-$225
-$447
$17,867
$74,832
-$189
-$376
$15,052
$67,064
6
-$210
-$409
$16,688
$90,901
-$170
-$332
$13,533
$80,094
7
-$196
-$375
$15,642
$105,972
-$153
-$293
$12,211
$91,860
8
-$182
-$343
$14,583
$120,030
-$137
-$258
$10,958
$102,423
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b 6% sales tax and 12% excise tax; insurance estimates are described in text.
0 Fuel expenditures calculated using retail fuel prices according to AEO2015 reference fuel price case.
Given the variety in the vocational market, the subcategory analysis becomes more
interesting. For example, Table 7-42 shows the payback for an intercity bus. Table 7-43 shows
the same information for a transit bus, while Table 7-44 shows this information for a school bus.
These tables highlight how much the payback period can vary depending on the level of
technology cost and fuel consumption improvement versus the number of miles driven. The
high VMT intercity bus (-80,000 miles/year) and transit bus (-60,000 miles/year) payback in the
1st and 2nd year, respectively, despite first year costs exceeding $6, 000 and $5, 000,
respectively. By contrast, the lower VMT school bus (-13,000 miles/year) pays back in the
7th year (or 8th year with 7 percent discounting) despite first year costs under $4, 000.

-------
Table 7-42 Discounted Owner Expenditures & Payback Period for MY2027 Intercity Bus under the Final
Program Vs. The Flat Baseline and using Method B
3% and 7% Discount Rates (2012$)a
Age
3% Discount Rate
7% Discount Rate
Technology
cost, taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
Technology
cost, taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures 0
Cumulative
expenditures
1
-$5,848
-$427
$6,739
$465
-$5,738
-$419
$6,612
$456
2
-$91
-$412
$6,628
$6,589
-$86
-$389
$6,260
$6,240
3
-$86
-$398
$6,522
$12,627
-$78
-$362
$5,929
$11,730
4
-$80
-$384
$6,415
$18,578
-$70
-$336
$5,614
$16,937
5
-$75
-$370
$6,313
$24,445
-$63
-$312
$5,318
$21,880
6
-$70
-$356
$6,199
$30,218
-$57
-$289
$5,027
$26,562
7
-$65
-$344
$6,127
$35,936
-$51
-$269
$4,783
$31,025
8
-$61
-$333
$6,044
$41,586
-$46
-$250
$4,542
$35,271
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b 6% sales tax and 12% excise tax; insurance estimates are described in text.
0 Fuel expenditures calculated using retail fuel prices according to AEO2015 reference fuel price case.
Table 7-43 Discounted Owner Expenditures & Payback Period for MY2027 Diesel Fueled Transit Bus under
the Final Program Vs. The Flat Baseline and using Method B
3% and 7% Discount Rates (2013$)a
Age
3% Discount Rate
7% Discount Rate
Technology
cost,
taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
Technology
cost,
taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
1
-$4,908
-$79
$3,437
-$1,550
-$4,815
-$78
$3,372
-$1,521
2
-$77
-$74
$3,273
$1,571
-$73
-$70
$3,091
$1,427
3
-$72
-$69
$3,118
$4,549
-$65
-$63
$2,835
$4,134
4
-$67
-$65
$2,967
$7,384
-$59
-$57
$2,597
$6,615
5
-$63
-$60
$2,826
$10,087
-$53
-$51
$2,381
$8,892
6
-$59
-$56
$2,687
$12,659
-$48
-$46
$2,179
$10,978
7
-$55
-$53
$2,573
$15,125
-$43
-$41
$2,009
$12,903
8
-$51
-$49
$2,456
$17,481
-$38
-$37
$1,846
$14,674
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b 6% sales tax and 12% excise tax; insurance estimates are described in text.
0 Fuel expenditures calculated using retail fuel prices according to AEO2015 reference fuel price case.

-------
Table 7-44 Discounted Owner Expenditures & Payback Period for MY2027 Diesel Fueled School Bus under
the Final Program Vs. The Less Dynamic Baseline and using Method B
3% and 7% Discount Rates (2012$)a
Age
3% Discount Rate
7% Discount Rate
Technology
cost,
taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
Technology
cost,
taxes,
insurance b
Maintenance
expenditures
Fuel
expenditures
C
Cumulative
expenditures
1
-$3,309
-$16
$573
-$2,752
-$3,247
-$15
$562
-$2,700
2
-$52
-$15
$563
-$2,255
-$49
-$14
$532
-$2,231
3
-$49
-$15
$554
-$1,764
-$44
-$13
$504
-$1,784
4
-$45
-$14
$545
-$1,278
-$40
-$12
$477
-$1,359
5
-$42
-$14
$537
-$798
-$36
-$11
$452
-$954
6
-$40
-$13
$527
-$323
-$32
-$11
$427
-$570
7
-$37
-$13
$521
$148
-$29
-$10
$407
-$202
8
-$34
-$12
$514
$615
-$26
-$9
$386
$149
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the less
dynamic baseline, la, and more dynamic baseline, lb, please see Preamble Section X.A.I
b 6% sales tax and 12% excise tax; insurance estimates are described in text.
0 Fuel expenditures calculated using retail fuel prices according to AEO2014 early release, reference case estimates.
We could present tables for each MOVES subcategory, but since all are calculated using
the same methodology, the detailed tables seem unnecessary. Instead, we provide Table 7-45
which summarizes the payback period for each MOVES subcategory at both 3 percent and 7
percent discount rates and for each fuel type.

-------
Table 7-45 Payback Periods Associated with the Final Program Vs. The Flat Baseline and using Method B
for MY2027 Vehicle Subcategories at 3% and 7% Discount Rates Payback occurs in Year Showna
Subcategory
3% Discount Rate
7% Discount Rate
Gasoline
Diesel

Gasoline
Diesel

HD Pickups & Vans (MY2027)
4
3

4
3

Vocational (MY2027 for each)






Intercity bus
N/A
1

N/A
1

Transit bus
2
2

2
2

School bus
8
7

9
8

Refuse truck
N/A
2

N/A
2

Single unit short haul
3
4

4
4

Single unit long haul
N/A
3

N/A
3

Motor home
27
29

>30
>30

Tractor/Trailer (MY2027 for each)






Combination short haul
N/A
2

N/A
2

Combination long haul
N/A
1

N/A
1

Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of
the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
N/A denotes no such vehicles in this segment.
7.2.5 Cost per Ton of CO2 Equivalent Reduced vs. the Flat Baseline and using
Method B
The agencies have calculated the cost per ton of GHG (CCh-equivalent, or CCheq)
reductions associated with this rulemaking using the costs presented in Chapter 7.2.1 and 7.2.2,
and the GHG emissions reductions described in Chapter 5 of this RIA but expressed here as
CO2-equivalents (CChe). These costs per ton-reduction values are presented in Table 7-46
through Table 7-49 for HD pickups & vans, vocational vehicles, tractor/trailers and all segments,
respectively. The cost per metric ton of CChe emissions reductions in 2050 represents the long-
term cost per ton of the emissions reduced. The agencies have also calculated the cost per metric
ton of CChe emission reductions including the savings associated with reduced fuel
consumption.
The calculations presented here include all engine- and vehicle-related costs but do not
include benefits associated with the final program such as those associated with criteria pollutant
reductions or energy security benefits (discussed in Chapter 8 of this RIA). By including the fuel
savings, the cost per ton-reduction is less than $0 since the estimated value of fuel savings
outweighs the program costs.

-------
Table 7-46 Annual Cost per Metric Ton of CCheq Emissions Reduced in the Final Program
Vs. The Flat Baseline and using Method B
HD Pickups and Vans only (dollar values are 2013$)a
Calendar
Year
Vehicle &
Maintenance
Costs
($Billions)
Fuel
Savings
($Billions)
GHG
Reduced
(MMT)
$/metric
ton w/o
fuel
$/metric
ton w/ fuel
2021
$0.6
$0.0
0.2
$2,800
$2,500
2024
$1.1
$0.7
3.1
$370
$140
2027
$1.7
$2.2
8.9
$190
-$57
2030
$1.6
$3.9
15
$110
-$150
2035
$1.7
$6.6
23
$73
-$220
2040
$1.7
$8.9
27
$63
-$270
2050
$1.8
$9.8
30
$59
-$270
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for
an explanation of the flat baseline, la, and dynamic baseline, lb, please see Preamble
Section X. A. 1 GHG reductions include CO2 and CO2 equivalents of CH4, and N20.
Table 7-47 Annual Cost per Metric Ton of CCheq Emissions Reduced in the Final Program
Vs. The Flat Baseline and using Method B
Vocational Vehicles only (dollar values are 2013$)a
Calendar
Year
Vehicle &
Maintenance
Costs
($Billions)
Fuel
Savings
($Billions)
GHG
Reduced
(MMT)
$/metric
ton w/o
fuel
$/metric
ton w/ fuel
2021
$0.7
$0.2
1.0
$710
$510
2024
$1.2
$1.0
4.4
$270
$53
2027
$1.5
$2.1
9.0
$160
-$69
2030
$1.5
$3.5
14
$110
-$140
2035
$1.7
$5.7
20
$81
-$200
2040
$1.8
$7.7
24
$76
-$240
2050
$2.2
$9.1
29
$77
-$240
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for
an explanation of the flat baseline, la, and dynamic baseline, lb, please see Preamble
Section X.A.I GHG reductions include CO2 and CO2 equivalents of CH4, N20 and
HFCs.

-------
Table 7-48 Annual Cost per Metric Ton of CCheq Emissions Reduced in the Final Program
Vs. The Flat Baseline and using Method B
Tractor/Trailers only (dollar values are 2013$)a
Calendar
Year
Vehicle &
Maintenance
Costs
($Billions)
Fuel
Savings
($Billions)
GHG
Reduced
(MMT)
$/metric
ton w/o
fuel
$/metric
ton w/ fuel
2021
$1.2
$1.0
5.0
$250
$46
2024
$1.9
$4.4
20
$94
-$120
2027
$2.3
$9.6
41
$54
-$180
2030
$2.3
$16
64
$36
-$210
2035
$2.6
$27
95
$27
-$250
2040
$3.0
$37
115
$26
-$290
2050
$3.5
$44
141
$25
-$290
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for
an explanation of the flat baseline, la, and more dynamic baseline, lb, please see
Preamble Section X.A.I GHG reductions include CO2 and CO2 equivalents of CH4, and
N20.
Table 7-49 Annual Cost per Metric Ton of CCheq Emissions Reduced in the Final Program
Vs. The Flat Baseline and using Method B
All Vehicle Segments (dollar values are 2013$)a
Calendar
Year
Vehicle &
Maintenance
Costs
($Billions)
Fuel
Savings
($Billions)
GHG
Reduced
(MMT)
$/metric
ton w/o
fuel
$/metric
ton w/ fuel
2021
$2.6
$1.3
6.2
$410
$210
2024
$4.2
$6.1
28
$150
-$65
2027
$5.4
$14
59
$91
-$140
2030
$5.5
$23
94
$59
-$190
2035
$5.9
$39
138
$43
-$240
2040
$6.5
$53
167
$39
-$280
2050
$7.5
$63
199
$38
-$280
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for
an explanation of the flat baseline, la, and dynamic baseline, lb, please see Preamble
Section X.A.I GHG reductions include CO2 and CO2 equivalents of CH4, N20 and
HFCs.
For comparison, Table 7-50 through Table 7-53 show the same information as it was
presented in Chapter 7 of the final RIA for the Phase 1 HD rule.7

-------
Table 7-50 Annual Cost per Metric Ton of CCheq Emissions Reduced in the HD Phase 1 Final Rule
HD Pickups and Vans only (dollar values are 2009$)
Calendar
Year
Vehicle &
Maintenance
Costs
($Billions)
Fuel
Savings
($Billions)
GHG
Reduced
(MMT)
$/metric
ton w/o
fuel
$/metric
ton w/ fuel
2020
$0,8
$0.9
3
$240
-$30
2030
$0.9
$3.0
10
$90
-$200
2040
$1.0
$4.3
14
$70
-$240
2050
$1.2
$5.5
16
$80
-$270
Table 7-51 Annual Cost per Metric Ton of CCheq Emissions Reduced in the HD Phase 1 Final Rule
Vocational Vehicles only (dollar values are 2009$)
Calendar
Year
Vehicle &
Maintenance
Costs
($Billions)
Fuel
Savings
($Billions)
GHG
Reduced
(MMT)
$/metric
ton w/o
fuel
$/metric
ton w/ fuel
2020
$0.2
$1.1
4
$50
-$210
2030
$0.2
$2.4
9
$20
-$250
2040
$0.3
$3.5
12
$30
-$270
2050
$0.4
$4.7
14
$30
-$310
Table 7-52 Annual Cost per Metric Ton of CCheq Emissions Reduced in the HD Phase 1 Final Rule
Tractor/Trailers only (dollar values are 2009$)
Calendar
Year
Vehicle &
Maintenance
Costs
($Billions)
Fuel
Savings
($Billions)
GHG
Reduced
(MMT)
$/metric
ton w/o
fuel
$/metric
ton w/ fuel
2020
$1.0
$7.7
32
$30
-$210
2030
$1.1
$15.3
57
$20
-$250
2040
$1.4
$20.2
68
$20
-$280
2050
$1.8
$26.4
78
$20
-$320
Table 7-53 Annual Cost per Metric Ton of CCheq Emissions Reduced in the HD Phase 1 Final Rule
All Vehicle Segments (dollar values are 2009$)
Calendar
Year
Vehicle &
Maintenance
Costs
($Billions)
Fuel
Savings
($Billions)
GHG
Reduced
(MMT)
$/metric
ton w/o
fuel
$/metric
ton w/ fuel
2020
$2.0
$9.6
39
$50
-$190
2030
$2.2
$20.6
76
$30
-$240
2040
$2.7
$28.0
94
$30
-$270
2050
$3.3
$36.5
108
$30
-$310

-------
7.2.6 Costs and Benefits for each Regulatory Subcategory using the Flat
Baseline and Method B
The full presentation of program costs and benefits is in Chapter 8 of this RIA. Please see
that chapter for details behind the social cost of carbon, non-GHG pollution benefits, energy
security, and all of the other metrics that go into developing the full cost and benefit analysis.
Here we present simply the high level cost, fuel savings, benefits and net benefits for each of the
3 regulatory subcategories: HD pickups and vans, vocational vehicles and tractor/trailers.
Table 7-54 Costs, Fuel Savings, Benefits & Net Benefits for each Regulatory Subcategory in the MY Lifetime
Analysis (Billions of 2013$)a b c


3% Discount Rate
7% Discount Rate
Costs (Technology & Maintenance)
HD Pickups & Vans
-$8.5
-$5.6
Vocational Vehicles
-$7.4
-$4.8
Tractor/Trailers
-$12.5
-$8.2
Total
-$28.4
-$18.6
Fuel Savings
HD Pickups & Vans
$26.9
$13.6
Vocational Vehicles
$25.1
$13.0
Tractor/Trailers
$117.1
$60.6
Total
$169.1
$87.2
Benefits
HD Pickups & Vans
$14.1
$9.8
Vocational Vehicles
$12.3
$8.8
Tractor/Trailers
$61.8
$43.6
Total
$88.2
$62.3
Net Benefits
HD Pickups & Vans
$32.4
$17.8
Vocational Vehicles
$30.0
$17.0
Tractor/Trailers
$166.4
$96.1
Total
$228.8
$130.9
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of
the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b The monetized GHG benefits presented in this analysis exclude the value of changes in HFC emissions
expected under this program (see RIA Chapter 8.5). Although EPA has not monetized changes in HFCs in
the main benefits analysis, the value of any increases or reductions should not be interpreted as zero.
0 GHG benefit estimates include reductions in CO2, CH4, and N20 but do not include the HFC reductions.
Note that net present value of reduced CO2 GHG emissions is calculated differently than other benefits.
The same discount rate used to discount the value of damages from future emissions (SC-CO2, SC-CH4,
and SC-N2O, each discounted at rates of at 5, 3, 2.5 percent) is used to calculate net present value of SC-
CO2, SC-CH4, and SC-N2O, respectively, SC-CO2 for internal consistency. Refer to the SC-CO2 TSD for
more detail.
7.3 Key Parameters Used in the Estimation of Costs and Fuel Savings
This section presents some of the parameters used in generating expenditure impacts
associated with the program. Table 7-55 presents estimated sales of complying vehicles by
calendar year. Table 7-56 presents $/gallon in the AEO 2015 reference fuel price case. Note
that AEO projects fuel prices out to 2040. Table 7-57 presents AEO 2014 final reference case
fuel prices which are used by both agencies in the CAFE Model for HD pickups and vans in the

-------
proposal, and used by EPA in this final rule. For that analysis, the retail (post-tax) prices are
increased for each year after 2040 by 0.2 percent for gasoline and 0.7 percent for diesel. For
years beyond 2040, EPA has kept fuel prices at the 2040 level rather than growing those fuel
prices at a rate consistent with years prior to 2040. Table 7-58 shows the depreciation rates used
in the payback period analysis presented in Chapter 7.2. Table 7-59 through Table 7-61 show
the policy and reference case VMT values used in MOVES modeling.
Table 7-55 Estimated Calendar Year Sales by Vehicle Type using Method B a'b
Calendar Year
HD Pickup &
Vans
Vocational
Vehicles
Tractors
Semi-trailers
2018
1,206,112
471,994
134,141
158,286
2019
1,192,088
476,252
138,240
163,123
2020
1,195,369
485,983
144,154
170,102
2021
1,184,184
484,752
144,737
170,790
2022
1,176,320
486,068
145,814
172,061
2023
1,174,470
487,849
146,257
172,583
2024
1,182,761
498,683
150,729
177,860
2025
1,191,602
508,256
152,898
180,420
2026
1,200,976
515,592
154,105
181,844
2027
1,201,868
523,805
155,682
183,705
2028
1,201,965
531,284
157,395
185,726
2029
1,200,297
539,624
160,217
189,056
2030
1,196,706
549,322
164,275
193,845
2031
1,191,071
557,981
168,017
198,260
2032
1,189,075
567,362
171,578
202,462
2033
1,191,398
580,104
176,600
208,388
2034
1,199,387
595,739
182,637
215,512
2035
1,207,377
610,539
188,035
221,881
2036
1,216,582
626,546
194,248
229,213
2037
1,224,403
641,706
200,144
236,170
2038
1,231,432
656,449
205,795
242,838
2039
1,235,794
669,757
210,833
248,783
2040
1,241,415
683,894
216,332
255,272
2041
1,247,054
696,936
220,387
260,057
2042
1,252,832
710,223
224,518
264,931
2043
1,258,753
723,769
228,725
269,896
2044
1,264,820
737,569
233,014
274,957
2045
1,271,039
751,639
237,381
280,110
2046
1,277,407
765,973
241,831
285,361
2047
1,283,920
780,577
246,364
290,710
2048
1,290,589
795,451
250,981
296,158
2049
1,297,425
810,611
255,686
301,709
2050
1,304,420
826,068
260,480
307,366
Notes:
a Sales are estimated using population data contained in MOVES. See Chapter 5 of this RIA for a description of
the MOVES modeling done in support of this rule.
b For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the
flat baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
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
Table 7-56 AEO 2015 Reference Fuel Price Case (2013$/gallon)
Pre-Tax
Retail
Gasoline
$2.30
$2.30
$2.35
$2.39
$2.43
$2.47
$2.52
$2.57
$2.62
$2.66
$2.71
$2.76
$2.82
$2.88
$2.95
$3.02
$3.09
$3.16
$3.23
$3.30
$3.38
$3.47
$3.54
$3.54
$3.54
$3.54
$3.54
$3.54
$3.54
$3.54
$3.54
$3.54
$3.54
Diesel
$2.62
$2.66
$2.72
$2.78
$2.86
$2.93
$2.99
$3.05
$3.13
$3.20
$3.27
$3.35
$3.42
$3.50
$3.59
$3.68
$3.76
$3.86
$3.95
$4.05
$4.16
$4.26
$4.36
$4.36
$4.36
$4.36
$4.36
$4.36
$4.36
$4.36
$4.36
$4.36
$4.36
Gasoline
$2.70
$2.70
$2.74
$2.78
$2.82
$2.86
$2.90
$2.95
$3.00
$3.04
$3.09
$3.14
$3.20
$3.26
$3.33
$3.39
$3.46
$3.53
$3.60
$3.66
$3.74
$3.83
$3.90
$3.90
$3.90
$3.90
$3.90
$3.90
$3.90
$3.90
$3.90
$3.90
$3.90
Diesel
$3.08
$3.12
$3.17
$3.23
$3.31
$3.37
$3.43
$3.49
$3.56
$3.63
$3.70
$3.77
$3.84
$3.92
$4.00
$4.09
$4.17
$4.26
$4.35
$4.45
$4.55
$4.65
$4.75
$4.75
$4.75
$4.75
$4.75
$4.75
$4.75
$4.75
$4.75
$4.75
$4.75

-------
Table 7-57 AEO 2014 Final Reference Fuel Price Case Used in the CAFE Model for HD Pickups and Vans;
Used by both Agencies in the Proposal & by EPA in this Final Rule (2012$/gallon)

Pre-Tax
Post-Tax
Calendar
Year
Gasoline
Diesel
Gasoline
Diesel
2018
$2.63
$3.10
$3.02
$3.53
2019
$2.64
$3.19
$3.03
$3.61
2020
$2.69
$3.25
$3.08
$3.67
2021
$2.74
$3.32
$3.12
$3.74
2022
$2.79
$3.41
$3.17
$3.82
2023
$2.84
$3.46
$3.22
$3.87
2024
$2.88
$3.51
$3.26
$3.92
2025
$2.92
$3.58
$3.29
$3.98
2026
$2.95
$3.62
$3.32
$4.02
2027
$2.99
$3.68
$3.36
$4.08
2028
$3.00
$3.73
$3.37
$4.12
2029
$3.03
$3.77
$3.40
$4.16
2030
$3.07
$3.81
$3.43
$4.20
2031
$3.10
$3.87
$3.46
$4.25
2032
$3.14
$3.92
$3.50
$4.30
2033
$3.18
$3.98
$3.54
$4.36
2034
$3.27
$4.06
$3.62
$4.43
2035
$3.30
$4.10
$3.65
$4.47
2036
$3.34
$4.14
$3.69
$4.51
2037
$3.38
$4.18
$3.73
$4.54
2038
$3.43
$4.22
$3.77
$4.58
2039
$3.49
$4.29
$3.83
$4.65
2040
$3.56
$4.38
$3.90
$4.73
2041
$3.57
$4.41
$3.91
$4.76
2042
$3.58
$4.44
$3.92
$4.80
2043
$3.59
$4.48
$3.93
$4.83
2044
$3.59
$4.51
$3.93
$4.86
2045
$3.60
$4.54
$3.94
$4.90
2046
$3.61
$4.58
$3.95
$4.93
2047
$3.62
$4.61
$3.96
$4.97
2048
$3.63
$4.65
$3.97
$5.00
2049
$3.63
$4.68
$3.97
$5.04
2050
$3.64
$4.72
$3.98
$5.07

-------
Table 7-58 Depreciation Schedule used in Payback Analysis for Method B a
Age
Depreciation
0
0%
1
3%
2
7%
3
10%
4
13%
5
17%
6
20%
7
23%
8
27%
9
30%
10
33%
11
37%
12
40%
13
43%
14
47%
15
50%
16
53%
17
57%
18
60%
19
63%
20
67%
21
70%
22
73%
23
77%
24
80%
25
83%
26
83%
27
83%
28
83%
29
83%
30
83%
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of
the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
Table 7-59 Reference Case and Policy Case Vehicle Miles Traveled (VMT)
For the Final Program relative to the Flat Baseline using Method B
HD Pickups and Vans a
Model Year
Reference case
Policy Case
Rebound VMT
2018
243,446,459,798
243,446,459,798
0
2019
240,544,622,588
240,544,622,588
0
2020
241,190,926,860
241,190,926,860
0
2021
238,846,698,033
241,426,272,670
2,579,574,637
2022
237,380,423,724
239,944,040,068
2,563,616,344
2023
237,153,891,479
239,715,144,759
2,561,253,281
2024
239,066,747,610
241,648,660,086
2,581,912,475
2025
241,062,399,725
243,665,842,812
2,603,443,087
2026
243,119,992,516
245,745,591,658
2,625,599,142
2027
243,534,755,232
246,164,989,408
2,630,234,175
2028
243,820,406,557
246,453,561,767
2,633,155,210
2029
243,718,985,090
246,351,151,547
2,632,166,457
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of
the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 7-60 Reference Case and Policy Case Vehicle Miles Traveled (VMT)
For the Final Program relative to the Flat Baseline using Method B
Vocational Vehicles a
Model Year
Reference case
Policy Case
Rebound VMT
2018
109,299,356,451
109,299,356,451
0
2019
109,171,917,190
109,171,917,190
0
2020
110,312,045,137
110,312,045,137
0
2021
108,908,544,746
109,235,261,181
326,716,435
2022
108,219,636,901
108,544,256,343
324,619,441
2023
107,648,982,428
107,971,989,117
323,006,689
2024
109,094,964,009
109,422,205,744
327,241,735
2025
110,256,962,536
110,587,744,520
330,781,984
2026
110,830,009,427
111,162,555,501
332,546,074
2027
111,618,481,660
111,953,312,055
334,830,395
2028
112,430,556,692
112,767,875,523
337,318,830
2029
113,281,216,713
113,621,013,246
339,796,533
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of
the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
Table 7-61 Reference Case and Policy Case Vehicle Miles Traveled (VMT)
For the Final Program relative to the Flat Baseline using Method B
Tractor/Trailera
Model Year
Reference case
Policy Case
Rebound VMT
2018
196,687,058,720
197,818,200,910
1,131,142,190
2019
202,939,391,970
204,223,800,600
1,284,408,630
2020
211,438,197,150
212,899,481,640
1,461,284,490
2021
211,708,016,651
213,296,020,576
1,588,003,925
2022
212,548,914,706
214,143,012,332
1,594,097,627
2023
212,247,027,109
213,838,908,002
1,591,880,893
2024
217,583,929,523
219,215,829,412
1,631,899,889
2025
219,380,061,614
221,025,659,538
1,645,597,923
2026
219,528,087,171
221,174,595,841
1,646,508,670
2027
220,148,138,758
221,799,166,682
1,651,027,924
2028
221,151,741,304
222,810,299,501
1,658,558,197
2029
223,496,517,913
225,173,028,204
1,676,510,291
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of
the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
References
1	Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy Standards; Final
Rulemaking 75 Fed. Reg. 25323 (May 7, 2010).
2	Final Rulemaking to Establish Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium-
and Heavy-Duty Engines and Vehicles, Regulatory Impact Analysis, EPA-420-R-11-901, August 2011.
3	All of EPA's calculations of costs and benefits (monetized) along with paybacks can be found in the docket on a
DVD titled, "GHGHD2 BCA."
4	Final Rulemaking to Establish Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium-
and Heavy-Duty Engines and Vehicles, Regulatory Impact Analysis, EPA-420-R-11-901, August 2011.
5	North American Council for Freight Efficiency. Confidence Findings on the Potential of 6x2 Axles. 2013. Pages
30-31.
6	Final Rulemaking for 2017-2025 Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average
Fuel Economy Standards, Regulatory Impact Analysis, EPA-420-R-12-016, August 2012.
7	Final Rulemaking to Establish Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium-
and Heavy-Duty Engines and Vehicles, Regulatory Impact Analysis, EPA-420-R-11-901, August 2011.

-------
Chapter 8. Economic and Other Impacts
8.1 Framework for Benefits and Costs
This chapter presents the costs, benefits and other economic impacts of the Phase 2
standards. It is important to note that NHTSA's fuel consumption standards and EPA's GHG
standards will both be in effect, and each will lead to average fuel efficiency increases and GHG
emission reductions.
The net benefits of the Phase 2 standards consist of the effects of the program on:
•	vehicle program costs (costs of complying with the vehicle CO2 and fuel consumption
standards)
•	changes in fuel expenditures associated with reduced fuel use resulting from more
efficient vehicles and increased fuel use associated with the "rebound" effect, both of
which result from the program
•	economic value of reductions in GHGs
•	economic value of reductions in other non-GHG pollutants
•	costs associated with increases in noise, congestion, and crashes resulting from increased
vehicle use
•	savings in drivers' time from less frequent refueling
•	benefits of increased vehicle use associated with the "rebound" effect
•	economic value of improvements in U.S. energy security
The benefits and costs of these rules are analyzed using 3 percent and 7 percent discount
rates, consistent with current OMB guidance.A These rates are intended to represent consumers'
preference for current over future consumption (3 percent), and the real rate of return on private
investment (7 percent) which indicates the opportunity cost of capital. However, neither of these
rates necessarily represents the discount rate that individual decision-makers use.
The program may also have other economic effects that are not included here. In
particular, as discussed in Chapter 2 of the RIA, the technology cost estimates developed here
take into account the costs to hold other vehicle attributes, such as size and performance,
constant. With these assumptions, and because welfare losses represent monetary estimates of
how much buyers would have to be compensated to be made as well off as they would have been
in the absence of this regulation,6 price increases for new vehicles measure the welfare losses to
A The range of Social Cost of Carbon (SCC) values uses several discount rates because the literature shows that the
SCC is quite sensitive to assumptions about the discount rate, and because no consensus exists on the appropriate
rate to use in an intergenerational context (where costs and benefits are incurred by different generations). Refer to
Section 8.4 for more information.
B This approach describes the economic concept of compensating variation, a payment of money after a change that
would make a consumer as well off after the change as before it. A related concept, equivalent variation, estimates
the income change that would be an alternative to the change taking place. The difference between them is whether

-------
the vehicle buyers.0 If the full technology cost gets passed along to the buyer as an increase in
price, the technology cost thus measures the primary welfare loss of the standards, including
impacts on buyers. Increasing fuel efficiency would have to lead to other changes in the vehicles
that buyers find undesirable for there to be additional welfare losses that are not included in the
technology costs.
As the 2012-2016 and 2017-2025 light-duty GHG/CAFE rules discussed, if other vehicle
attributes are not held constant, then the technology cost estimates do not capture the losses to
vehicle buyers associated with these changes.1 The light-duty rules also discussed other
potential issues that could affect the calculation of the welfare impacts of these types of changes,
such as aspects of buyers' behavior that might affect the demand for technology investments,
uncertainty in buyers' investment horizons, and the rate at which truck owners' trade off higher
vehicle purchase price against future fuel savings.
Where possible, we identify the uncertain aspects of these economic impacts and attempt
to quantify them (e.g., sensitivity ranges associated with quantified and monetized GHG impacts;
range of dollar-per-ton values to monetize non-GHG health benefits; uncertainty with respect to
learning and markups). The agencies have examined the sensitivity of our estimates of savings
in fuel expenditures to alternative assumptions about future fuel prices; results of this sensitivity
analysis can be found in Chapter 8.12 of this RIA. NHTSA's EIS also characterizes the
uncertainty in economic impacts associated with the HD national program. For other impacts,
however, there is inadequate information to inform a thorough, quantitative assessment of
uncertainty. EPA and NHTSA continue to work toward developing a comprehensive strategy for
characterizing the aggregate impact of uncertainty in key elements of its analyses and we will
continue to work to refine these uncertainty analyses in the future as time and resources permit.
This and other chapters of the RIA address section 317 of the Clean Air Act on economic
assessment of standards implementing section 202 of the Act. Chapter 8.11 addresses section
321 of the Clean Air Act on evaluation of potential loss of shifts of employment. The total
monetized benefits and costs of the program are summarized in Chapter 8.10 for the final
program and in Chapter 11 for all alternatives.
8.2 Conceptual Framework for Evaluating Impacts
The HD Phase 2 standards will implement both the 2007 Energy Independence and
Security Act requirement that NHTSA establish fuel efficiency standards for medium- and
heavy-duty vehicles and the Clean Air Act requirement that EPA adopt technology-based
standards to control pollutant emissions from new motor vehicles and engines contributing to air
the consumer's point of reference is her welfare before the change (compensating variation) or after the change
(equivalent variation). In practice, these two measures are typically very close together.
c Indeed, it is likely to be an overestimate of the loss to the consumer, because the buyer has choices other than
buying the same vehicle with a higher price; she could choose a different vehicle, or decide not to buy a new
vehicle. The buyer would choose one of those options only if the alternative involves less loss than paying the
higher price. Thus, the increase in price that the buyer faces would be the upper bound of loss of consumer welfare,
unless there are other changes to the vehicle due to the fuel efficiency improvements that make the vehicle less
desirable to consumers.

-------
pollution that endangers public health and welfare. NHTSA's statutory mandate is intended to
further the agency's long-standing goals of reducing U.S. consumption and imports of petroleum
energy to improve the nation's energy security.
From an economics perspective, government actions to improve our nation's energy
security and to protect our nation from the potential threats of climate change address
"externalities," or economic consequences of decisions by individuals and businesses that extend
beyond those who make these decisions. For example, users of transportation fuels increase the
entire U.S. economy's risk of having to make costly adjustments due to rapid increases in oil
prices, but these users generally do not consider such costs when they decide to consume more
fuel.
Similarly, consuming transportation fuel also increases emissions of greenhouse gases
and other more localized air pollutants that occur when fuel is refined, distributed, and
consumed. Some of these emissions increase the likelihood and severity of potential climate-
related economic damages, and others cause economic damages by adversely affecting human
health. The need to address these external costs and other adverse effects provides a well-
established economic rationale that supports the statutory direction given to government agencies
to establish regulatory programs that reduce the magnitude of these adverse effects at reasonable
costs.
The Phase 2 standards will require manufacturers of new heavy-duty vehicles, including
trailers (HDVs), to improve the fuel efficiency of the products that they produce. As HDV users
purchase and operate these new vehicles, they would consume significantly less fuel, in turn
reducing U.S. petroleum consumption and imports as well as emissions of GHGs and other air
pollutants. Thus as a consequence of the agencies' efforts to meet NHTSA statutory obligations
to improve U.S. energy security and EPA's obligation to issue standards "to regulate emissions
of the deleterious pollutant... from motor vehicles" that endangers public health and welfare,2
the fuel efficiency and GHG emission standards would also reduce HDV operators' outlays for
fuel purchases. These fuel savings are one measure of the final rule's effectiveness in promoting
NHTSA's statutory goal of conserving energy, as well as EPA's obligation under section 202 (a)
(1) and (2) of the Clean Air Act to assess the cost of standards. Although these savings are not
the agencies' primary motivation for adopting higher fuel efficiency standards, these substantial
fuel savings represent significant additional economic benefits of these rules.
Potential savings in fuel costs appear to offer HDV buyers' strong incentives to pay
higher prices for vehicles that feature technology or equipment that reduces fuel consumption.
These potential savings also appear to offer HDV manufacturers similarly strong incentives to
produce more fuel-efficient vehicles. Economic theory suggests that interactions between
vehicle buyers and sellers in a normally-functioning competitive market would lead HDV
manufacturers to incorporate all technologies that contribute to lower net costs into the vehicles
they offer, and buyers to purchase them willingly. Nevertheless, many readily available
technologies that appear to offer cost-effective increases in HDV fuel efficiency (when evaluated
over their expected lifetimes using conventional discount rates) have not been widely adopted,
despite their potential to repay buyers' initial investments rapidly.

-------
This economic situation is commonly known as the "energy efficiency gap" or "energy
paradox." This situation is perhaps more challenging to understand with respect to the heavy-
duty sector versus the light-duty vehicle sector. Unlike light-duty vehicles - which are
purchased and used mainly by individuals and households - the vast majority of HDVs are
purchased and operated by profit-seeking businesses for which fuel costs represent a substantial
operating expense. Nevertheless, on the basis of evidence reviewed below, the agencies believe
that a significant number of fuel efficiency improving technologies will remain far less widely
adopted in the absence of these standards. The economic analysis of these standards is based in
the engineering analysis of the costs and effectiveness of the technologies. The agencies have
detailed their findings on costs and effectiveness in Preamble Sections III, IV, V, and VI, and
RIA Chapter 2. If these cost and effectiveness estimates are correct, and if the agencies have not
omitted key costs or benefits, then the efficiency gap exists, even if it seems implausible.
Explaining why the gap exists is a separate and difficult challenge from observing the existence
of the gap, because of the difficulties involved in developing tests of the different possible
explanations. As discussed below, there is very little empirical evidence on behaviors that might
lead to the gap, even while there continues to be substantial evidence, via the cost and
effectiveness analysis, of the gap's existence.
Economic research offers several possible explanations for why the prospect of these
apparent savings might not lead HDV manufacturers and buyers to adopt technologies that
would be expected to reduce HDV operating costs. Some of these explanations involve failures
of the HDV market for reasons other than the externalities caused by producing and consuming
fuel. Examples include situations where information about the performance of fuel economy
technologies is incomplete, costly to obtain, or available only to one party to a transaction (or
"asymmetrical"), as well as behavioral rigidities in either the HDV manufacturing or HDV-
operating industries, such as standardized or inflexibly administered operating procedures, or
requirements of other regulations on HDVs. Examples that do not involve market failures
include possible effects on the performance, reliability, carrying capacity, maintenance
requirements of new technology under the demands of everyday use, or transactions or
adjustment costs. We note again that these and other hypotheses are presented as potential
explanations of the finding of an efficiency gap based on an engineering analysis. They are not
themselves the basis for regulation.
In the HD Phase 1 rulemaking (which, in contrast to these standards, did not apply to
trailers), and in the Phase 2 NPRM, the agencies raised various hypotheses that might explain
this energy efficiency gap or paradox:
• Imperfect information in the new vehicle market: information available to prospective
buyers about the effectiveness of some fuel-saving technologies for new vehicles may
be inadequate or unreliable. If reliable information on their effectiveness in reducing
fuel consumption is unavailable or difficult to obtain, HDV buyers will
understandably be reluctant to pay higher prices to purchase vehicles equipped with
unproven technologies.
As discussed in the NPRM, one common theme from recent research3 is the inability of
HDV buyers to obtain reliable information about the fuel savings, reliability, and maintenance
costs of technologies that improve fuel efficiency. See 80 FR 40436. In the trucking industry, the

-------
performance of fuel-saving technology is likely to depend on many firm-specific attributes,
including the intensity of HDV use, the typical distance and routing of HDV trips, driver
characteristics, road conditions, regional geography and traffic patterns. As a result, businesses
that operate HDVs have strong preferences for testing fuel-saving technologies "in-house"
because they are concerned that their patterns of vehicle use may lead to different results from
those reported in published information. Businesses with less capability to do in-house testing
often seek information from peers, yet often remain skeptical of its applicability due to
differences in the nature of their operations.
•	Imperfect information in the resale market: buyers in the used vehicle market may not
be willing to pay adequate premiums for more fuel efficient vehicles when they are
offered for resale to ensure that buyers of new vehicles can recover the remaining
value of their original investment in higher fuel efficiency. The prospect of an
inadequate return on their original owners' investments in higher fuel efficiency may
contribute to the short payback periods that buyers of new vehicles appear to
demand.4
The recent research cited above (Klemick et al. 2015, Roeth et al. 2013, Aarnink et al. 2012)
found mixed evidence for imperfect information in the market for used HDVs. On the one hand,
some studies noted that fuel-saving technology is often not appreciated in the used vehicle
market, because of imperfect information about its benefits, or greater mistrust of its
performance among buyers in the used vehicle market than among buyers of new vehicles.
When buyers of new vehicles considered features that would affect value in the secondary
market, those features were rarely related to fuel economy. In addition, some used-vehicle
buyers might have a larger "knowledge gap" than new-vehicle buyers. In other cases, the lack of
interest might be due to the intended use of the used HDVs, which may not reward the presence
of certain fuel-saving technologies. In other cases, however, fuel-saving technology can lead to
a premium in the used market, as for instance to meet the more stringent requirements for HDVs
operating in California.
•	Principal-agent problems causing split incentives: an HDV buyer may not be directly
responsible for its future fuel costs, or the individual who will be responsible for fuel
costs may not participate in the HDV purchase decision. In these cases, the signal to
invest in higher fuel efficiency normally provided by savings in fuel costs may not be
transmitted effectively to HDV buyers, and the incentives of HDV buyers and fuel
buyers will diverge, or be "split." The trailers towed by heavy-duty tractors, which
are typically not supplied by the tractor manufacturer or seller, present an obvious
potential situation of split incentives that was not addressed in the HD Phase 1
rulemaking, but it may apply in this rulemaking. If there is inadequate pass-through
of price signals from trailer users to their buyers, then low adoption of fuel-saving
technologies may result.
All of the recent research identifies split incentives, or principal-agent problems, as a
potential barrier to technology adoption. Vernon and Meier (2012) estimate that 23 percent of
trailers may be exposed to split incentives due to businesses that own and lease trailers to HDV
operators not having an incentive to invest in trailer-specific fuel-saving technology.5 They also
estimate that 5 percent of HDV fuel use is subject to split incentives that arise when the firm

-------
paying fuel costs does not make the tractor investment decision (e.g., because a carrier
subcontracts to an owner-operator but still pays for fuel). They do not quantify the financial
significance of these problems.
Klemick et al. (2015), Aarnink et al. (2012), and Roeth et al. (2013) provide mixed
evidence on the severity of the split-incentive problem. Focus groups often identify diverging
incentives between drivers and the decision-makers responsible for purchasing vehicles.
Aarnink et al. (2012) and Roeth et al. (2013) cite examples of split incentives involving trailers
and fuel surcharges, although the latter also cites other examples where these same issues do not
lead to split incentives. In an effort to minimize problems that can arise form split incentives,
many businesses that operate HDVs also train drivers in the use of specific technologies or to
modify their driving behavior in order to improve fuel efficiency, while some also offer financial
incentives to their drivers to conserve fuel. All of these options can help to reduce the split
incentive problem, although they may not be effective where it arises from different ownership
of combination tractors and trailers.
•	Uncertainty about future fuel cost savings: HDV buyers may be uncertain about
future fuel prices, or about maintenance costs and reliability of some fuel efficiency
technologies. In contrast, the costs of fuel-saving technologies are immediate. If
buyers are loss-averse, they may react to this uncertainty by underinvesting in
technologies to improve fuel economy. In contrast, the costs of fuel-saving or
maintenance-reducing technologies are immediate and thus not subject to
discounting. In this situation, potential variability about buyers' expected returns on
capital investments to achieve higher fuel efficiency may shorten the payback period
- the time required to repay those investments - they demand in order to make them.
Questions related to uncertainty about future costs for fuel and maintenance, as well as
about the reliability of new technology that could result in costly downtime, illustrate the
problem of uncertain or unreliable information about the actual performance of fuel efficiency
technology discussed above. Roeth et al. (2013) and Klemick et al. (2015) both document the
short payback periods that HDV buyers require on their investments — usually about 2 years —
which may be partly attributable to these uncertainties.
•	Adjustment and transactions costs: potential resistance to new technologies -
stemming, for example, from drivers' reluctance or slowness to adjust to changes in
the way vehicles operate - may slow or inhibit new technology adoption. If a
conservative approach to new technologies leads HDV buyers to adopt them slowly,
then successful new technologies will be adopted over time without market
intervention, but only with potentially significant delays in achieving the fuel saving,
environmental, and energy security benefits they offer. There also may be costs
associated with training drivers to realize potential fuel savings enabled by new
technologies, or with accelerating fleet operators' scheduled fleet turnover and
replacement to hasten their acquisition of vehicles equipped with these technologies.
These factors might present real resource costs to firms that are not reflected in a
typical engineering analysis.

-------
Klemick et al. (2015), Roeth et al. (2013), and Aarnink et al. (2012) provide some
support for the view that adjustment and transactions costs may impede HDV buyers from
investing in higher fuel efficiency. These studies note that HDV buyers are less likely to select
new technology when it is not available from their preferred manufacturers. Some technologies
are only available as after-market additions, which can add other costs to adopting them.
•	Driver acceptance of new equipment or technologies as a barrier to their adoption.
HDV driver turnover is high in the U.S., and businesses that operate HDVs are
concerned about retaining their best drivers. Therefore, they may avoid technologies
that require significant new training or adjustments in driver behavior.
For some technologies that can be used to meet these standards, such as automatic tire
inflation systems, training costs are likely to be minimal. Other technologies, such as stop-start
systems, may require drivers to adjust their expectations about vehicle operation, and it is
difficult for the agencies to anticipate how drivers will respond to such changes.0
•	Constraints on access to capital for investment. If buyers of new vehicles have
limited funds available, then they must choose between investing in fuel-saving
technology and other vehicle technologies or attributes.
There would be tradeoffs if capital markets are constrained, and fuel-saving technologies
do not provide returns sufficient to achieve the hurdle rates that the buyers require. Klemick et
al. (2015) did not find capital constraints to be a problem for the medium- and large-sized
businesses participating in Klemick et al.'s (2013) study. On the other hand, Roeth et al. (2013)
noted that access to capital can be a significant challenge to smaller or independent businesses,
and that price is always a concern to buyers.
•	"Network externalities," where the benefits to new users of a technology depend on
how many others have already adopted it. If the value of a technology increases with
increasing adoption, then it can be difficult for the adoption process to begin: each
potential adopter has an incentive to wait for others to adopt before making the
investment. If all adopters wait for others, then adoption may not happen.
One example where network externalities seem likely to arise is the market for natural
gas-fueled HDVs: the limited availability of refueling stations may reduce potential buyers'
willingness to purchase natural gas-fueled HDVs, while the small number of such HDVs in use
does not provide sufficient economic incentive to construct more natural gas refueling stations.
Some businesses that operate HDVs may also be concerned about the difficulty in locating repair
facilities or replacement parts, such as single-wide tires, wherever their vehicles operate. When
a technology has been widely adopted, then it is likely to be serviceable even in remote or rural
places, but until it becomes widely available, its early adopters may face difficulties with repairs
D The distinction between simply requiring drivers (or mechanics) to adjust their expectations and compromises in
vehicle performance or utility is subtle. While the former may not impose significant compliance costs in the long
run, the latter would represent additional economic costs of complying with the standard.

-------
or replacements. By accelerating the widespread adoption of these technologies, the standards
may assist in overcoming these difficulties.
• First-mover disadvantage. Many manufacturers prefer to observe the market and
follow other manufacturers rather than be the first to market with a specific
technology. The "first-mover disadvantage" has been recognized in other research
where the "first-mover" pays a higher proportion of the costs of developing
technology, but loses the long-term advantage when other businesses follow
quickly.6'E In this way, there may be barriers to innovation on the supply side that
result in lower adoption rates of fuel-efficiency technology than would be optimal.
Roeth et al. (2013) noted that HDV buyers often prefer to have technology or equipment
installed by their favored original equipment manufacturers. However, some technologies may
not be available through these preferred sources, or may be available only as after-market
installations from third parties (Aarnink et al. 2012, Roeth et al. 2013). Manufacturers may be
hesitant to offer technologies for which there is not strong demand, especially if the technologies
require significant research and development expenses and other costs of bringing the technology
to a market of uncertain demand. Roeth et al. (2013) noted that it can take years, and sometimes
as much as a decade, for a specific technology to become available from all manufacturers.
In summary, the agencies recognize that businesses that operate HDVs are under
competitive pressure to reduce operating costs, which should compel HDV buyers to identify and
rapidly adopt cost-effective fuel-saving technologies. Outlays for labor and fuel generally
constitute the two largest shares of HDV operating costs, depending on the price of fuel, distance
traveled, type of HDV, and commodity transported (if any), so businesses that operate HDVs
face strong incentives to reduce these costs.7'8
However, the relatively short payback periods that buyers of new HDVs appear to require
suggest that some combination of the factors cited above impedes this process. Markets for both
new and used HDVs may face these problems, although it is difficult to assess empirically the
degree to which they actually do. Even if the benefits from widespread adoption of fuel-saving
technologies exceed their costs, their use may remain limited or spread slowly because their
early adopters bear a disproportionate share of those costs. In this case, the standards may help
to overcome such barriers by ensuring that these measures will be widely adopted.
Providing information about fuel-saving technologies, offering incentives for their
adoption, and sharing HDV operators' real-world experiences with their performance through
voluntary programs such as EPA's SmartWay Transport Partnership should assist in the adoption
of new cost-saving technologies. Nevertheless, other barriers that impede the diffusion of new
technologies are likely to remain. Buyers who are willing to experiment with new technologies
expect to find cost savings, but those savings may be difficult to verify or replicate. As noted
previously, because benefits from employing these technologies are likely to vary with the
characteristics of individual routes and traffic patterns, buyers of new HDVs may find it difficult
E This first-mover disadvantage must be large enough to overcome the potential incentive for first movers to earn

-------
to identify or verify the effects of fuel-saving technologies in their operations. Risk-averse
buyers may also avoid new technologies out of concerns over the possibility of inadequate
returns on their investments, or with other possible adverse impacts.
Competitive pressures in the HDV freight transport industry can provide a strong
incentive to reduce fuel consumption and improve environmental performance. Nevertheless,
HDV manufacturers may delay in investing in the development and production of new
technologies, instead waiting for other manufacturers to bear the initial risks of those
investments. In addition, not every HDV operator has the requisite ability or interest to access
and utilize the technical information, or the resources necessary to evaluate this information
within the context of his or her own operations.
As discussed previously, whether the technologies available to improve HDVs' fuel
efficiency would be adopted widely in the absence of the program is challenging to assess. To
the extent that these technologies would be adopted in its absence, neither their costs nor their
benefits should be attributed to the program.
The agencies will continue to explore reasons for the slow adoption of readily available
and apparently cost-effective technologies for improving fuel efficiency.
8.3 Analysis of the Rebound Effect
The "rebound effect" has been defined in a variety of different ways in the energy policy
and economics literature. One common definition states that the rebound effect is the increase in
demand for an energy service when the cost of the energy service is reduced due to efficiency
improvements.9'10'11 In the context of heavy-duty vehicles (HDVs), this can be interpreted as an
increase in HDV fuel consumption resulting from more intensive vehicle use in response to
increased vehicle fuel efficiency.17 Although much of this vehicle use increase is likely to take
the form of increases in the number of miles vehicles are driven, it can also take the form of
increases in the loaded weight at which vehicles operate or changes in traffic and road conditions
vehicles encounter as operators alter their routes and schedules in response to improved fuel
efficiency. Because this more intensive use consumes fuel and generates emissions, it reduces
the fuel savings and avoided emissions that would otherwise be expected to result from the
increases in fuel efficiency in this rulemaking.
Unlike the light-duty vehicle (LDV) rebound effect, the HDV rebound effect has not
been extensively studied. According to a 2010 HDV report published by the National Research
Council of the National Academies (NRC)12, it is "not possible to provide a confident measure of
the rebound effect," yet NRC concluded that a HDV rebound effect probably exists and that,
"estimates of fuel savings from regulatory standards will be somewhat misestimated if the
rebound effect is not considered." Although we believe the HDV rebound effect needs to be
F We discuss other potential rebound effects in Section 8.3.3.2, such as the indirect and economy-wide rebound
effects. Note also that there is more than one way to measure HDV energy services and vehicle use. The agencies'
analyses use VMT as a measure (as discussed below); other potential measures include ton-miles, cube-miles, and
fuel consumption.

-------
studied in more detail, we have nevertheless attempted to capture its potential effect in our
analysis of these final rules, rather than to await further study. We have elected to do so because
the magnitude of the rebound effect is an important determinant of the actual fuel savings and
emission reductions that are likely to result from adopting stricter fuel efficiency and GHG
emission standards.
In our analysis and discussion below, we focus on one widely-used metric to estimate the
rebound effect associated with all types of more intensive vehicle use, the increase in vehicle
miles traveled (VMT) that results from improved fuel efficiency. VMT can often provide a
reasonable approximation for all types of more intensive vehicle use. For simplicity, we refer to
this as "the VMT rebound effect" or "the direct VMT rebound" throughout this section, although
we acknowledge that it is an approximation to the rebound effect associated with all types of
more intensive vehicle use. The agencies use our VMT rebound estimates to generate VMT
inputs that are then entered into the EPA MOVES national emissions inventory model and the
Volpe Center's HD CAFE model. Both of these models use these inputs along with many others
to generate projected emissions and fuel consumption changes resulting from each of the
regulatory alternatives analyzed.
Using VMT rebound to approximate the fuel consumption impact from all types of more
intensive vehicle use may not be completely accurate. Many factors other than distance traveled
- for example, a vehicle's loaded weight - play a role in determining its fuel consumption, so it
is also important to consider how changes in these factors are correlated with variation in vehicle
miles traveled. Empirical estimates of the effect of weight on HDV fuel consumption vary, but
universally show that loaded weight has some effect on fuel consumption that is independent of
distance traveled. Therefore, the product of vehicle payload and miles traveled, which typically
is expressed in units of "ton-miles" or "ton-kilometers," has also been considered as a metric to
approximate the rebound effect. Because this metric's value depends on both payload and
distance, it is important to note that changes in these two variables can have different impacts on
HDV fuel consumption. This is because the fuel consumed by HDV freight transport is
determined by several vehicle attributes including engine and accessory efficiencies,
aerodynamic characteristics, tire rolling resistance and total vehicle mass—including payload
carried, if any. Other factors such as vehicle route and traffic patterns can also affect how each
of these vehicle attributes contributes to the overall fuel consumption of a vehicle. While it
seems intuitive that if all of these other conditions remain constant, a vehicle driving the same
route and distance twice will consume twice as much fuel as driving that same route once.
However, because of the other vehicle attributes, it is less intuitive how a change in vehicle
payload would affect vehicle fuel consumption.
Because the factors influencing HDV VMT rebound are generally different from those
affecting LDV VMT rebound, much of the research on the LDV sector is likely to not apply to
the HDV sector. For example, the owners and operators of LDVs may respond to the costs and
benefits associated with changes in their personal vehicle's fuel efficiency very differently than a
HDV fleet owner or operator would view the costs and benefits (e.g., profits, offering more
competitive prices for services) associated with changes in their HDVs' fuel efficiency. To the
extent the response differs, such differences may be smaller for HD pickups and vans, which
share some similarities with LDVs. As discussed in the 2010 NRC HD report, one difference
from the LDV case is that when calculating the change in HDV costs that causes the rebound

-------
effect, it is more important to consider all components of HDV operating costs. The costs of
labor and fuel generally constitute the two largest shares of HDV operating costs, depending on
the price of petroleum, distance traveled, type of vehicle, and commodity transported (if
any).1314 Equipment depreciation costs associated with the purchase or lease of an HDV are
another significant component of total operating costs (Figure 8-1).15 Even when HDV
purchases involve upfront, one-time payments, HDV operators must recover the depreciation in
the value of their vehicles resulting from their use, so this is likely to be considered as an
operating cost they will attempt to pass on to final consumers of HDV operator services.
Reference Case Total Truck Operation Costs Per Mile
Source: ATRI, 2015
Truck Insurance Premium,
$0.071, 4%
Permits, License,
Tolls, $0,042,2%
Repair, Maintenance,Tires
$0,202 ,12%
Fuel Cost, $0,583 , 34%
Truck/Trailer Lease or Purchase
Payments, $0.215,13%
Driver Benefit,
$0,129
Driver Wage, $0,462 , 27%
i Fuel Cost
¦	Driver Wage
Driver Benefit
i Truck/Trailer Lease or Purchase Payments
3 Repair, Maintenance, Tires
«Truck Insurance Premium
¦	Permits, License, Tolls
Total Cost Per Mile $1,703
Figure 8-1 Average Truck Operation Costs
Estimates of the impact of fuel efficiency standards on HDV VMT, and hence fuel
consumption, should account for changes in all of these components of HDV operating costs.
The higher the net savings in total operating costs is, the higher the expected rebound effect
would be. Conversely, if higher HDV purchase costs and any other ancillary costs of adopting
more fuel-efficiency vehicles outweigh future cost savings, and total operating costs increase,
HDV costs could rise, which would likely result in a decrease in HDV VMT. In theory, other
cost changes resulting from any requirement to achieve higher fuel efficiency, such as changes in
maintenance costs or insurance rates, should also be taken into account, although information on
these elements of HDV operating costs is extremely limited. In this analysis, the agencies adapt

-------
estimates of the VMT rebound effect to project the response of HDV use to the estimated
changes in total operating costs that result from the Phase 2 standards.
Since businesses are profit-driven, one would expect their decisions to be based on the
costs and benefits of different operating decisions, both in the near-term and long-term.
Specifically, one would expect commercial HDV operators to take into account changes in
overall operating costs per mile when making decisions about HDV use and setting rates they
charge for their services. If demand for those services is sensitive to the rates HDV operators
charge, HDV VMT could change in response to the effect of higher fuel efficiency on the rates
HDV operators charge. If demand for HDV services is insensitive to price (e.g., due to lack of
good substitutes), however, or if changes in HDV operating costs due to the standards are not
passed on to final consumers of HDV operator services, the standards may have a limited impact
on HDV VMT.
The following sections describe the factors affecting the magnitude of HDV VMT
rebound; review the econometric and other evidence related to HDV VMT rebound; and
summarize how we estimated the HDV rebound effect for this rulemaking.
8.3.1 Factors Affecting the Magnitude of HDV VMT Rebound
The magnitude and timing of HDV VMT rebound result from the interaction of many
different factors.16 Fuel savings resulting from fuel efficiency standards may cause HDV
operators and their customers to change their patterns of HDV use and fuel consumption in a
variety of ways. As discussed later in this section and in the proposal to this rule, HDV VMT
rebound estimates determined via other proxy elasticities vary, but in no case has there been an
estimate that fully offsets the fuel saved due to efficiency improvements (i.e., no direct rebound
effect greater than or equal to 100 percent). G
If fuel cost savings are passed on to the HDV operators' customers (e.g., logistics
businesses, manufacturers, retailers, municipalities, utilities consumers), those customers might
reorganize their logistics and distribution networks over time to take advantage of lower
operating costs. For example, customers might order more frequent shipments or choose
products that entail longer shipping distances, while freight carriers might divert some shipments
to trucks from other shipping modes such as rail, barge or air. In addition, customers might
choose to reduce their number of warehouses, reduce shipment rates or make smaller but more
frequent shipments, all of which could lead to an increase in HDV VMT. Ultimately, fuel cost
savings could ripple through the entire economy, thus increasing demand for goods and services
shipped by trucks, and therefore increase HDV VMT due to increased gross domestic product
(GDP).
G Elasticity is the measurement of how responsive an economic variable is to a change in another. For example:
price elasticity of demand is a measure used in economics to show the responsiveness, or elasticity, of the quantity
demanded of a good or service to a change in its price. More precisely, it gives the percentage change in quantity
demanded in response to a one percent change in price.

-------
Conversely, if fuel efficiency standards lead to net increases in the total costs of HDV
operation because fuel cost savings do not fully offset the increase in HDV purchase prices and
associated depreciation costs, then the price of HDV services could rise. This is likely to spur a
decrease in HDV VMT, and perhaps a shift to alternative shipping modes. These effects could
also ripple through the economy and affect GDP. Note, however, that we project fuel cost
savings will offset technology costs in our analysis supporting the final standards.
It is also important to note that any increase in VMT on HDVs impacted by the final
standards may be offset, to some extent, by a decrease in VMT on older HDVs. This may occur
if lower fuel costs resulting from our standards cause multi-vehicle fleet operators to shift VMT
to newer, more efficient HDVs in their fleet or cause operators with newer, more efficient HDVs
to be more successful at winning contracts than operators with older HDVs.
Also, as discussed in Chapter 8.2 of this RIA, the magnitude of the rebound effect is
likely to be influenced by the extent of any market failures that affect the demand for more fuel
efficient HDVs, as well as by HDV operators' responses to their perception of the tradeoff
between higher upfront HDV purchase costs versus lower but uncertain future expenditures on
fuel.
8.3,2 Recent Econometric and Other Evidence Related to HDV VMT Rebound
As discussed above, HDV VMT rebound is defined as the change in HDV VMT that
occurs in response to an increase in HDV fuel efficiency. We are not aware of any studies that
directly estimate this elasticity for the U.S. In the RIA that accompanied the proposal, we
discussed a number of econometric analyses of other related elasticities that could potentially be
used as a proxy for measuring HDV VMT rebound, as well as several other analyses that may
provide insight into the magnitude of HDV VMT rebound.17 These studies produced a wide
range of estimates for HDV VMT rebound however, and we were unable to draw any strong
conclusions about the magnitude of rebound based on this available literature.
We also discussed several challenges that researchers face in attempting to quantify the
VMT rebound effect for HDVs,18 including limited data on the HD sector and the difficulty of
specifying mathematical models that reflect the complex set of factors that influence HD VMT.
Given these limitations, the agencies requested comment on a number of aspects of the proposed
VMT rebound analysis, including procedures for measuring the rebound effect and the studies
discussed in the proposal. The agencies also committed to reviewing and considering revisions
to VMT rebound estimates for the final rule based on submissions from public commenters and
new research on the rebound effect. This section reviews new econometric analyses that have
been produced since the release of the proposal. All of these analyses study the change in HDV
use (measured in VMT, ton-mile, or fuel consumption) in response to changes in fuel price
($/gallon) or fuel cost ($/mile or $/ton-mile). The studies presented below attempt to estimate
these elasticities in the HDV sector using varying approaches and data sources.
During the same period that agencies were developing the proposal for this rulemaking,
EPA contracted with Energy and Environmental Research Associates (EERA) to analyze the
HDV rebound effect for regulatory assessment purposes. Excerpts of EERA's initial report to
EPA are included in the NPRM docket and contain detailed qualitative discussions of the

-------
rebound effect as well as data sources that could be used in quantitative analysis.19 EERA also
conducted follow-on quantitative analyses focused on estimating the impact of fuel prices on
VMT and fuel consumption. We included a Working Paper in the NPRM docket that described
much of this work.20 Note that EERA's Working Paper was not available at the time the
agencies conducted the analysis of the rebound effect for the proposal, but the agencies agreed to
consider this work, and any other work, in the final rule.
At the time of the filing of the NPRM, Winebrake et al. (2015) published two papers in
Transportation Research Part D: Transport and Environment based on EERA work mentioned
above.21 These two papers have been filed in the NPRM docket and received public review and
comment. In the first paper, the fuel price elasticities of VMT and fuel consumption for
combination trucks are estimated with regression models. The combination trucks paper uses
annual data for the period 1970-2012. VMT and fuel consumption are used as the dependent
variables. The control variables include: a macroeconomic variable (e.g., gross domestic product
(GDP)), imports/exports, and fuel price, among other variables. In the second paper, the fuel
price elasticity of VMT for single unit vehicles is estimated by using annual data for the period
1980-2012. The single unit vehicle paper uses similar control variables but includes additional
variables related to lane miles and housing construction. VMT is the only dependent variable
modeled in the single unit vehicle paper (i.e., fuel consumption is not modeled).
The results in Winebrake et al. are that the null hypothesis - which states that the fuel
price elasticity of VMT and the fuel price elasticity of fuel consumption are zero - cannot be
rejected with statistical confidence. The papers hypothesize that low elasticities may be due to a
range of possibilities including: (1) the common use of fuel surcharges; (2) adjustments in other
operational costs, such as labor; (3) possible principal-agent problems affecting driver behavior;
and (4) the nature of freight transportation as an input to a larger supply chain system that is
driven by other factors. These two papers suggest that previous regulatory analysis that uses a
five percent rebound effect for combination trucks and a 15 percent rebound effect for single unit
trucks may be overestimating the direct VMT rebound effect.
To the best of our knowledge, the Winebrake et al. paper represents the first peer-
reviewed work in the last two decades, after Gately (1990)H that attempts to estimate
quantitatively the impact of a change in fuel costs on HDV VMT in the U.S. context. A
subsequent paper by Wadud, discussed in more detail below, states that there is "only one
creditable study" on "the responses of different [heavy duty] vehicle sectors to fuel price or
income changes," specifically the Winebrake et al. combination truck work.
However, there is other recent work that has not been peer reviewed, or that studies HDV
VMT rebound in other countries, that bears mentioning as well. Resources for the Future (RFF)
filed a comment on the proposal with a Working Paper by Leard et al. (2015) to address HDV
rebound effects during the comment period of NPRM.22 Leard et al.'s Working Paper uses
detailed truck-level micro-data from Vehicle Use and Inventory Survey (VIUS) for six survey
years (specifically, 1977, 1982, 1987, 1992, 1997, and 2002). The "rebound effect" in this paper
H Gately, D., 1990. The U.S. demand for highway travel and motor fuel. Energy Journal. 11, p. 59-74.

-------
is defined to be a combination of a "VMT elasticity with respect to fuel costs per mile" ($/mile);
and a "truck count elasticity with respect to fuel costs per mile." Fuel costs per mile are defined
as fuel price ($/gal) divided by efficiency (mpg). Because the agencies do not estimate the
directional impact of this rulemaking on vehicle sales, the portion of Leard et al.'s estimates
associated with VMT rebound with respect to fuel costs per mile are the most useful point of
comparison to our estimates in the proposal.
Leard et al. report a VMT rebound effect result of 18.5 percent with respect to fuel costs
per mile for combination trucks.1 This finding suggests that previous estimates of combination
truck rebound effects used in this proposal, a five percent rebound effect, may be
underestimating the true rebound effect. Leard et al. also report a VMT rebound effect with
respect to fuel costs per mile of 12.2 percent for single unit trucks/ This finding suggests that
the previous use of a 15 percent rebound effect for single unit vehicles in the proposed rule may
be overestimating the true rebound effect. As noted, VIUS was discontinued in 2002, so the
most recent data in this study is 2002 which is fourteen years old. In addition, as noted, Leard et
al. Working Paper has not been peer reviewed or published.
Recently, Wadud (2016) has estimated price elasticities of diesel demand in the U.K.k
The paper aims to model diesel demand elasticity for different freight duty vehicle types in the
U.K. Wadud uses a similar model specification as Winebrake et al. in the regression analysis.
Wadud finds that diesel consumption in freight vehicles overall is quite inelastic. Diesel demand
from articulated trucks and light goods vehicles (similar to combination trucks in the U.S.) does
not respond to changes in diesel prices at all. Demand in rigid trucks (similar to single unit
trucks in the U.S.) responds to fuel price changes with a 15 percent elasticity. Wadud's work
presents empirical results in the U.K., which might not be appropriate to apply to the U.S.
8.3.3 How the Agencies Estimated the HDV Rebound Effect for this Final
Rulemaking
8.3.3.1 Values Used in the Phase 2 NPRM Analysis
At the time the agencies conducted their analysis of the Proposed Phase 2 fuel efficiency
and GHG emissions standards, the agencies determined that the evidence did not lend itself to
any changes in the values used to estimate the VMT rebound effect in the HD Phase 1
rulemaking. The agencies used the rebound effect estimates of 15 percent for vocational
vehicles, five percent for combination tractors, and 10 percent for HD pickup trucks and vans
from the HD Phase 1 rulemaking
1 Leard et al. report a total VMT rebound effect result of 29.7 percent for combination trucks, which is a sum of
separate estimates associated with both VMT elasticity and truck count elasticity with respect to fuel costs per mile.
1 For vocational trucks, Leard et al. report an overall 9.3 percent rebound value, which is a sum of separate estimates
associated with both VMT elasticity and truck count elasticity with respect to fuel costs per mile.
K Wadud, Zia, Diesel Demand in the Road Freight Sector in the UK: Estimates for Different Vehicle Types, Applied
Energy 165 (2016), p. 849-857.

-------
8.3.3.2 How the Agencies Analyzed VMT Rebound in this Final Rulemaking
The emergence of new information as well as the public comments are cause for updating
the quantitative values used to estimate the VMT rebound effect from those estimated by the
analysis conducted for the HDV Phase 1 rulemaking. For vocational trucks, the Winebrake et al.
study found no responsiveness of truck travel to diesel fuel prices, suggesting a VMT rebound
effect of essentially zero. Leard et al. suggested a VMT rebound effect for vocational trucks of
roughly 12 percent. For combination trucks, the Winebrake et al. study found a rebound effect
of essentially zero percent. The Leard et al. study found a VMT elasticity rebound effect of
roughly 18 percent for combination trucks. In addition to the RFF comments to which Leard et
al. was included, EPA and NHTSA received ten other comments on HDV rebound during the
comment period for the proposal, six of which were substantive. One of these commenters
suggested that the agencies' rebound numbers "appear reasonable." The five others commented
that the rebound estimates for both combination and vocational vehicles used in the proposal
were overestimated, and suggested using the Winebrake et al. estimates.
In revising the HD VMT rebound estimates, we give somewhat greater consideration to
the findings of Winebrake et al. because it is peer-reviewed and published, whereas Leard et al.
is a Working Paper. Based on this consideration and on the comments that we received in
response to the proposal, the agencies have chosen to revise the VMT rebound estimate for
vocational trucks down to five percent, and have elected to maintain the use of the five percent
rebound effect for combination tractor-trailers. We note that while the Winebrake et al. work
supports rebound estimates of zero percent for vocational vehicles and combination tractor-
trailers, using a five percent value is conservative and leaves some consideration of uncertainty,
as well as some consideration of the (un-peer reviewed and unpublished) findings of the Leard et
al. study. The five percent value is in range of the two U.S. studies and generally addresses the
issues raised by the commenters.
We did not receive new data or comments on our estimated VMT rebound effect for
heavy-duty pick-up trucks and vans. Therefore, we have elected to use the 10 percent value used
at proposal. It should be noted that the rebound estimates we have selected for our analysis
represent the VMT impact from our final standards with respect to changes in the fuel cost per
mile driven.
As described previously, the HDV rebound effect should ideally be a measure of the
change in fuel consumed with respect to the change in overall operating costs due to a change in
HDV fuel efficiency. Such a measure would incorporate all impacts from our rules, including
those from incremental increases in vehicle prices that reflect costs for improving their fuel
efficiency. Therefore, VMT rebound estimates with respect to fuel costs per mile must be
"scaled" to apply to total operating costs, by dividing them by the fraction of total operating
costs accounted for by fuel use.
The agencies scaled the VMT rebound calculations to total operating costs using the most
recent information from the American Transportation Research Institute (ATRI), which has been
updated for this final rulemaking.23 ATRI estimates that the average motor carrier cost per mile
is $1,703 for 2014. Other elements of the total costs are listed below in Table 8-1.

-------
Table 8-1 Elements of the Operating Costs per Mile
OPERATING COST PER MILE
ATRI
Fuel Cost
$0,583
New Vehicle Cost
$0,215
Maintenance & Repair Cost
$0,158
All Other (labor, insurance, etc.)
$0,747
Total Motor Carrier Costs
$1,703
For the final rulemaking, the agencies determined VMT rebound separately for each
HDV category and for each alternative. However, the agencies made simplifying assumptions in
the VMT rebound analysis for this rulemaking, similar to the approach taken in the HD GHG
Phase 1 final rule. Chapter 7 of the RIA presents VMT rebound values for each HDV sector that
we estimated for the final program. These VMT impacts are reflected in the estimates of total
fuel savings and reductions in emissions of GHG and other air pollutants presented in Section VI
and VII of the Preamble for all categories.
For the purposes of this final rulemaking, we made simplifying assumptions when
applying the overall rebound effect to each class of truck. For example, we assumed that per
mile vehicle costs were based on the new vehicle cost (e.g., $175,000 for the reference case
Class 8 combination tractor with three box trailers, $40,000 for the reference case HD pickups,
and $100,000 for the vocational vehicles)24 divided by the total lifetime number of expected
vehicle miles (e.g., 1.12 million miles for a Class 8 combination tractor-trailer, 169,249 miles for
2b/3 trucks, and 203,548 miles for vocational vehicles). 25 We recognize that this calculation
implicitly assumes that truck depreciation is strictly a function of usage, and that it does not take
into account the opportunity cost of alternative uses of capital. As a result, the new vehicle cost
per mile assumptions used in these calculations represent a smaller percentage of total operating
costs compared to the ATRI and CSI examples.
The agencies assumed in this final rulemaking an "average" incremental technology cost
for the alternatives, as shown in Table 8-2. Due to timing constraints, the agencies were not able
to determine the technology costs for the final alternatives prior to conducting the emission
inventory modeling. Therefore, the technology costs for Alternatives lb, 2, 3, and 4 were
assumed to be the values developed in the HD Phase 2 NPRM.26 The agencies did not develop a
technology package cost for Alternative 5 because we believe that there would be such
substantial additional costs related to pulling ahead the development of so many additional
technologies that we cannot accurately predict these costs (even assuming that it is technically
feasible to do so, which is at the least doubtful).27 For the rebound calculation, the technology
package cost of Alternative 5 was assumed to be twice the cost of Alternative 4.

-------
Table 8-2 Technology Costs Used to Determine the Rebound Effect of Each Alternative

ALTERNATIVE

Vehicle
Category
lb
2
Final Program
4
5
Combination
Tractors
$362
$8,358
$12,849
$12,849
$25,698
HD Pickup
& Vans
$15
$714
$1,342
$1,841
$3,682
Vocational
Vehicles
$0
$380
$3,381
$3,382
$6,762
The fuel costs per mile in the analysis were calculated using EIA's Annual Energy
Outlook 2015's projections for diesel fuel price.28 The average fuel economy for each category
was determined using MOVES2014a. The combination tractor-trailer fuel economy used was
6.2 mpg, the vocational vehicle category was 9.8 mpg, and the HD pickup category was 14.5
mpg. The technology effectiveness of the alternatives in the final rules was assumed to be equal
to the technology effectiveness developed for each alternative in the Phase 2 NPRM, as show in
Table 8-3.29
Table 8-3 Technology Effectiveness Used to Determine the Rebound Effect of Each Alternative

AL1
fERNATIVE

Vehicle
Category
lb
2
Final Program
4
5
Combination
Tractors
2.1%
12.1%
20.4%
20.5%
25.5%
HD Pickup
& Vans
2.9%
9.6%
16.2%
16.3%
18.5%
Vocational
Vehicles
0%
3.2%
11.8%
11.9%
17.4%
The operating costs calculated based on all of these inputs are shown below in Table 8-4.

-------
Table 8-4 Operating Costs for the Reference and Final Program
OPERATING COST PER MILE
REFERENCE CASE
FINAL
PROGRAM
Tractor-Trailers
Fuel Cost
$0,586
$0.49
New Vehicle Cost
$0,156
$0.17
Maintenance & Repair Cost
$0,158
$0,158
All Other (labor, insurance, etc.)
$0,747
$0,747
Total Motor Carrier Costs
$1,647
$1,559
Picku
ps and Vans
Fuel Cost
$0,250
$0,220
New Vehicle Cost
$0,236
$0,240
Maintenance & Repair Cost
$0,158
$0,158
All Other (labor, insurance, etc.)
$0,747
$0,747
Total Motor Carrier Costs
$1,392
$1,365
Vocational Vehicles
Fuel Cost
$0.37
$0.33
New Vehicle Cost
$0,491
$0,508
Maintenance & Repair Cost
$0,158
$0,158
All Other (labor, insurance, etc.)
$0,747
$0,747
Total Motor Carrier Costs
$1,767
$1,744
Other simplifying assumptions include the use of an average cost rather than a marginal
cost. Some trucking firms may use a marginal cost to determine whether to increase their fuel
usage, however we do not have any data on when firms might use a marginal cost calculation
rather than an average cost calculation. Although using a marginal cost might be more
appropriate for calculating the rebound effect, we do not have a methodology for calculating the
marginal cost.30
In the costs and benefits summarized in Preamble Section IX.K, we have not explicitly
taken into account any potential fuel savings or GHG emission reductions from the rail, air or
water-borne shipping sectors due to mode shifting because estimates of this effect seem too
speculative at this time. Likewise we have not taken into account any fuel savings or GHG
emissions reductions from the potential shift in VMT from older HDVs to newer, more efficient
HDVs. As discussed in the Preamble at Section IX.E, we have found limited evidence of the
impact of HDV fuel efficiency standards on mode shifting and no evidence on shifting activity
away from older HDVs to newer HDVs. The agencies requested comment on these assumptions
in the NPRM, but did not receive any.
In addition, we have not attempted to capture the extent of how current market failures
might impact the rebound effect. The direction and magnitude of the rebound effect in the HD
truck market are expected to vary depending on the existence and types of market failures
affecting the fuel efficiency of the trucking fleet. If firms are already accurately accounting for
the costs and benefits of these technologies and fuel savings, then these regulations would
increase their net costs, because trucks would already include all cost-effective fuel saving
technologies. As a result, the rebound effect would actually be negative and truck VMT would
decrease as a result of these regulations.

-------
However, if firms are not optimizing their behavior today due to factors such as lack of
reliable information (see Preamble Section IX. A or RIA Chapter 8.2 for further discussion), it is
more likely that truck VMT would increase. If firms recognize their lower net costs as a result of
these regulations and pass those costs along to their customers, then the rebound effect would
increase truck VMT. This response assumes that trucking rates include both truck purchase costs
and fuel costs, and that the truck purchase costs included in the rates spread those costs over the
full expected lifetime of the trucks. If those costs are spread over a shorter period, as the
expected short payback period implies, then those purchase costs will inhibit reduction of freight
rates, and to the extent that they do so the rebound effect will be proportionally smaller.
As discussed in more detail in Preamble Section IX.A and RIA Chapter 8.2, if there are
market failures such as split incentives, estimating the rebound effect may depend on the nature
of the failures. For example, if the original purchaser cannot fully recoup the higher upfront
costs through fuel savings before selling the vehicle nor pass those costs onto the resale buyer,
the firm would be expected to raise shipping rates. A firm purchasing the truck second-hand
might lower shipping rates if the firm recognizes the cost savings after operating the vehicle,
leading to an increase in VMT. Similarly, if there are split incentives and the vehicle buyer is
not the same entity that purchases the fuel, than there would theoretically be a positive rebound
effect. In this scenario, fuel savings would lower the net costs to the fuel purchaser, which
would result in a larger increase in truck VMT.
Note that while we focus on the VMT rebound effect in our analysis of this final rule,
there are at least two other types of rebound effects discussed in the energy policy and economics
literature. In addition to VMT rebound effects, there are "indirect" rebound effects, which refers
to the purchase of other goods or services (that consume energy) with the costs savings from
energy efficiency improvements; and "economy-wide" rebound effects, which refers to the
increased demand for energy throughout the economy in response to the reduced market price of
energy that happens as a result of energy efficiency improvements. One commenter pointed out
that consumers may use their savings from lower fuel costs as a result of the direct rebound
effect to buy more goods and services, which indirectly increases the use of energy (i.e., the
indirect rebound effect).L The commenter states that the indirect rebound effect represents a
positive economic result for consumers, since consumer welfare increases, although it could
result in increased energy use and GHG emissions. We agree with this commenter's observation
that, to the extent that indirect rebound does occur, it could have both positive and negative
impacts.
Another commenter suggested that the indirect or economy-wide rebound effect could be
large enough so as to fully offset the fuel savings and GHG emissions benefits of the rule.M The
commenter provides multiple estimates of the potential size of the indirect rebound effect.
However, the unpublished methodology used to perform these estimates has not undergone peer
review and, as explained in the response to comment document, the agencies find it to be
dubious. Further, as discussed in detail in the proposed rule and our response to comment
L EP A-HQ-0 AR-2014-0827-1336.
M EP A-HQ-0 AR-2014-0827-1467.

-------
document, there are a number of other important questions not addressed by the commenter that
must be examined before we can have enough confidence in these kinds of estimates to include
them in our economic analysis.
As discussed in this proposed rule, all of the fuel costs savings will not necessarily be
passed through to the consumer in terms of cheaper goods and services. First, there may be
market barriers that impede trucking companies from passing along the fuel cost savings from
the rule in the form of lower rates. Second, there are upfront vehicle costs (and potentially
transaction or transition costs associated with the adoption of new technologies) that would
partially offset some of the fuel cost savings from our rule, thereby limiting the magnitude of the
impact on prices of final goods and services. Also, it is not clear how the fuel savings from the
rule would be utilized by trucking firms. For example, trucking firms may reinvest fuel savings
in their own company; retain fuel savings as profits; pass fuel savings onto customers or others;
or increase driver pay. Finally, it is not clear how the different pathways that fuel savings would
be utilized would affect greenhouse gas emissions.
Research on indirect and economy-wide rebound effects is scant, and we have not
identified any peer-reviewed research that attempts to quantify indirect or economy-wide
rebound effects for FED Vs. In particular, the agencies are not aware of any peer-reviewed
approach which indicates that the magnitude of indirect or economy-wide rebound effects, if
any, would be significant for this final rule.N Therefore, we rely on the analysis of vehicle miles
traveled to estimate the rebound effect in this rule, as we did for the HD Phase 1 rule, where we
attempted to quantify only rebound effects from our rule that impact HDV VMT.
In order to test the effect of alternative assumptions about the rebound effect, NHTSA
examined the sensitivity of its estimates of benefits and costs of the proposed Phase 2 program
for HD pickups and vans to alternative assumptions about the rebound effect. While the main
analysis for pickups and vans assumes a 10 percent rebound effect, the sensitivity analysis
estimates the benefits and costs of the standards under the assumptions of 5, 15, and 20 percent
rebound effects. This sensitivity analysis can be found in Section IX.E.3 of the NPRM
Preamble0 and shows that (a) using a 5 percent value for the rebound effect reduced benefits and
costs of the proposed standards by identical amounts, leaving net benefits unaffected; and (b)
values of the rebound effect above 10 percent increased costs and reduced benefits from their
values in the main analysis, thus reducing net benefits of the proposed standards. Nevertheless,
the proposed and now the final program have significant net benefits and these alternative values
of the rebound effect would not have affected the agencies' selection of the final program
stringency, as that selection is based on NHTSA's assessment of the maximum feasible fuel
N The same entity responsible for these comments also sought reconsideration of the Phase 1 rule on the grounds
that indirect rebound effects had not been considered by the agencies and could negate all of the benefits of the
standards. This assertion rested on an unsupported affidavit lacking any peer review or other indicia of objectivity.
This affidavit cited only one published study. The study cited did not deal with vehicle efficiency, has
methodological limitations (many of them acknowledged), and otherwise was not pertinent. EPA and NHTSA thus
declined to reconsider the Phase 1 rule based on these speculative assertions. See generally 77 FR 51703-04 (Aug.
27, 2012) and 77 FR 51502-03 (Aug. 24, 2012). The analysis in this entity's comments on this rulemaking rests
largely on that same unsupported affidavit.
0 80 FR 40137.

-------
efficiency standards and EPA's selection of appropriate GHG standards to address energy
security and the environment.
8.4 Impact on Class Shifting, Fleet Turnover, and Sales
The agencies considered two additional potential indirect effects which may lead to
unintended consequences of the program to improve the fuel efficiency and reduce GHG
emissions from HD trucks. The next sections cover the agencies' qualitative discussions on
potential class shifting and fleet turnover effects.
8,4,1 Class Shifting
Heavy-duty vehicles are typically configured and purchased to perform a function. For
example, a concrete mixer truck is purchased to transport concrete, a combination tractor is
purchased to move freight with the use of a trailer, and a Class 3 pickup truck could be
purchased by a landscape company to pull a trailer carrying lawnmowers. The purchaser makes
decisions based on many attributes of the vehicle, including the gross vehicle weight rating of the
vehicle, which in part determines the amount of freight or equipment that can be carried. If the
Phase 2 standards impact either the performance of the vehicle or the marginal cost of the vehicle
relative to the other vehicle classes, then consumers may choose to purchase a different vehicle,
resulting in the unintended consequence of increased fuel consumption and GHG emissions in-
use.
The agencies, along with the NAS panel, found that there is little or no literature which
evaluates class shifting between trucks.31 In addition, the agencies did not receive comments
specifically raising concerns about class shifting. NHTSA and EPA qualitatively evaluated the
final rules in light of potential class shifting. The agencies looked at four potential cases of
shifting: - from light-duty pickup trucks to heavy-duty pickup trucks; from sleeper cabs to day
cabs; from combination tractors to vocational vehicles; and within vocational vehicles.
Light-duty pickup trucks, those with a GVWR of less than 8,500 pounds, are currently
regulated under the existing GHG/CAFE standards for light duty vehicles. The increased
stringency of the light-duty 2017-2025 MY vehicle rule has led some to speculate that vehicle
consumers may choose to purchase heavy-duty pickup trucks that are currently regulated under
the HD Phase 1 program if the cost of the light-duty regulation is high relative to the cost to buy
the larger heavy-duty pickup trucks. Since fuel consumption and GHG emissions rise
significantly with vehicle mass, a shift from light-duty trucks to heavy-duty trucks would likely
lead to higher fuel consumption and GHG emissions, an untended consequence of the
regulations. Given the significant price premium of a heavy-duty truck (often five to ten
thousand dollars more than a light-duty pickup), we believe that such a class shift would be
unlikely whether or not this program existed. These final rules would continue to diminish any
incentive for such a class shift because they would narrow the GHG and fuel efficiency
performance gap between light-duty and heavy-duty pickup trucks. The regulations for the HD
pickup trucks, and similarly for vans, are based on similar technologies and therefore reflect a
similar expected increase in cost when compared to the light-duty GHG regulation. Hence, the
combination of the two regulations provides little incentive for a shift from light-duty trucks to
HD trucks. To the extent that this regulation of heavy-duty pickups and vans could conceivably

-------
encourage a class shift towards lighter pickups, this unintended consequence would in fact be
expected to lead to lower fuel consumption and GHG emissions as the smaller light-duty pickups
have significantly better fuel economy ratings than heavy-duty pickup trucks.
The projected cost increases for this action differ between Class 8 day cabs and Class 8
sleeper cabs, reflecting our conservative assumption for purposes of this analysis on shifting that
compliance with the standards would lead truck consumers to specify sleeper cabs equipped with
APUs or alternatives to APU while day cab consumers would not. Since Class 8 day cab and
sleeper cab trucks perform essentially the same function when hauling a trailer, this raises the
possibility that the additional cost for an APU or alternatives to APU equipped sleeper cab could
lead to a shift from sleeper cab to day cab trucks. We do not believe that such an intended
consequence would occur for the following reasons. The addition of a sleeper berth to a tractor
cab is not a consumer-selectable attribute in quite the same way as other vehicle features. The
sleeper cab provides a utility that long-distance trucking fleets need to conduct their operations —
an on-board sleeping berth that lets a driver comply with federally-mandated rest periods, as
required by the Department of Transportation Federal Motor Carrier Safety Administration's
hours-of-service regulations. The cost of sleeper trucks is already higher than the cost of day
cabs, yet the fleets that need this utility purchase them.32 A day cab simply cannot provide this
utility with a single driver. The need for this utility would not be changed even if the additional
costs to reduce greenhouse gas emissions from sleeper cabs exceed those reducing greenhouse
gas emissions from day cabs.p
A trucking fleet could instead decide to put its drivers in hotels in lieu of using sleeper
berths, and switch to day cabs. However, this is unlikely to occur in any great number, since the
added cost for the hotel stays would far overwhelm differences in the marginal cost between day
and sleeper cabs. Even if some fleets do opt to buy hotel rooms and switch to day cabs, they
would be highly unlikely to purchase a day cab that was aerodynamically worse than the sleeper
cab they replaced, since the need for features optimized for long-distance hauling would not have
changed. So in practice, there would likely be little difference to the environment for any
switching that might occur. Further, while our projected costs in the NPRM assumed the
purchase of an APU for compliance for nearly all sleeper cabs, the updated analysis reflects
additional flexibility in the final rules that would allow manufacturers to use several other
alternatives to APUs that would be much less expensive. Thus, even though we are now
projecting that APU costs will be somewhat higher than what we projected for the NPRM,
manufacturers and consumers will not be required to use them. In fact our regulatory structure
would allow compliance using a near zero cost software utility that eliminates tractor idling after
five minutes. Using this compliance approach, the cost difference between a Class 8 sleeper cab
and day cab due to our regulations is small. We are proposing this alternative compliance
approach reflecting that some sleeper cabs are used in team driving situations where one driver
sleeps while the other drives. In that situation, an APU is unnecessary since the tractor is
continually being driven when occupied. When it is parked, it would automatically eliminate
p The average marginal cost difference between sleeper cabs and day cabs in the final rule is roughly $2,500.

-------
any additional idling through the shutdown software. If trucking businesses choose this option,
then costs based on purchase of APUs may overestimate the costs of this program to this sector.
Class shifting from combination tractors to vocational vehicles may occur if a customer
deems the additional marginal cost of tractors due to the regulation to be greater than the utility
provided by the tractor. The agencies initially considered this issue when deciding whether to
include Class 7 tractors with the Class 8 tractors or regulate them as vocational vehicles. The
agencies' evaluation of the combined vehicle weight rating of the Class 7 shows that if these
vehicles were treated significantly differently from the Class 8 tractors, then they could be easily
substituted for Class 8 tractors. Therefore, the agencies would continue to include both classes
in the tractor category. The agencies believe that a shift from tractors to vocational vehicles
would be limited because of the ability of tractors to pick up and drop off trailers at locations
which cannot be done by vocational vehicles.
The agencies do not envision that the regulatory program would cause class shifting
within the vocational vehicle class. As vocational vehicles include a wide variety of vehicle
types, and serve a wide range of functions, the diversity in the vocational vehicle segment can be
primarily attributed to the variety of customer needs for specialized vehicle bodies and added
equipment, rather than to the chassis. The new standards are projected to lead to a small increase
in the incremental cost per vehicle. However, these cost increases are consistent across the board
for both vocational vehicles and the engines used in the vehicle (Table V-30 at Preamble Section
V.C.3). The agencies believe that the utility gained from the additional technology package
would outweigh the additional cost for vocational vehicles.Q
In conclusion, NHTSA and EPA believe that the regulatory structure for HD vehicles and
engines would not significantly change the current competitive and market factors that determine
purchaser preferences. Furthermore, even if a small amount of shifting would occur, any
resulting GHG impacts would likely to be negligible because any vehicle class that sees an
uptick in sales is also being regulated for GHG emission control and fuel efficiency. Therefore,
the agencies did not include an impact of class shifting on the vehicle populations used to assess
the benefits of the program.
8.4.2 Fleet Turnover and Sales Effects
A regulation that affects the cost to purchase and/or operate trucks could affect whether a
consumer decides to purchase a new truck and the timing of that purchase. The term pre-buy
refers to the idea that truck purchases may occur earlier than otherwise planned to avoid the
additional costs associated with a new regulatory requirement. Slower fleet turnover, or low-
Q The final rule projects the average per-vehicle costs associated with the 2027 MY standards are projected to be
generally less than six percent of the overall price of a new vehicle. The cost-effectiveness of these vocational
vehicle standards in dollars per ton is similar to the cost effectiveness estimated for light-duty trucks in the 2017-
2025 light duty greenhouse gas standards (Preamble Section V.C.3) (which the agencies found to be highly cost-
effective, even without considering payback due to fuel savings).

-------
buys, may occur when owners opt to keep their existing truck rather than purchase a new truck
due to the incremental cost of the regulation.
The 2010 NAS HD Report discussed the topics associated with medium- and heavy-duty
vehicle fleet turnover. NAS noted that there is some empirical evidence of pre-buy behavior in
response to the 2004 and 2007 heavy-duty engine emission standards, with larger impacts
occurring in response to higher costs.33 However, those regulations increased upfront costs to
firms without any offsetting future cost savings from reduced fuel purchases. In summary, NAS
stated that:
.. .during periods of stable or growing demand in the freight sector, pre-buy behavior may
have significant impact on purchase patterns, especially for larger fleets with better
access to capital and financing. Under these same conditions, smaller operators may
simply elect to keep their current equipment on the road longer, all the more likely given
continued improvements in diesel engine durability over time. On the other hand, to the
extent that fuel economy improvements can offset incremental purchase costs, these
impacts will be lessened. Nevertheless, when it comes to efficiency investments, most
heavy-duty fleet operators require relatively quick payback periods, on the order of two
to three years.34
The regulations are projected to return fuel savings to the vehicle owners that offset the
cost of the regulation within a few years. The effects of the regulation on purchasing behavior
and sales will depend on the nature of the market failures and the extent to which firms consider
the projected future fuel savings in their purchasing decisions.
If trucking firms or other buyers account for the rapid payback, they are unlikely to
strategically accelerate or delay their purchase plans at additional cost in capital to avoid a
regulation that will lower their overall operating costs. As discussed in Chapter 8.2, this scenario
may occur if this program reduces uncertainty about fuel-saving technologies. More reliable
information about ways to reduce fuel consumption allows truck purchasers to evaluate better the
benefits and costs of additional fuel savings, primarily in the original vehicle market, but
possibly in the resale market as well. In addition, the standards are expected to lead
manufacturers to install more fuel-saving technologies and promote their purchase; the increased
availability and promotion may encourage sales.
Other market failures may leave open the possibility of some pre-buy or delayed
purchasing behavior. Firms may not consider the full value of the future fuel savings for several
reasons. For instance, truck purchasers may not want to invest in fuel efficiency because of
uncertainty about fuel prices. Another explanation is that the resale market may not fully
recognize the value of fuel savings, due to lack of trust of new technologies or changes in the
uses of the vehicles. Lack of coordination (also called split incentives—see Chapter 8.2)
between truck purchasers (who may emphasize the up-front costs of the trucks) and truck
operators, who like the fuel savings, can also lead to pre-buy or delayed purchasing behavior. If
these market failures prevent firms from fully internalizing fuel savings when deciding on
vehicle purchases, then pre-buy and delayed purchase could occur and could result in a slight
decrease in the GHG benefits of the regulation.

-------
Thus, whether pre-buy or delayed purchase is likely to play a significant role in the truck
market depends on the specific behaviors of purchasers in that market. Without additional
information about which scenario is more likely to be prevalent, the agencies are not projecting a
change in fleet turnover characteristics due to this regulation.
Whether vehicle sales appear to be affected by the HD Phase 1 standards could provide
some insight into the impacts of the standards. The Environmental Defense Fund observes that
MY 2014 heavy-duty trucks had the highest sales since 2005. Any trends in sales are likely to be
affected by macroeconomic conditions, which have been recovering since 2009-2010. The
standards may have affected sales, but the size of that effect is likely to be swamped by the
effects of the economic recovery. It is unlikely to be possible to separate the effects of the
existing standards from other confounding factors.
8.5 Monetized GHG Impacts
8.5,1 Monetized CO2 Impacts - Social Cost of Carbon
We estimate the global social benefits of CO2 emission reductions expected from the HD
Phase 2 program using the social cost of carbon (SC-CO2) estimates presented in the Technical
Support Document: Technical Update of the Social Cost of Carbon for Regulatory Impact
Analysis Under Executive Order 12866 (May 2013, Revised July 2015) ("current SC-CO2
TSD"). We refer to these estimates, which were developed by the U.S. government, as "SC-CO2
estimates." The SC-CO2 is a metric that estimates the monetary value of impacts associated with
marginal changes in CO2 emissions in a given year. It includes a wide range of anticipated
climate impacts, such as net changes in agricultural productivity and human health, property
damage from increased flood risk, and changes in energy system costs, such as reduced costs for
heating and increased costs for air conditioning. It is typically used to assess the avoided
damages as a result of regulatory actions (i.e., benefits of rulemakings that lead to an incremental
reduction in cumulative global CO2 emissions).
The SC-CO2 estimates used in this analysis were developed over many years, using the
best science available, and with input from the public. Specifically, an interagency working
group (IWG) that included the EPA and other executive branch agencies and offices used three
integrated assessment models (IAMs) to develop the SC-CO2 estimates and recommended four
global values for use in regulatory analyses. The SC-CO2 estimates were first released in
February 2010 and updated in 2013 using new versions of each IAM. The 2013 update did not
revisit the 2010 modeling decisions with regards to the discount rate, reference case
socioeconomic and emission scenarios, and equilibrium climate sensitivity distribution. Rather,
improvements in the way damages are modeled are confined to those that have been incorporated
into the latest versions of the models by the developers themselves and published in the peer-
reviewed literature. The 2010 SC-CO2 Technical Support Document (2010 SC-CO2 TSD)
provides a complete discussion of the methods used to develop these estimates and the current

-------
SC-CO2 TSD presents and discusses the 2013 update (including recent minor technical
corrections to the estimates).11
One key methodological aspect discussed in the SC-CO2 TSDs is the global scope of the
estimates. The SC-CO2 estimates represent global measures because of the distinctive nature of
the climate change, which is highly unusual in at least three respects. First, emissions of most
GHGs contribute to damages around the world independent of the country in which they are
emitted. Second, the U.S. operates in a global, highly interconnected economy, such that
impacts on the other side of the world can affect our economy. This means that the true costs of
climate change to U.S. are much larger than the direct impacts that simply occur in the U.S.
Third, climate change represents a classic public goods problem because each country's
reductions benefit everyone else and no country can be excluded from enjoying the benefits of
other countries' reductions, even if it provides no reductions itself. In this situation, the only
way to achieve an economically efficient level of emissions reductions is for countries to
cooperate in providing mutually beneficial reductions beyond the level that would be justified
only by their own domestic benefits. In reference to the public good nature of mitigation and its
role in foreign relations, thirteen prominent academics noted that these "are compelling reasons
to focus on a global SCC" (Pizer et al., 2014). In addition, the IWG recently noted that there is
no bright line between domestic and global damages. Adverse impacts on other countries can
have spillover effects on the United States, particularly in the areas of national security,
international trade, public health and humanitarian concerns.s
The 2010 SC-CO2 TSD also noted a number of limitations to the SC-CO2 analysis,
including the incomplete way in which the IAMs capture catastrophic and non-catastrophic
impacts, their incomplete treatment of adaptation and technological change, uncertainty in the
extrapolation of damages to high temperatures, and assumptions regarding risk aversion.
Currently IAMs do not assign value to all of the important physical, ecological, and economic
impacts of climate change recognized in the climate change literature due to a lack of precise
information on the nature of damages and because the science incorporated into these models
understandably lags behind the most recent research.1 The limited amount of research linking
climate impacts to economic damages makes the modeling exercise even more difficult. These
individual limitations do not all work in the same direction in terms of their influence on the SC-
CO2 estimates, though taken together they suggest that the SC-CO2 estimates are likely
conservative. In particular, the IPCC Fourth Assessment Report (2007) concluded that "It is very
likely that [SC-CO2 estimates] underestimate the damage costs because they cannot include
many non-quantifiable impacts." Since then, the peer-reviewed literature has continued to
R Both the 2010 SC-CO2 TSD and the current SC-CO2 TSD are available at:
https://www.whitehouse.gov/omb/oira/social-cost-of-carbon.
s See Response to Comments: Social Cost of Carbon for Regulatory Impact Analysis under Executive Order 12866,
July 2015, page 31, at https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-comments-final-
july-2015.pdf.
T Climate change impacts and social cost of greenhouse gases modeling is an area of active research. For example,
see: (1) Howard, Peter, "Omitted Damages: What's Missing from the Social Cost of Carbon." March 13, 2014,
http://costofcarbon.org/files/Omitted_Damages_Whats_Missing_From_the_Social_Cost_of_Carbon.pdf; and (2)
Electric Power Research Institute, "Understanding the Social Cost of carbon: A Technical Assessment," October
2014, www.epri.com.

-------
support this conclusion. For example, the IPCC Fifth Assessment report (2014) observed that
SC-CO2 estimates continue to omit various impacts, such as "the effects of the loss of
biodiversity among pollinators and wild crops on agriculture." Nonetheless, these estimates and
the discussion of their limitations represent the best available information about the social
benefits of CO2 reductions to inform benefit-cost analysis. The new versions of the models used
to estimate the values presented below offer some improvements in these areas, although further
work is warranted.
Accordingly, EPA and other agencies continue to engage in research on modeling and
valuation of climate impacts with the goal to improve these estimates. The EPA and other
federal agencies also continue to consider feedback on the SC-CO2 estimates from stakeholders
through a range of channels, including public comments on Agency rulemakings that use the SC-
CO2 in supporting analyses and through regular interactions with stakeholders and research
analysts implementing the SC-CO2 methodology used by the IWG. The SC-CO2 comments
received on this rulemaking covered the technical details of the modeling conducted to develop
the SC-CO2 estimates and some also provided constructive recommendations for potential
opportunities to improve the SC-CO2 estimates in future updates. Section 11.8 of the RTC
document provides a summary and response to the SC-CO2 comments submitted to this
rulemaking. In addition, OMB sought public comment on the approach used to develop the SC-
CO2 estimates through a separate comment period and published a response to those comments
in 2015.u
After careful evaluation of the full range of comments submitted to OMB, the IWG
continues to recommend the use of the SC-CO2 estimates in regulatory impact analysis. With
the July 2015 release of the response to comments, the IWG announced plans to obtain expert
independent advice from the National Academies of Sciences, Engineering and Medicine to
ensure that the SC-CO2 estimates continue to reflect the best available scientific and economic
information on climate change.v The Academies then convened a committee, "Assessing
Approaches to Updating the Social Cost of Carbon," (Committee) which is reviewing the state of
the science on estimating the SC-CO2, and will provide expert, independent advice on the merits
of different technical approaches for modeling and highlight research priorities going forward.
EPA will evaluate its approach based upon any feedback received from the Academies' panel.
To date, the Committee has released an interim report, which recommended against doing
a near term update of the SC-CO2 estimates. For future revisions, the Committee recommended
the IWG move efforts towards a broader update of the climate system module consistent with the
most recent, best available science, and also offered recommendations for how to enhance the
discussion and presentation of uncertainty in the SC-CO2 estimates. Specifically, the Committee
recommended that "the IWG provide guidance in their technical support documents about how
[SC-CO2] uncertainty should be represented and discussed in individual regulatory impact
u See https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-comments-final-july-2015.pdf
v The Academies' review will be informed by public comments and focus on the technical merits and challenges of
potential approaches to improving the SC-CO2 estimates in future updates. See
https://www.whitehouse.gov/blog/2015/07/02/estimating-benefits-carbon-dioxide-emissions-reductions.

-------
analyses that use the [SC-CO2]" and that the technical support document for each update of the
estimates present a section discussing the uncertainty in the overall approach, in the models used,
and uncertainty that may not be included in the estimates.w At the time of this writing, the IWG
is reviewing the interim report and considering the recommendations. EPA looks forward to
working with the IWG to respond to the recommendations and will continue to follow IWG
guidance on SC-CO2.
The four SC-CO2 estimates are as follows: $13, $46, $68, and $140 per metric ton of
CO2 emissions in the year 2020 (2013$).x Table 8-5 presents the SC-CO2 estimates in selected
years, rounded to two significant digits. The first three values are based on the average SC-CO2
from the three IAMs, at discount rates of 5, 3, and 2.5 percent, respectively. SC-CO2 estimates
for several discount rates are included because the literature shows that the SC-CO2 is quite
sensitive to assumptions about the discount rate, and because no consensus exists on the
appropriate rate to use in an intergenerational context (where costs and benefits are incurred by
different generations). The fourth value is the 95th percentile of the SC-CO2 from all three
models at a 3 percent discount rate. It is included to represent lower probability but higher
impact outcomes from climate change, which are captured further out in the tail of the SC-CO2
distribution, and while less likely than those reflected by the average SC-CO2 estimates, would
be much more harmful to society and therefore, are relevant to policy makers. The SC-CO2
increases over time because future emissions are expected to produce larger incremental
damages as economies grow and physical and economic systems become more stressed in
response to greater climate change.
Table 8-5 Social Cost of CO2,2012 - 2050a (in 2013$ per Metric Ton)
CALENDAR
YEAR
DISCOUNT RATE AND STATISTIC
5% Average
3% Average
2.5% Average
3%
95th percentile
2012
$12
$36
$58
$100
2015
$12
$40
$62
$120
2020
$13
$46
$68
$140
2025
$15
$51
$75
$150
2030
$18
$55
$80
$170
2035
$20
$60
$86
$180
2040
$23
$66
$92
$200
2045
$25
$70
$98
$220
2050
$29
$76
$100
$230
Note:
" The SC-CO2 values are dollar-year and emissions-year specific and have been rounded to two
significant digits. Unrounded numbers from the current SC-CO2 TSD were adjusted to 2013$ and
used to calculate the CO2 benefits.
w National Academies of Sciences, Engineering, and Medicine. (2016). Assessment of Approaches to Updating the
Social Cost of Carbon: Phase 1 Report on a Near-Term Update. Committee on Assessing Approaches to Updating
the Social Cost of Carbon, Board on Environmental Change and Society. Washington, DC: The National Academies
Press, doe 10.17226/21898. See Executive Summary, page I, for quoted text.
x The SC-CO2 values have been rounded to two significant digits. Unrounded numbers from the current SC-CO2
TSD were adjusted to 2013$ and used to calculate the CO2 benefits.

-------
Applying the global SC-CO2 estimates, shown in Table 8-5, to the estimated reductions
in domestic CO2 emissions for the program, we estimate the dollar value of the climate related
benefits for each analysis year. In order to calculate the dollar value for emission reductions, the
SC-CO2 estimate for each emissions year will be applied to changes in CO2 emissions for that
year, and then discounted back to the analysis year using the same discount rate used to estimate
the SC-CO2. For internal consistency, the annual benefits are discounted back to net present
value terms using the same discount rate as each SC-CO2 estimate (i.e. 5 percent, 3 percent, and
2.5 percent) rather than the discount rates of 3 percent and 7 percent used to derive the net
present value of other streams of costs and benefits of the final rule.Y The SC-CO2 estimates
and the associated CO2 benefit estimates for each calendar year are shown in Table 8-6.
Table 8-6 Annual Upstream and Downstream CO2 Benefits and Net Present Values for the Given SC-CO2
Value for the Final Program Relative to the Flat Baseline and using Method B,a b (Millions of 2012$)
CALEND
5%
3%
2.5%
3%
AR
(AVERAGE SC-CO2
(AVERAGE SC-CO2
(AVERAGE SC-CO2
(95™ PERCENTILE =
YEAR
=
=
=
$100 IN 2012)

$12 IN 2012)
$36 IN 2012)
$58 IN 2012)

2018
$6.5
$22
$33
$63
2019
$13
$46
$68
$130
2020
$21
$73
$110
$210
2021
$80
$280
$420
$840
2022
$170
$550
$820
$1,700
2023
$250
$850
$1,300
$2,600
2024
$390
$1,300
$2,000
$4,000
2025
$560
$1,800
$2,700
$5,500
2026
$700
$2,400
$3,500
$7,100
2027
$950
$3,000
$4,400
$9,100
2028
$1,100
$3,700
$5,400
$11,000
2029
$1,300
$4,300
$6,400
$13,000
2030
$1,600
$5,000
$7,300
$15,000
2035
$2,700
$8,100
$11,000
$25,000
2040
$3,700
$11,000
$15,000
$33,000
2050
$5,500
$15,000
$20,000
$45,000
NPVb
$24,000
$110,000
$180,000
$340,000
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b The SC-CO2 values are dollar-year and emissions-year specific. Note that discounted values of reduced GHG
emissions are calculated differently than other benefits. The same discount rate used to discount the value of
damages from future emissions (SC-CO2 at 5, 3, and 2.5 percent) is used to calculate discounted values of SC-CO2
for internal consistency. Refer to SC-CO2 TSD for more detail.
Y See more discussion on the appropriate discounting of climate benefits using SC-CO2 in the 2010 SC-CO2 TSD.
Other benefits and costs of regulations unrelated to CO2 emissions are discounted at the 3% and 7% rates specified
in OMB guidance for regulatory analysis.

-------
We also conducted a separate analysis of the CO2 benefits over the model year lifetimes
of vehicles sold in the regulatory timeframe. In contrast to the calendar year analysis, the model
year lifetime analysis shows the impacts of the program on each of these MY fleets over the
course of their lifetimes. Full details of the inputs to this analysis can be found in RIA Chapter
5. The CO2 benefits in the context of this MY lifetime analysis are shown in Table 8-7 for each
of the four different social cost of carbon values. The CO2 benefits shown for each model year
represent the net present value of the benefits in each year in the model year life discounted back
to the first year of the model year. The same discount rate used to discount the value of damages
from future emissions (SC-CO2 at 5, 3, and 2.5 percent) is used to calculate the net present value
of SC-CO2 for internal consistency.
Table 8-7 Discounted Model Year Lifetime Upstream & Downstream CO2 Benefits for the Given SC-CO2
Value for the Final Program Relative to the Less Dynamic Baseline and using Method B (Millions of 2012$)a b
MODEL
5%
3%
2.5%
3%
YEAR
(AVERAGE SC-CO2
(AVERAGE SC-CO2
(AVERAGE SC-CO2
(95™ PERCENTILE

$12 IN 2012)
$36 IN 2012)
$58 IN 2012)
$100 IN 2012)
2018
$38
$150
$230
$450
2019
$36
$140
$220
$430
2020
$34
$140
$220
$420
2021
$560
$2,300
$3,600
$7,000
2022
$590
$2,500
$3,900
$7,500
2023
$610
$2,600
$4,000
$7,800
2024
$920
$4,000
$6,200
$12,000
2025
$940
$4,100
$6,400
$12,000
2026
$950
$4,200
$6,600
$13,000
2027
$1,200
$5,400
$8,500
$16,000
2028
$1,200
$5,300
$8,400
$16,000
2029
$1,200
$5,300
$8,400
$16,000
Sum
$8,200
$36,000
$57,000
$110,000
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b The SC-CO2 values are dollar-year and emissions-year specific. Note that discounted values of reduced GHG
emissions are calculated differently than other benefits. The same discount rate used to discount the value of
damages from future emissions (SC-CO2 at 5, 3, and 2.5 percent) is used to calculate discounted values of SC-CO2
for internal consistency. Refer to SC-CO2 TSD for more detail.
8.5.2 Non-CCh GHG Impacts
EPA calculated the global social benefits of CH4 and N2O emissions reductions expected
from the final rulemaking using estimates of the social cost of methane (SC-CH4) and the social
cost of nitrous oxide (SC-N2O). Similar to the SC- CO2, the SC- CH4 and SC- N2O estimate the
monetary value of impacts associated with marginal changes in CH4 and N2O emissions,
respectively, in a given year. Each metric includes a wide range of anticipated climate impacts,
such as net changes in agricultural productivity and human health, property damage from
increased flood risk, and changes in energy system costs, such as reduced costs for heating and
increased costs for air conditioning. The SC-CH4 and SC-N2O estimates applied in this analysis
were developed by Marten et al. (2014) and are discussed in greater detail below. EPA is

-------
unaware of analogous estimates of HFC-134a and has therefore presented a sensitivity analysis,
separate from the main benefit cost analysis, that approximates the benefits of HFC-134a
reductions based on global warming potential (GWP) gas comparison metrics ("GWP
approach"). Other unquantified non-CCh benefits are discussed in this section as well.
8.5.2.1 Monetized CH4 and N2O Impacts
As discussed in the proposed rulemaking, a challenge particularly relevant to the
monetization of non- CO2 GHG impacts is that the IWG did not estimate the social costs of non-
CO2 GHG emissions at the time the SC-CO2 estimates were developed. While there are other
estimates of the social cost of non- CO2 GHGs in the peer review literature, none of those
estimates are consistent with the SC- CO2 estimates developed by the IWG and most are likely
underestimates due to changes in the underlying science subsequent to their publication.2
However, in the time leading up to the proposal for this rulemaking, a paper by Marten et
al. (2014) provided the first set of published SC-CH4 and SC-N2O estimates in the peer-
reviewed literature that are consistent with the modeling assumptions the IWG used to develop
the SC-CO2 estimates. Specifically, the estimation approach Marten et al. used incorporated the
same set of three IAMs, five socioeconomic and emissions scenarios, equilibrium climate
sensitivity distribution, three constant discount rates, and the aggregation approach used by the
IWG to develop the SC-CO2 estimates. The aggregation method involved distilling the 45
distributions of each metric produced for each emissions year into four estimates: the mean
across all models and scenarios using a 2.5 percent, 3 percent, and 5 percent discount rate, and
the 95th percentile of the pooled estimates from all models and scenarios using a 3 percent
discount rate. Marten et al. also used the same rationale as the IWG to develop global estimates
of the SC-CH4 and the SC-N2O, given that CH4 and N2O are global pollutants.
In addition, the atmospheric lifetime and radiative efficacy of methane used by Marten et
al. is based on the estimates reported by the IPCC in their Fourth Assessment Report (AR4,
2007), including an adjustment in the radiative efficacy of methane to account for its role as a
precursor for tropospheric ozone and stratospheric water. These values represent the same ones
used by the IPCC in AR4 for calculating GWPs. At the time Marten et al. developed their
estimates of the SC-CH4, AR4 was the latest assessment report by the IPCC. The IPCC updates
GWP estimates with each new assessment, and in the most recent assessment, AR5, the latest
estimate of the methane GWP ranged from 28-36, compared to a GWP of 25 in AR4. The
updated values reflect a number of changes: changes in the lifetime and radiative efficiency
estimates for CO2, changes in the lifetime estimate for methane, and changes in the correction
factor applied to methane's GWP to reflect the effect of methane emissions on other climatically
z As discussed in previous RIAs (e.g., EPA 2012), there is considerable variation among these published estimates
in the models and input assumptions they employ. These studies differ in the emission perturbation year, employ a
wide range of constant and variable discount rate specifications, and consider a range of baseline socioeconomic and
emissions scenarios that have been developed over the last 20 years. The researchers cited in EPA 2012 include:
Fankhauser (1994); Kandlikar (1995); Hammitt et al. (1996); Tol et al. (2003); Tol (2004); and Hope and Newberry
(2006).

-------
important substances such as tropospheric ozone and stratospheric water vapor. In addition, the
range presented in the latest IPCC report reflects different choices regarding whether to account
for climate feedbacks on the carbon cycle for both methane and CO2 (rather than just for CO2 as
was done in AR4).aa'bb
The resulting SC-CH4 and SC-N2O estimates are presented in Table 8-8. Marten etal. (2014)
discuss these estimates and compare them with other recent estimates in the literature. The authors
noted that a direct comparison of their estimates with all of the other published estimates is
difficult, given the differences in the models and socioeconomic and emissions scenarios, but
results from three relatively recent studies offer a better basis for comparison (see Hope (2006),
Marten and Newbold (2012), Waldhoff et al. (2014)). Marten et al. found that, in general, the SC-
CH4 estimates from their 2014 paper are higher than previous estimates and the SC-N2O estimates
from their 2014 paper fall within the range from Waldhoff et al. The higher SC- CH4 estimates
are partially driven by the higher effective radiative forcing due to the inclusion of indirect effects
from methane emissions in their modeling. Marten et al., similar to other recent studies, also find
that their directly modeled SC- CH4 and SC-N2O estimates are higher than the GWP-weighted
estimates. More detailed results and a comparison to other published estimates can be found in
Marten et al. (2014).
Table 8-8 Social Cost of CH4 and N20,2012 - 2050a [2013$ per metric ton]
(Source: Marten et al. (2014)b)
YEAR
SC-CH4
SC-N2O

5%
3%
2.5%
3%
5%
3%
2.5%
3%

Average
Average
Average
95th percentile
Average
Average
Average
95th
percentile
2012
$440
$1,000
$1,400
$2,800
$4,000
$14,000
$21,000
$36,000
2015
490
1,100
1,500
3,100
4,400
14,000
22,000
38,000
2020
590
1,300
1,800
3,500
5,200
16,000
24,000
43,000
2025
710
1,500
2,000
4,100
6,000
19,000
26,000
48,000
2030
830
1,800
2,200
4,600
6,900
21,000
30,000
54,000
2035
990
2,000
2,500
5,400
8,100
23,000
32,000
60,000
2040
1,100
2,200
2,900
6,000
9,200
25,000
35,000
66,000
2045
1,300
2,500
3,100
6,700
10,000
27,000
37,000
73,000
2050
1,400
2,700
3,400
7,400
12,000
30,000
41,000
79,000
Notes:
a The values are emissions-year specific and have been rounded to two significant digits, as shown in Marten et al.
(2014). These rounded numbers were used to calculate the GHG benefits.
^ Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment
Report of the Intergovernmental Panel on Climate Change [Stacker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K.
Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge,
United Kingdom and New York, NY, USA.
BB Note that this analysis uses a GWP value for methane of 25 for CO2 equivalency calculations, consistent with the
GHG emissions inventories and the IPCC Fourth Assessment Report (AR4).

-------
b The estimates in this table have been adjusted to reflect the minor technical corrections to the SC-CO2 estimates
described above. See the Corrigendum to Marten et al. (2014),
http://www.tandfonline.com/doi/abs/10.1080/14693062.2015.1070550
In addition to requesting comment on these estimates in the proposed rulemaking, EPA
noted that it had initiated a peer review of the application of the Marten et al (2014) non- CO2
social cost estimates in regulatory analysis.cc EPA also stated that, pending a favorable peer
review, it planned to use the Marten et al (2014) estimates to monetize benefits of CH4 and N2O
emission reduction in the main benefit-cost analysis of the final rule.
Since then, EPA received responses that supported this application. Three reviewers
considered seven charge questions that covered issues such as the EPA's interpretation of the
Marten et al. estimates, the consistency of the estimates with the SC-CO2 estimates, the EPA's
characterization of the limits of the GWP-approach to value non-CCh GHG impacts, and the
appropriateness of using the Marten et al. estimates in regulatory impact analyses. The
reviewers agreed with the EPA's interpretation of Marten et al.'s estimates, generally found the
estimates to be consistent with the SC-CO2 estimates, and concurred with the limitations of the
GWP approach, finding directly modeled estimates to be more appropriate. While outside of the
scope of the review, the reviewers briefly considered the limitations in the SC-CO2 methodology
(e.g., those discussed earlier in this section) and noted that because the SC-CO2 and SC-CH4 and
SC-N2O methodologies are similar, the limitations also apply to the resulting SC-CH4 and SC-
N2O estimates. Two of the reviewers concluded that use of the SC-CH4 and SC-N2O estimates
developed by Marten et al. and published in the peer-reviewed literature is appropriate in RIAs,
provided that the Agency discuss the limitations, similar to the discussion provided for SC-CO2
and other economic analyses. All three reviewers encouraged continued improvements in the
SC-CO2 estimates and suggested that as those improvements are realized they should also be
reflected in the SC-CH4 and SC-N2O estimates, with one reviewer suggesting the SC-CH4 and
SC-N2O estimates lag this process. The EPA supports continued improvement in the SC-CO2
estimates developed by the U.S. government and agrees that improvements in the SC-CO2
estimates should also be reflected in the SC-CH4 and SC-N2O estimates. The fact that the
reviewers agree that the SC-CH4 and SC-N2O estimates are generally consistent with the SC-
CO2 estimates that are recommended by OMB's guidance on valuing CO2 emissions reductions,
leads the EPA to conclude that use of the SC-CH4 and SC-N2O estimates is an analytical
improvement over excluding CH4 and N2O emissions from the monetized portion of the benefit
cost analysis.
The EPA also carefully considered the full range of public comments and associated
technical issues on the Marten et al. estimates received through this rulemaking and determined
that it would continue to use the estimates in the final rulemaking analysis. Based on the
evaluation of the public comments on this rulemaking, the favorable peer review of the
cc For a copy of the peer review and the responses, see
https://cfpub.epa.gov/si/si_public_pra_view.cfm?dirEntryID=291976 (see "SCCH4 EPA PEER REVIEW
FILES.PDF").

-------
application of Marten et al. estimates, and past comments urging EPA to value non-CC>2 GHG
impacts in its rulemakings,DD EPA concluded that the estimates represent the best scientific
information on the impacts of climate change available in a form appropriate for incorporating
the damages from incremental CH4 and N2O emissions changes into regulatory analysis and has
therefore included those benefits in the main benefits analysis. Please see the Response to
Comments document, Section X, for detailed responses to the comments on non-CCh GHG
valuation.
The application of directly modeled estimates from Marten et al. (2014) to benefit-cost
analysis of a regulatory action is analogous to the use of the SC-CO2 estimates. Specifically, the
SC-CH4 and SC-N2O estimates in Table 8-8 are used to monetize the benefits of reductions in
CH4 and N2O emissions, respectively, expected as a result of the rulemaking. Forecasted
changes in CH4 (N2O) emissions in a given year, expected as a result of the regulatory action,
are multiplied by the SC-CH4 (SC-N2O) estimate for that year. To obtain a present value
estimate, the monetized stream of future non-CC>2 benefits are discounted back to the analysis
year using the same discount rate used to estimate the social cost of the non-CCh GHG emission
changes. The limitations for the SC-CO2 estimates discussed above likewise apply to the SC-
CH4 and SC-N2O estimates, given the consistency in the methodology.
The CH4 and N2O benefits based on Marten et al. (2014) are presented for each calendar
year in Table 8-9 and Table 8-10, respectively.
DD EPA sought public comments on the valuation of non-CCh GHG impacts in previous rulemakings (e.g., U.S.
EPA 2012b, 2012d). In general, the commenters that support valuation of CO2 impacts strongly encouraged EPA to
incorporate the monetized value of non-CCh GHG impacts into the benefit cost analysis, however they noted the
challenges associated with the GWP-approach, as discussed later in this section, and encouraged the use of directly-
modeled estimates of the social cost of non- CO2 GHGs to overcome those challenges.

-------
Table 8-9 Annual Upstream and Downstream CH4 GHG Benefits and Net Present Values for the Given SC-
CH4 Value for the Final Program Relative to the Flat Baseline and using Method B, using the Directly
Modeled Approach, Calendar Year Analysis (Millions of 2013$)a b
CALENDAR
5%
3%
2.5%
3%
YEAR
(AVERAGE SC-CH4
(AVERAGE SC-CH4
(AVERAGE SC-CH4
(95™ PERCENTILE

$440 IN 2012)
$1000 IN 2012)
$1400 IN 2012)
$2800 IN 2012)
2018
$0.3
$0.6
$0.8
$1.6
2019
$0.6
$1.3
$1.7
$3.4
2020
$0.9
$2.0
$2.7
$5.4
2021
$3.7
$8.2
$11
$22
2022
$7.4
$16
$21
$43
2023
$12
$26
$33
$68
2024
$19
$40
$52
$110
2025
$26
$56
$72
$150
2026
$34
$72
$92
$190
2027
$44
$94
$120
$250
2028
$54
$120
$150
$300
2029
$65
$140
$170
$360
2030
$76
$160
$200
$420
2035
$130
$260
$340
$720
2040
$180
$360
$460
$980
2050
$280
$530
$660
$1,400
NPVb
$1,200
$3,800
$5,400
$10,000
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b The SC-CH4 values are dollar-year and emissions-year specific. Note that discounted values of reduced GHG
emissions are calculated differently than other benefits. The same discount rate used to discount the value of
damages from future emissions (SC-CH4 at 5, 3, and 2.5 percent) is used to calculate discounted values of SC-CH4
for internal consistency.

-------
Table 8-10 Annual Upstream and Downstream N2O GHG Benefits and Net Present Values for the Given SC-
N2O Value for the Final Program Relative to the Less Dynamic Baseline and using Method B, using the
Directly Modeled Approach, Calendar Year Analysis (Millions of 2013$)a b
CALENDAR
5%
3%
2.5%
3%
YEAR
(AVERAGE SC-
(AVERAGE SC-N2O =
(AVERAGE SC-N2O =
(95™ PERCENTILE =

n2o =
$14000 IN 2012)
$21000 IN 2012)
$36000 IN 2012)

$4000 IN 2012)



2018
$0.0
$0.0
$0.0
$0.1
2019
$0.0
$0.1
$0.1
$0.2
2020
$0.0
$0.1
$0.2
$0.3
2021
$0.1
$0.4
$0.5
$1.0
2022
$0.2
$0.7
$1.1
$1.9
2023
$0.4
$1.2
$1.7
$3.0
2024
$0.6
$1.8
$2.6
$4.7
2025
$0.8
$2.5
$3.6
$6.6
2026
$1.1
$3.3
$4.6
$8.5
2027
$1.4
$4.2
$6.0
$11
2028
$1.7
$5.2
$7.4
$13
2029
$2.0
$6.2
00
00
&
$16
2030
$2.4
$7.2
$10
$19
2035
$4.1
$12
$16
$31
2040
$5.7
$16
$22
$41
2050
$8.9
$22
$30
$58
NPVb
$37
$160
$250
$430
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b The SC-N2O values are dollar-year and emissions-year specific. Note that discounted values of reduced GHG
emissions are calculated differently than other benefits. The same discount rate used to discount the value of
damages from future emissions (N2O at 5, 3, and 2.5 percent) is used to calculate discounted values of N20 for
internal consistency.
8.5.2.2 Sensitivity Analysis - HFC-134a Benefits Based on the GWP
Approximation Approach
While the rulemaking will result in reductions of HFC-134a, EPA is unaware of estimates
of the social cost of HFC-134a that are analogous to the SC- CO2, SC- CH4, and SC- N2O
estimates discussed in the previous section. Therefore, EPA has used an alternative approach to
approximate the value of HFC-134a impacts and presents the results in this sensitivity analysis,
separate from the main benefit cost analysis. Specifically, EPA has used the GWP for HFC-134a
to convert the emissions of this gas to CO2 equivalents, which are then valued using the SC-CO2
estimates.
The GWP measures the cumulative radiative forcing from a perturbation of a non-CCh
GHG relative to a perturbation of CO2 over a fixed time horizon, often 100 years. The GWP
mainly reflects differences in the radiative efficiency of gases and differences in their
atmospheric lifetimes. While the GWP is a simple, transparent, and well-established metric for
assessing the relative impacts of non-CC>2 emissions compared to CO2 on a purely physical
basis, there are several well-documented limitations in using it to value non-CC>2 GHG benefits,
as discussed in the 2010 SC-CO2 TSD and previous rulemakings (e.g., U.S. EPA 2012b,

-------
2012d).EE In particular, several recent studies found that GWP-weighted benefit estimates for
CH4 and N2O are likely to be lower than the estimates derived using directly modeled social
cost estimates for these gases (Marten and Newbold, 2012; Marten et al. 2014; and Waldhoff et
al. 2014). Gas comparison metrics, such as the GWP, are designed to measure the impact of
non-CCh GHG emissions relative to CO2 at a specific point along the pathway from emissions to
monetized damages (Depicted in Figure 8-2), and this point may differ across measures.








Environmental


Emissions
-~
Atmospheric
Concentration
->
Radiative
Forcing

Climate
Impacts

and Socio-
Economic
Impacts

Monetized
Damages
Figure 8-2 Path from GHG Emissions to Monetized Damages (Source: Marten et al., 2014)
The GWP is not ideally suited for use in benefit-cost analyses to approximate the social
cost of non-CCh GHGs because they ignore important nonlinear relationships beyond radiative
forcing in the chain between emissions and damages. These can become relevant because gases
have different lifetimes and the SC-CO2 takes into account the fact that marginal damages from
an increase in temperature are a function of existing temperature levels. Another limitation of
gas comparison metrics for this purpose is that some environmental and socioeconomic impacts
are not linked to all of the gases under consideration, or radiative forcing for that matter, and will
therefore be incorrectly allocated. For example, the economic impacts associated with increased
agricultural productivity due to higher atmospheric CO2 concentrations included in the SC-CO2
will be incorrectly allocated to HFC-134a emissions with the GWP-based valuation approach.
Furthermore, the assumptions made in estimating the GWP are not consistent with the
assumptions underlying SC-CO2 estimates in general, and the SC-CO2 estimates developed by
the IWG more specifically. For example the 100-year time horizon usually used in estimating
the GWP is less than the 300-year horizon the IWG used in developing the SC-CO2 estimates.
The GWP approach also treats all impacts within the time horizon equally, independent of the
time at which they occur. This is inconsistent with the role of discounting in economic analysis,
which accounts for a basic preference for earlier over later gains in utility and expectations
regarding future levels of economic growth. While EPA is unaware of studies that have
examined HFC-134a specifically, which has a relatively short lifetime compared to CO2, the
findings from Marten and Newbold 2012 suggest that the temporal independence of the GWP
could lead the GWP approach to underestimate the SC-HFC-134a with a larger downward bias
under higher discount rates. Additionally, because HFC-134a does not contribute to CO2
fertilization, that would also lead the GWP approach to underestimate the SC-HFC-134a (Marten
and Newbold 2012).FF
EE See also Reilly and Richards, 1993; Schmalensee, 1993; Fankhauser, 1994; Marten and Newbold, 2012.
FF The average atmospheric lifetime of HFC-134a is about 13 years. Marten and Newbold (2012) examined CH4,
which also has a relatively short atmospheric lifetime compared to CO2, and found that the GWP approach could
underestimate the SC-CH4. We note that the truncation of the time period in the GWP calculation could lead to an

-------
Although directly modeled estimates of the social cost of HFC-134a may offer an
improvement over the GWP approach, EPA is unaware of published estimates that are consistent
with the SC-CO2 estimates developed by the IWG. Therefore, EPA has continued to apply the
GWP approach to approximate the HFC-134a benefits. Given the limitations discussed above,
EPA also continues to present the results in a sensitivity analysis rather than the main benefit-
cost analysis.00
Under the GWP approach, EPA converted HFC-134a to CO2 equivalents for each
calendar year using the AR4 100-year GWP for HFC-134a (1,43 0).35 These CCh-equivalent
emission reductions are multiplied by the SC-CO2 estimate corresponding to each year of
emission reductions. As with the calculation of annual benefits of CO2 emission reductions, the
annual benefits of HFC-134a emission reductions based on the GWP approach are discounted
back to net present value terms using the same discount rate as each SC-CO2 estimate. The
estimated HFC-134a benefits using the GWP approach are presented in Table 8-11.
overestimate of the SC-non- CO2 for near term perturbation years in cases where the SC-CO2 is based on a
sufficiently low or steeply declining discount rate.
GG For example, the 2012 New Source Performance Standards and Amendments to the National Emissions
Standards for Hazardous Air Pollutants for the Oil and Natural Gas Industry are expected to reduce methane
emissions by 900,000 metric tons annually, see http://www.gpo.gov/fdsys/pkg/FR-2012-08-16/pdf/2012-16806.pdf.
Additionally, the 2017-2025 Light-duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel
Economy Standards, promulgated jointly with the National Highway Traffic Safety Administration, is expected to
reduce methane emissions by over 100,000 metric tons in 2025 increasing to nearly 500,000 metric tons in 2050, see
http://www.gpo.gov/fdsys/pkg/FR-2012-10-15/pdf/2012-21972.pdf.

-------
Table 8-11 Annual Upstream and Downstream HFC-134a GHG Benefits and Net Present Values for the
Given SC-CO2 Value for Final Program Relative to the Flat Baseline and using Method B, using the GWP
Approach (Millions of 2013$)a b
CALENDA
5%
3%
2.5%
3%
R
(AVERAGE SC-CO2
(AVERAGE SC-CO2
(AVERAGE SC-CO2
(95™ PERCENTILE
YEAR
=
=
=
=

$12 IN 2012)
$36 IN 2012)
$58 IN 2012)
$100 IN 2012)
2018
$0.0
$0.0
$0.0
$0.0
2019
$0.0
$0.0
$0.0
$0.0
2020
$0.0
$0.0
$0.0
$0.0
2021
$0.2
$0.8
$1.3
$2.5
2022
$0.5
$1.7
$2.6
$5.1
2023
$0.8
$2.6
$3.9
$7.9
2024
$1.1
$3.6
$5.3
$11
2025
$1.4
$4.7
$7.0
$14
2026
$1.7
$5.9
$8.6
$18
2027
$2.2
$7.1
$10
$21
2028
$2.6
$8.3
$12
$25
2029
$2.9
$10
$14
$29
2030
$3.5
$11
$16
$33
2035
$5.0
$15
$22
$47
2040
$6.2
$18
$25
$54
2050
$8.6
$23
$31
$70
NPVb
$44
$200
$320
$620

Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
b The SC-CO2 values are dollar-year and emissions-year specific. Note that discounted values of reduced GHG
emissions are calculated differently than other benefits. The same discount rate used to discount the value of
damages from future emissions (SC-CO2 at 5, 3, and 2.5 percent) is used to calculate discounted values of SC-CO2
for internal consistency. Refer to 2010 SC-CO2 TSD for more detail.
8.5.2.3 Additional non-CCh GHGs Co-Benefits
In determining the relative social costs of the different gases, the Marten et al. (2014)
analysis accounts for differences in lifetime and radiative efficiency between the non-CCh GHGs
and CO2. The analysis also accounts for radiative forcing resulting from methane's effects on
tropospheric ozone and stratospheric water vapor, and for at least some of the fertilization effects
of elevated carbon dioxide concentrations. However, there exist several other differences
between these gases that have not yet been captured in this analysis, for example the non-
radiative effects of methane-driven elevated tropospheric ozone levels on human health,
agriculture, and ecosystems, and the effects of carbon dioxide on ocean acidification. Inclusion
of these additional non-radiative effects would potentially change both the absolute and relative
value of the various gases.
Of these effects, the human health effect of elevated tropospheric ozone levels resulting
from methane emissions is the closest to being monetized in a way that would be comparable to
the SCC. Premature ozone-related cardiopulmonary deaths resulting from global increases in
tropospheric ozone concentrations produced by the methane oxidation process have been the

-------
focus of a number of studies over the past decade (e.g., West et al. 200636). Recent studies have
produced an estimate of a monetized benefit of methane emissions reductions, with results on the
order of $1000 per metric ton of CH4 emissions reduced (Anenberg et al. 201237; Shindell et al.
20 1 238; Sarofim et al. 20 1 539), an estimate similar in magnitude to the climate benefits of CH4
reductions estimated by the Marten et al. or GWP methods. However, though EPA is continuing
to monitor this area of research as it evolves, EPA is not applying them for benefit estimates at
this time.
8.6 Quantified and Monetized Non-GHG Health and Environmental Impacts
This section discusses the economic benefits from reductions in health and environmental
impacts resulting from non-GHG emission reductions that can be expected to occur as a result of
the Phase 2 standards. CO2 emissions are predominantly the byproduct of fossil fuel combustion
processes that also produce criteria and hazardous air pollutant emissions. The vehicles that are
subject to the Phase 2 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 and will also affect emissions from upstream sources that occur during
the refining and distribution of fuel. Changes in ambient concentrations of ozone, PM2.5, and air
toxics that will result from the Phase 2 standards are expected to affect human health by reducing
premature deaths and other serious human health effects, as well as other important
improvements in public health and welfare. Children especially benefit from reduced exposures
to criteria and toxic pollutants, because they tend to be more sensitive to the effects of these
respiratory pollutants. Ozone and particulate matter have been associated with increased
incidence of asthma and other respiratory effects in children, and particulate matter has been
associated with a decrease in lung maturation. Some minority groups and children living under
the poverty line are even more vulnerable with higher prevalence of asthma.
It is important to quantify the health and environmental impacts associated with the
standards because a failure to adequately consider ancillary impacts could lead to an incorrect
assessment of their costs and benefits. Moreover, the health and other impacts of exposure to
criteria air pollutants and airborne toxics tend to occur in the near term, while most effects from
reduced climate change are likely to occur only over a time frame of several decades or longer.
As mentioned in Chapter 7, impacts such as emissions reductions, costs and benefits are
presented in this analysis from two perspectives:
•	A "model year lifetime analysis" (MY), which shows impacts of the program that occur
over the lifetime of the vehicles produced during the model years subject to the Phase 2
standards (MYs 2018 through 2029); and,
•	A "calendar year analysis" (CY), which shows annual costs and benefits of the Phase 2
standards for each year from 2018 through 2050. We assume the standard in the last
model year subject to the standards applies to all subsequent MY fleets developed in the
future.
In previous light-duty and heavy-duty GHG rulemakings, EPA has quantified and
monetized non-GHG health impacts using two different methods. For the MY analysis, EPA
applies PM-related "benefits per-ton" values to the stream of lifetime estimated emission

-------
reductions as a reduced-form approach to estimating the PIVh.s-related benefits of the rule.40'™
For the CY analysis, EPA typically conducts full-scale photochemical air quality modeling to
quantify and monetize the PM2.5- and ozone-related health impacts of a single representative
future year. EPA then assumes these benefits are repeated in subsequent future years when
criteria pollutant emission reductions are equal to or greater than those modeled in the
representative future year.
This two-pronged approach to estimating non-GHG impacts is precipitated by the length
of time needed to prepare the necessary emissions inventories and the processing time associated
with full-scale photochemical air quality modeling for a single representative future year. The
timing requirements (along with other resource limitations) preclude EPA from being able to do
the more detailed photochemical modeling for every year that we include in our benefit and cost
estimates, and require EPA to make air quality modeling input decisions early in the analytical
process. As a result, it was necessary to use emissions from the proposed program to conduct the
air quality modeling for this action.
The chief limitation when using air quality inventories based on emissions from the
proposal in the CY modeling analysis is that they can diverge from the estimated emissions of
the final rulemaking. How much the emissions might diverge and how that difference would
impact the air quality modeling and health benefit results is difficult to anticipate. For the FRM,
EPA concluded that when comparing the proposal and final rule inventories, the differences were
enough to justify the move of the typical CY benefits analysis (based on air quality modeling)
from the primary estimate of costs and benefits to a supplemental analysis in an appendix to the
RIA (See Appendix 8A).n While we believe this supplemental analysis is still illustrative of the
standard's potential benefits, EPA has instead chosen to characterize the CY benefits in a manner
consistent with the MY lifetime analysis. That is, we apply the PM-related "benefits per-ton"
values to the CY final rule emission reductions to estimate the PM-related benefits of the final
rule.
This section presents the benefits-per-ton values used to monetize the benefits from
reducing population exposure to PM associated with the standards. EPA bases its analyses on
peer-reviewed studies of air quality and health and welfare effects and peer-reviewed studies of
the monetary values of public health and welfare improvements, and is generally consistent with
benefits analyses performed for the analysis of the final Tier 3 Vehicle Rule,41 the final 2012 PM
NAAQS Revision,42 and the final 2017-2025 Light Duty Vehicle GHG Rule.43
EPA is also requiring that rebuilt engines installed in new incomplete vehicles (i.e.,
"glider kit" vehicles) meet the emission standards applicable in the year of assembly of the new
vehicle, including all applicable standards for criteria pollutants (Section XIII.B of the
Preamble). For the final rule, EPA has updated its analysis of the environmental impacts of these
glider kit vehicles (see Section XIII.B. 1 of the Preamble). These standards will decrease PM and
1111 See: htto://www3.epa.gov/airqualitv/benmap/sabpt.html. The current values available on the webpage have been
updated since the publication of the Fann et al., 2012 paper. For more information regarding the updated values, see
https://www3.epa.gov/sites/production/files/2014-10/documents/sourceapportionmentbpttsd.pdf (accessed May 2,
2016).
11 Chapter 5 of the RIA discusses the reasons for these differences in more detail.

-------
NOx emissions dramatically, leading to substantial public health-related benefits. Although we
only present these benefits as a sensitivity analysis in Section XIII of the Preamble, it is clear that
removing even a fraction of glider kit vehicles from the road will yield substantial health-related
benefits that are not captured by the primary estimate of monetized non-GHG health impacts
described in this section.
8.6.1 Economic Value of Reductions in Particulate Matter
As described in Chapter 5, the standards will reduce emissions of several criteria and
toxic pollutants and their precursors. In this analysis, EPA only estimates the economic value of
the human health benefits associated with the resulting reductions in PM2.5 exposure. Due to
analytical limitations with the benefit per-ton method, this analysis does not estimate benefits
resulting from reductions in population exposure to other criteria pollutants such as ozone.JJ
Furthermore, the benefits per-ton method, like all air quality impact analyses, does not monetize
all of the potential health and welfare effects associated with reduced concentrations of PM2.5.
This analysis uses estimates of the benefits from reducing the incidence of the specific
PM2.5-related health impacts described below. These estimates, which are expressed per ton of
PM2.5-related emissions eliminated by the final program, represent the total monetized value of
human health benefits (including reduction in both premature mortality and premature
morbidity) from reducing each ton of directly emitted PM2.5, or its precursors (SO2 and NOx ),
from a specified source. Ideally, the human health benefits would be estimated based on changes
in ambient PM2.5 as determined by full-scale air quality modeling. However, the length of time
needed to prepare the necessary emissions inventories, in addition to the processing time
associated with the modeling itself, has precluded us from performing air quality modeling for
this final program.
The PM-related dollar-per-ton benefit estimates used in this analysis are provided in
Table 8-12. As the table indicates, these values differ among pollutants, and also depend on their
original source, because emissions from different sources can result in different degrees of
population exposure and resulting health impacts. In the summary of costs and benefits, Chapter
8.10, EPA presents the monetized value of PM-related improvements associated with the final
program.
11 The air quality modeling that underlies the PM-related benefit per ton values also produced estimates of ozone
levels attributable to each sector. However, the complex non-linear chemistry governing ozone formation prevented
EPA from developing a complementary array of ozone benefit per ton values. This limitation notwithstanding, we
anticipate that the ozone-related benefits associated with reducing emissions of NOx and VOC could be substantial.

-------
Table 8-12 PM-related Benefits-per-ton Values (thousands, 2013$)a
YEAR0
ON-ROAD MOBILE SOURCES
UPSTREAM SOURCES15
Direct PM2 5
S02
NOx
Direct PM2 5
S02
NOx
Estimated Using a 3 Percent Discount Rateb
2016
$380-$870
$20-$46
$7.8-$18
$330-$760
$71-$160
$6.9-$16
2020
$410-$920
$22-$50
$8.2-$18
$350-$800
$76-$ 170
$7.5-$17
2025
$450-$l,000
$25-$56
$9.0-$20
$400-$890
$84-$ 190
$8.2-$18
2030
$490-$l,100
$28-$62
$9.7-$22
$430-$960
$92-$200
$8.9-$20
Estimated Using a 7 Percent Discount Rateb
2016
$340-$780
$18-$42
$7.1-$16
$300-$680
$64-$ 140
$6.3-$14
2020
$370-$830
$20-$45
$7.5-$17
$320-$730
$68-$150
$6.7-$15
2025
$410-$920
$22-$50
$8.1-$18
$350-$800
$76-$ 170
$7.4-$17
2030
$440-$990
$25-$56
$8.8-$20
$380-$870
$82-$ 180
$8.0-$18
Notes:
" The benefit-per-ton estimates presented in this table are based on a range of premature mortality estimates derived
from the ACS study (Krewski et al., 2009) and the Six-Cities study (Lepeule et al., 2012).
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.
0 Benefit-per-ton values were estimated for the years 2016, 2020, 2025 and 2030. We hold values constant for
intervening years (e.g., the 2016 values are assumed to apply to years 2017-2019; 2020 values for years 2021-2024;
2030 values for years 2031 and beyond).
d We assume for the purpose of this analysis that "upstream emissions" are most closely associated with refinery
sector benefit per-ton values. The majority of upstream emission reductions associated with the final rule are related
to domestic onsite refinery emissions and domestic crude production. While upstream emissions also include
storage and transport sources, as well as upstream refinery sources, we have chosen to simply apply the refinery
values.
The benefit per-ton technique has been used in previous analyses, including EPA's 2017-
2025 Light-Duty Vehicle Greenhouse Gas Rule,44 the Reciprocating Internal Combustion Engine
rules,45'46 and the Residential Wood Heaters NSPS.47 Table 8-13 shows the quantified PM2.5-
related co-benefits captured in those benefit per-ton estimates, as well as unquantified effects the
benefits per-ton estimates are unable to capture.
Table 8-13 Human Health and Welfare Effects of PM2.5
POLLUTANT
QUANTIFIED AND MONETIZED
IN PRIMARY ESTIMATES
UNQUANTIFIED EFFECTS
CHANGES IN:
pm25
Adult premature mortality
Acute bronchitis
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
Chronic and subchronic bronchitis cases
Strokes and cerebrovascular disease
Low birth weight
Pulmonary function
Chronic respiratory diseases other than chronic
bronchitis
Non-asthma respiratory emergency room visits
Visibility
Household soiling

-------
Consistent with the cost-benefit analysis that accompanied the 2012 PM NAAQS
revision, the benefits estimates utilize the concentration-response functions as reported in the
epidemiology literature. KK'48 To calculate the total monetized impacts associated with
quantified health impacts, EPA applies values derived from a number of sources. For premature
mortality, EPA applies a value of a statistical life (VSL) derived from the mortality valuation
literature. For certain health impacts, such as respiratory-related ailments, EPA applies
willingness-to-pay estimates derived from the valuation literature. For the remaining health
impacts, EPA applies values derived from current cost-of-illness and/or wage estimates.
Readers interested in reviewing the complete methodology for creating the benefit-per-
ton estimates used in this analysis can consult EPA's "Technical Support Document: Estimating
the Benefit per Ton of Reducing PM2.5 Precursors from 17 Sectors."LL Readers can also refer to
Fann et al. (2012)49 for a detailed description of the benefit-per-ton methodology.
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 on-road mobile sources; direct PM emitted from
electricity generating units). 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.
As Table 8-12 indicates, EPA projects that the per-ton values for reducing emissions of
non-GHG pollutants from both vehicle use and upstream sources such as fuel refineries will
increase over time.MM These projected increases reflect rising income levels, which increase
affected individuals' willingness to pay for reduced exposure to health threats from air
pollution.^ They also reflect future population growth and increased life expectancy, which
expands the size of the population exposed to air pollution in both urban and rural areas,
especially among older age groups with the highest mortality risk.00
KK Although we summarize the main issues in this chapter, we encourage interested readers to see the benefits
chapter of the RIA that accompanied the PM NAAQS for a more detailed description of recent changes to the
quantification and monetization of PM benefits. Note that the cost-benefit analysis was prepared solely for purposes
of fulfilling analysis requirements under Executive Order 12866 and was not considered, or otherwise played any
part, in the decision to revise the PM NAAQS.
LL For more information regarding the updated values, see:
http://www3.epa.gov/airquality/benmap/models/Source_Apportionment_BPT_TSD_l_3 l_13.pdf (accessed
September 9, 2014).
VIVI As we discuss in the emissions chapter (Chapter 5), the rule would yield emission reductions from upstream
refining and fuel distribution due to decreased petroleum consumption.
The issue is discussed in more detail in the 2012 PM NAAQS RIA, Section 5.6.8. See U.S. Environmental
Protection Agency. (2012). Regulatory Impact Analysis for the Final Revisions to the National Ambient Air Quality
Standards for Particulate Matter, Health and Environmental Impacts Division, Office of Air Quality Planning and
Standards, EPA-452-R-12-005, December 2012. Available on the internet:
http://www3.epa.gov/ttnecasl/regdata/RIAs/finalria.pdf.
00 For more information about EPA's population projections, please refer to the following:
http://www3.epa.gov/air/benmap/models/BenMAPManualAppendicesAugust2010.pdf (See Appendix K).

-------
The benefit-per-ton estimates are subject to a number of assumptions and uncertainties:
•	The benefit-per-ton estimates used here reflect specific geographic patterns of
emissions reductions and specific air quality and benefits modeling assumptions
associated with the derivation of those estimates (see the TSD describing the
calculation of the national benefit-per-ton estimates). 50'pp Consequently, these
estimates may not reflect local variability in population density, meteorology,
exposure, baseline health incidence rates, or other local factors associated with the
current analysis. Therefore, use of these benefit-per-ton values to estimate non-
GHG benefits may lead to higher or lower benefit estimates than if these benefits
were calculated based on direct air quality modeling.
•	This analysis assumes that all fine particles, regardless of their chemical
composition, are equally potent in causing premature mortality. This is an
important assumption, because PM2.5 produced via transported precursors emitted
from stationary sources may differ significantly from direct PM2.5 released from
diesel engines and other industrial sources. The PM ISA, which was twice
reviewed by SAB-CASAC, concluded that "many constituents of PM2.5 can be
linked with multiple health effects, and the evidence is not yet sufficient to allow
differentiation of those constituents or sources that are more closely related to
specific outcomes."51 PM composition and the size distribution of those particles
vary within and between areas due to source characteristics. Any specific
location could have higher or lower contributions of certain PM species and other
pollutants than the national average, meaning potential regional differences in
health impact of given control strategies. Depending on the toxicity of each PM
species reduced by the standards, assuming equal toxicity could over or
underestimate benefits.
•	This analysis assumes that the health impact function for fine particles is linear
within the range of ambient concentrations under consideration. Thus, the
estimates include health benefits from reducing fine particles in areas with varied
concentrations of PM2.5, including regions that are in attainment with the fine
particle standard. The direction of bias that assuming linear-no threshold model
or alternative model introduces depends upon the "true" functional from of the
relationship and the specific assumptions and data in a particular analysis. For
example, if the true function identifies a threshold below which health effects do
not occur, benefits may be overestimated if a substantial portion of those benefits
were estimated to occur below that threshold. Alternately, if a substantial portion
of the benefits occurred above that threshold, the benefits may be underestimated
because an assumed linear no-threshold function may not reflect the steeper slope
above that threshold to account for all health effects occurring above that
threshold.
pp See also: http://www3.epa.gov/airqualitv/benmap/sabpt.html. The current values available on the webpage have
been updated since the publication of the Fann et al., 2012 paper. For more information regarding the updated
values, see: http://www3.epa.gov/airqualitv/benmap/models/Source Apportionment BPT TSD 1 31 13.pdf
(accessed September 9, 2014).

-------
•	There are several health benefit categories that EPA was unable to quantify due to
limitations associated with using benefits-per-ton estimates, several of which
could be substantial. Because the NOx and VOC emission reductions associated
with the final program are also precursors to ozone, reductions in NOx and VOC
would also reduce ozone formation and the health effects associated with ozone
exposure. Unfortunately, ozone-related benefits-per-ton estimates do not exist
due to issues associated with the complexity of the atmospheric air chemistry and
nonlinearities associated with ozone formation. The PM-related benefits-per-ton
estimates also do not include any human welfare or ecological benefits.
•	There are many uncertainties associated with the health impact functions that
underlie the benefits-per-ton estimates. These include: within-study variability
(the precision with which a given study estimates the relationship between air
quality changes and health effects); across-study variation (different published
studies of the same pollutant/health effect relationship typically do not report
identical findings and in some instances the differences are substantial); the
application of concentration-response functions nationwide (does not account for
any relationship between region and health effect, to the extent that such a
relationship exists); extrapolation of impact functions across population (we
assumed that certain health impact functions applied to age ranges broader than
that considered in the original epidemiological study); and various uncertainties in
the concentration-response function, including causality and thresholds. These
uncertainties may under- or over-estimate benefits.
•	EPA has investigated methods to characterize uncertainty in the relationship
between PM2.5 exposure and premature mortality. EPA's final PM2.5 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.52
•	The benefit-per-ton unit values used in this analysis incorporate projections of key
variables, including atmospheric conditions, source level emissions, population,
health baselines, incomes, and technology. These projections introduce some
uncertainties to the benefit per ton estimates.
8.6,2 Unquantified Health and Environmental Impacts
In addition to the co-pollutant health impacts EPA quantifies in this analysis, there are a
number of other health and human welfare endpoints that we will not be able to quantify because
of current limitations in the methods or available data. These impacts are associated with
emissions of air toxics (including benzene, 1,3-butadiene, formaldehyde, acetaldehyde, acrolein,
and ethanol), ambient ozone, and ambient PM2.5 exposures. For example, we have not quantified
a number of known or suspected health effects linked with ozone and PM for which appropriate
health impact functions are not available or which do not provide easily interpretable outcomes
(i.e., changes in heart rate variability). In addition, we are currently 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.

-------
Although there will be impacts associated with air toxic pollutant emission changes that
result from this action, 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). EPA's 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.53 While EPA has since improved these
tools, there remain critical limitations for estimating incidence and assessing benefits of reducing
mobile source air toxics.
As part of the second prospective analysis of the benefits and costs of the Clean Air
Act,54 EPA conducted a case study analysis of the health effects associated with reducing
exposure to benzene in Houston from implementation of the Clean Air Act. While reviewing the
report, EPA's Advisory Council on Clean Air Compliance Analysis concluded that "the
challenges for assessing progress in health improvement as a result of reductions in emissions of
hazardous air pollutants (HAPs) are daunting...due to a lack of exposure-response functions,
uncertainties in emissions inventories and background levels, the difficulty of extrapolating risk
estimates to low doses and the challenges of tracking health progress for diseases, such as
cancer, that have long latency periods."55 EPA continues to work to address these limitations;
however, EPA did not have the methods and tools available for national-scale application in time
for the analysis of the final action.QQ
8.7 Additional Impacts
8.7.1 Cost of Noise, Congestion, and Crashes
Chapter 8.3 discusses the likely sign of the rebound effect. If net operating costs of the
vehicle decline, then we expect a positive rebound effect. Increased vehicle use associated with
a positive rebound effect also contributes to increased traffic congestion, motor vehicle crashes,
and highway noise. Depending on how the additional travel is distributed throughout the day
and on where it takes place, additional vehicle use can contribute to traffic congestion and delays
by increasing traffic volumes on facilities that are already heavily traveled during peak periods.
These added delays impose higher costs on drivers and other vehicle occupants in the form of
increased travel time and operating expenses. Because drivers do not take these added costs into
account in deciding when and where to travel, they must be accounted for separately as a cost of
the added driving associated with the rebound effect.
QQ In April, 2009, EPA hosted a workshop on estimating the benefits or reducing hazardous air pollutants. This
workshop built upon the work accomplished in the June 2000 Science Advisory Board/EPA Workshop on the
Benefits of Reductions in Exposure to Hazardous Air Pollutants, which generated thoughtful discussion on
approaches to estimating human health benefits from reductions in air toxics exposure, but no consensus was
reached on methods that could be implemented in the near term for a broad selection of air toxics. Please visit
http://epa.gov/air/toxicair/2009workshop.html for more information about the workshop and its associated materials.

-------
Our approach in this final rule is identical to that used in the proposal. EPA and NHTSA
rely on estimates of congestion, crash, and noise costs caused by pickup trucks and vans, single
unit trucks, buses, and combination tractors developed by the Federal Highway Administration to
estimate the increased external costs caused by added driving due to the rebound effect.56 The
FHWA estimates are intended to measure the increases in costs from added congestion, property
damages and injuries in traffic crashes, and noise levels caused by various classes of trucks that
are borne by persons other than their drivers (or "marginal" external costs). EPA and NHTSA
employed estimates from this source previously in the analysis accompanying the light-duty
2012-2016 vehicle rulemaking. The agencies continue 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.
FHWA's congestion cost estimates for trucks, which are weighted averages based on the
estimated fractions of peak and off-peak freeway travel for each class of trucks, already account
for the fact that trucks make up a smaller fraction of peak period traffic on congested roads
because they try to avoid peak periods when possible. FHWA's congestion cost estimates focus
on freeways because non-freeway effects are less serious due to lower traffic volumes and
opportunities to re-route around the congestion. The agencies, however, applied the congestion
cost to the overall VMT increase, though the fraction of VMT on each road type used in MOVES
range from 27 to 29 percent of the vehicle miles on freeways for vocational vehicles and 53
percent for combination tractors. The results of this analysis potentially overestimate the
congestions costs associated with increased truck use, and thus lead to a conservative estimate of
benefits.
EPA and NHTSA estimated the costs of additional vocational vehicle travel using a
weighted average of 15 percent of the FHWA estimate for bus costs and 85 percent of the
FHWA estimate for single unit truck costs to reflect the make-up of this segment. The low, mid,
and high cost estimates from FHWA updated to 2012 dollars are included in Table 8-14.
Table 8-14 Low-Mid-High Cost Estimates (2013$/mile)
NOISE

High
Middle
Low
Pickup Truck, Van
$0,002
$0,001
$0,000
Vocational Vehicle
$0,024
$0,009
$0,003
Combination Tractor
$0,055
$0,021
$0,006
Crashes

High
Middle
Low
Pickup Truck, Van
$0,088
$0,028
$0,015
Vocational Vehicle
$0,051
$0,017
$0,009
Combination Tractor
$0,074
$0,023
$0,011
Congestion

High
Middle
Low
Pickup Truck, Van
$0,153
$0,052
$0,014
Vocational Vehicle
$0,350
$0,119
$0,032
Combination Tractor
$0,337
$0,115
$0,030
The agencies are using FHWA's "Middle" estimates for marginal congestion, crash, and
noise costs caused by increased travel from trucks.57 This approach is consistent with the

-------
methodology used in the HD GHG Phase 1 rule and both LD GHG rules. These costs are
multiplied by the annual increases in vehicle miles travelled from the rebound effect to yield the
estimated increases in congestion, crash, and noise externality costs during each future year. The
results are shown in Table 8-15 through Table 8-17.
Table 8-15 Annual Costs & Net Present Values Associated with Increased Noise, Crashes and Congestion for
the Final Program Relative to the Flat Baseline and using Method B (Millions of 2013$)a
CALENDAR
YEAR
HD PICKUP AND
VANS
VOCATIONAL
TRACTOR/TRAILER
SUM
2018
$0
$0
$0
$0
2019
$0
$0
$0
$0
2020
$0
$0
$0
$0
2021
$17
$4
$77
$99
2022
$34
$8
$97
$139
2023
$51
$12
$116
$178
2024
$67
$16
$134
$216
2025
$82
$20
$150
$252
2026
$97
$23
$165
$285
2027
$111
$26
$179
$317
2028
$124
$30
$192
$345
2029
$136
$32
$203
$372
2030
$147
$35
$214
$396
2035
$188
$45
$255
$487
2040
$206
$50
$285
$541
2050
$218
$57
$329
$604
NPV, 3%
$2,462
$599
$3,694
$6,755
NPV, 7%
$1,100
$266
$1,704
$3,070
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
Table 8-16 Discounted Model Year Lifetime Costs Associated with Increased Noise, Crashes and Congestion
for the Final Program Relative to the Flat Baseline and using Method B (3% discount rate, Millions of 2013$)
a
MODEL
YEAR
HD PICKUP AND
VANS
VOCATIONAL
TRACTOR/TRAILER
SUM
2018
$0
$0
$124
$124
2019
$0
$0
$140
$140
2020
$0
$0
$158
$158
2021
$141
$32
$170
$343
2022
$136
$31
$166
$333
2023
$132
$30
$161
$323
2024
$129
$30
$160
$319
2025
$127
$29
$157
$313
2026
$124
$28
$153
$305
2027
$121
$28
$149
$297
2028
$117
$27
$145
$289
2029
$114
$26
$142
$283
Sum
$1,140
$261
$1,825
$3,227
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 8-17 Discounted Model Year Lifetime Costs Associated with Increased Noise, Crashes and Congestion
for the Final Program Relative to the Flat Baseline and using Method B (7% discount rate, Millions of 2013$)
a
MODEL
YEAR
HD PICKUP AND
VANS
VOCATIONAL
TRACTOR/TRAILER
SUM
2018
$0
$0
$80
$80
2019
$0
$0
$89
$89
2020
$0
$0
$100
$100
2021
$88
$20
$106
$215
2022
$82
$19
$100
$201
2023
$76
$18
$93
$187
2024
$72
$17
$90
$178
2025
$68
$16
$84
$168
2026
$64
$15
$79
$158
2027
$60
$14
$74
$148
2028
$56
$13
$70
$139
2029
$53
$12
$66
$131
Sum
$619
$143
$1,030
$1,793
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
8.7,2 Benefits of Reduced Refueling Time
Reducing the fuel consumption of heavy-duty trucks will either increase their driving
range before they require refueling, or lead truck manufacturers to offer, and truck purchasers to
buy, smaller fuel tanks. Keeping the fuel tank the same size will allow truck operators to reduce
the frequency with which drivers typically refuel their vehicles, by extending the upper limit on

-------
the distance they can travel before requiring refueling. Alternatively, if truck purchasers and
manufacturers respond to improved fuel economy by reducing the size of fuel tanks, the smaller
tank will require less time to fill during each refueling stop.
Because refueling time represents a time cost of truck operation, these time savings
should be incorporated into truck purchasers' decisions about how much fuel-saving technology
they purchase as part of their choices of new vehicles. The savings calculated here thus raise the
same questions discussed in Preamble Section IX VIII. A and RIA Chapter 8.2: does the
apparent existence of these savings reflect failures in the market for fuel economy, or does it
reflect costs that are not addressed in this analysis? The response to these questions could vary
across truck segment.
No direct estimates of the value of extended vehicle range or reduced fuel tank size are
readily available. Instead, this analysis calculates the reduction in the annual amount of time a
driver of each type of truck would spend filling its fuel tank; this reduced time could result either
from fewer refueling events, if new trucks' fuel tanks stay the same size, or from less time spent
filling the tank during each refueling stop, if new trucks' fuels tank are made proportionately
smaller. As discussed in Chapter 8.3 in this RIA, the average number of miles each type of
vehicle is driven annually would likely increase under the regulation, as truck operators respond
to lower fuel expenditures (the "rebound effect"). The estimates of refueling time with the rule
in effect allow for this increase in truck use. However, the estimate of the rebound effect does
not account for any reduction in net operating costs from lower refueling time. Because the
rebound effect should measure the change in VMT with respect to the net change in overall
operating costs, refueling time costs would ideally factor into this calculation. The effect of this
omission is expected to be minor because refueling time savings are generally small relative to
the value of reduced fuel expenditures.
Our approach to calculating refueling savings in this final rule is identical to the approach
used in the proposal. The savings in refueling time are calculated as the total amount of time the
driver of a typical truck in each class would save each year as a consequence of pumping less
fuel into the vehicle's tank. The calculation also includes a fixed time per refill event of 3.5
minutes which would not occur as frequently due to the fewer number of refills.
The calculation uses the reduced number of gallons consumed by truck type and divides
that value by the tank volume and refill amount to get the number of refills, then multiplies that
by the time per refill to determine the number of hours saved in a given year. The calculation
then applies DOT-recommended values of travel time savings to convert the resulting time
savings to their economic value. The input metrics used in the analysis are included in Table
8-18. The equation for the calculation is shown below:
_	_ f G(llreference ^^lpolicy\ ( Gal per refill	\ ( $ \
Refueling Benefit = 			—	 x	1- time per refill x —
y Gal per refill J \Fuel dispense rate	) V"r/;afcor
The annual impacts associated with reduced refueling time are shown in Table 8-19 and the MY
lifetime impacts are shown in Table 8-20 and Table 8-21.

-------
Table 8-18 Inputs to Calculate Refueling Time Savings

HD PICKUP AND VAN
VOCATIONAL
VEHICLE
TRACTOR
Fuel Dispensing Rate
(gallon/minute)58
10
10
20
Refueling fixed time
(minutes/refill)59
3.5
3.5
3.5
Tank volume (gallons)3
30
40
200
Refill amount
(%volume/refill)a
60%
75%
75%
Resultant time/refill
(minutes/refill)
5.3
6.5
11.0
Wage rate
(2012$/hr)60'b
$27.22
31.01
28.56
Notes:
" HD pickup and van values based on a NHTSA survey, other are estimated.
b A wage growth rate of 1.2% has been assumed for future years.
Table 8-19 Annual Refueling Benefits and Net Present Values for the Final Program Relative to the Flat
Baseline and using Method B
(Dollar Values in Millions of 2013$)a
CALENDAR
YEAR
HD
PICKUP
AND
VANS
VOCATIONAL
TRACTOR/TRAILER
SUM OF BENEFITS
2018
$0
$0
$1
$1
2019
$0
$0
$3
$3
2020
$0
$0
$5
$5
2021
$3
$9
$14
$27
2022
$11
$19
$27
$56
2023
$23
$28
$40
$91
2024
$41
$43
$60
$144
2025
$63
$58
$81
$202
2026
$90
$73
$101
$264
2027
$122
$93
$128
$342
2028
$153
$112
$154
$420
2029
$184
$131
$180
$495
2030
$214
$150
$205
$570
2035
$344
$231
$321
$895
2040
$434
$293
$415
$1,141
2050
$542
$386
$569
$1,497
NPV, 3%
$4,444
$3,119
$4,422
$11,985
NPV, 7%
$1,814
$1,290
$1,821
$4,925
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
Table 8-20 Discounted Model Year Lifetime Refueling Benefits at 3% for the Final Program Relative to the
Flat Baseline and using Method B (Millions of 2013$)a
MODEL
HD PICKUP
VOCATIONAL
TRACTOR/
SUM
YEAR
AND VANS

TRAILER

2018
$0
$0
$9
$9
2019
$0
$0
$9
$9
2020
$0
$0
$8
$8
2021
$25
$82
$111
$218
2022
$66
$80
$109
$255
2023
$104
$78
$107
$290
2024
$142
$119
$167
$428
2025
$178
$119
$165
$461
2026
$212
$117
$162
$491
2027
$243
$154
$212
$609
2028
$239
$153
$209
$601
2029
$235
$151
$208
$594
Sum
$1,445
$1,054
$1,478
$3,976
Note:
a For an explanation of analytical Methods A and B, please see Preamble
Section I.D; for an explanation of the flat baseline, la, and dynamic
baseline, lb, please see Preamble Section X.A.I
Table 8-21 Discounted Model Year Lifetime Refueling Benefits at 7% for the Final Program Relative to the
Flat Baseline and using Method B (Millions of 2013$)a
MODEL
HD PICKUP
VOCATIONAL
TRACTOR/
SUM
YEAR
AND VANS

TRAILER

2018
$0
$0
$7
$7
2019
$0
$0
$6
$6
2020
$0
$0
$6
$6
2021
$15
$51
$68
$135
2022
$39
$48
$65
$152
2023
$60
$45
$61
$166
2024
$78
$66
$92
$236
2025
$94
$63
$88
$245
2026
$108
$60
$83
$251
2027
$120
$76
$104
$300
2028
$114
$73
$99
$285
2029
$108
$69
$95
$272
Sum
$737
$551
$773
$2,061
Note:
a For an explanation of analytical Methods A and B, please see Preamble
Section I.D; for an explanation of the flat baseline, la, and dynamic
baseline, lb, please see Preamble Section X.A.I
8.7.3 Benefits of Increased Travel Associated with Rebound Driving
The increase in travel associated with the rebound effect produces additional benefits to
vehicle owners and operators, which reflect the value of the added (or more desirable) social and
economic opportunities that become accessible with additional travel. The analysis estimates the
economic benefits from increased rebound-effect driving as the sum of fuel expenditures

-------
incurred plus the vehicle owner/operator surplus from the additional accessibility it provides. As
evidenced by the fact that vehicles make more frequent or longer trips when the cost of driving
declines, the benefits from this added travel exceed added expenditures for the fuel consumed.
Note that the amount by which the benefits from this increased driving exceed its increased fuel
costs measures the net benefits from the additional travel, usually referred to as increased
consumer surplus or, in this case, increased owner/operator surplus. The equation for the
calculation of the total travel benefit is shown below:
The analysis in this final rule is identical to that used in the proposal. The agencies'
analysis estimates the economic value of the increased owner/operator surplus provided by added
driving using the conventional approximation, which is one half of the product of the decline in
vehicle operating costs per vehicle-mile and the resulting increase in the annual number of miles
driven. Because it depends on the extent of improvement in fuel economy, the value of benefits
from increased vehicle use changes by model year and varies among alternative standards.
Under even those alternatives that will impose the highest standards, however, the magnitude of
the surplus from additional vehicle use represents a small fraction of this benefit. The benefits
are shown in Table 8-22 through Table 8-24
Table 8-22 Annual Value of Increased Travel and Net Present Values at 3% and 7% Discount Rates for the
Final Program Relative to the Flat Baseline and using Method B (Millions of 2013$)a
CALENDAR
YEAR
HD PICKUP AND
VANS
VOCATIONAL
TRACTOR/TRAILER
SUM
2018
$0
$0
$0
$0
2019
$0
$0
$0
$0
2020
$0
$0
$0
$0
2021
$43
$9
$247
$298
2022
$86
$18
$314
$417
2023
$128
$27
$379
$534
2024
$171
$36
$442
$648
2025
$212
$45
$502
$759
2026
$253
$53
$559
$866
2027
$292
$62
$613
$967
2028
$330
$70
$664
$1,064
2029
$367
$78
$712
$1,157
2030
$402
$85
$759
$1,247
2035
$558
$120
$982
$1,660
2040
$678
$149
$1,215
$2,043
2050
$721
$169
$1,394
$2,284
NPV, 3%
$7,427
$1,627
$14,303
$23,357
NPV, 7%
$3,232
$701
$6,410
$10,343
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Travel Benefit = (VMTrebound)	+ Q (VMTrebound)
policy
reference
policy

-------
Table 8-23 Discounted Model Year Lifetime Value of Increased Travel for the Final Program Relative to the
Flat Baseline and using Method B (3% discount rate, Millions of 2013$)a
MODEL
YEAR
HD PICKUP AND
VANS
VOCATIONAL
TRACTOR/TRAILER
SUM
2018
$0
$0
$452
$452
2019
$0
$0
$511
$511
2020
$0
$0
$580
$580
2021
$383
$77
$594
$1,054
2022
$372
$76
$590
$1,038
2023
$362
$74
$583
$1,020
2024
$357
$73
$572
$1,001
2025
$351
$73
$570
$994
2026
$346
$72
$564
$982
2027
$338
$70
$542
$951
2028
$335
$70
$538
$942
2029
$331
$70
$536
$937
Sum
$3,174
$655
$6,633
$10,462
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
Table 8-24 Discounted Model Year Lifetime Value of Increased Travel for the Final Program Relative to the
Flat Baseline and using Method B (7% discount rate, Millions of 2013$)a
MODEL
YEAR
HD PICKUP AND
VANS
VOCATIONAL
TRACTOR/TRAILER
SUM
2018
$0
$0
$285
$285
2019
$0
$0
$319
$319
2020
$0
$0
$358
$358
2021
$236
$47
$364
$647
2022
$220
$45
$348
$613
2023
$206
$43
$331
$580
2024
$196
$40
$313
$549
2025
$186
$39
$301
$525
2026
$176
$37
$287
$500
2027
$166
$35
$266
$466
2028
$158
$33
$254
$445
2029
$151
$32
$244
$427
Sum
$1,694
$351
$3,671
$5,715
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I

-------
8.8 Petroleum, Energy and National Security Impacts
8,8,1 Energy Security Impacts
The Phase 2 standards are designed to require improvements in the fuel efficiency of
medium- and heavy-duty vehicles and, thereby, reduce fuel consumption and GHG emissions.
In turn, the Phase 2 standards help to reduce U.S. petroleum imports. A reduction of U.S.
petroleum imports reduces both financial and strategic risks caused by potential sudden
disruptions in the supply of imported petroleum to the U.S. This reduction in risk increases U.S.
energy security. This section summarizes the agency's estimates of U.S. oil import reductions
and energy security benefits of the final Phase 2 standards. Additional discussion of this issue
can be found in Section IX.I of the Preamble.
The U.S., as a large oil importer and oil consumer, is economically vulnerable to
outcomes in a volatile global oil market that relies on oil supplies from potentially unstable
sources. Much of the world's oil and gas supplies are located in countries facing social,
economic, and demographic challenges, thus making them vulnerable to potential local
instability. In 2010, just over 40 percent of world oil supply came from OPEC (e.g.,
Organization of Petroleum Exporting Countries) nations and the Annual Energy Outlook 2015
projects that that this share will stay high; dipping slightly from 37 percent by 2020 and then
rising gradually to over 40 percent by 2035 and thereafter.1®
Approximately 30 percent of global supply is from Middle East and North African
countries alone, a share that is expected to grow.ss Measured in terms of the share of world oil
resources or the share of global oil export supply, rather than oil production, the concentration of
global petroleum resources in OPEC nations is even larger. As another measure of
concentration, of the 137 countries/principalities that export either crude or product, the top 12
have recently accounted for over 55 percent of exports.61 Eight of these countries are members
of OPEC, and a ninth is Russia.TT In a market where even a 1-2 percent supply loss raises prices
noticeably, and where a 10 percent supply loss could lead to an unprecedented price shock, this
regional concentration is of concern. ^ Historically, the countries of the Middle East have been
1111 The agencies used the AEO 2015 since this version of AEO was available at the time that fuel savings from the
rule were being estimated.
ss Middle East and North African oil supply share reaches over 40 percent in 2040 in the AEO 2015 Reference Case.
TT The other three are Norway, Canada, and the EU, an exporter of product.
1111 For example, the 2005 Hurricanes Katrina/Rita and the 2011 Libyan conflict both led to a 1.8 percent reduction in
global crude supply. While the price impact of the latter is not easily distinguished given the rapidly rising post-
recession prices, the former event was associated with a 10-15 percent world oil price increase. There are a range of
smaller events with smaller but noticeable impacts. Somewhat larger events, such as the 2002/3 Venezuelan Strike
and the War in Iraq, corresponded to about a 2.9 percent sustained loss of supply, and was associated with a 28
percent world oil price increase. (Compiled from EIA oil price data, IEA2012 [IEA Response System for Oil Supply
Emergencies] (http://www.iea.org/publications/freepublications/publication/EPPD Brochure English 2012 02.pdf)
See table onP. 11. and Hamilton 2011 "Historical Oil Shocks."dittp://econweb.ucsd.edu/~ihamilto/oil historv.pdf)
in *Routledge Handbook of Major Events in Economic History*, pp. 239-265, edited by Randall E. Parker and
Robert Whaples, New York: Routledge Taylor and Francis Group, 2013).

-------
the source of eight of the ten major world oil disruptions, with the ninth originating in
Venezuela, an OPEC country, and the tenth being Hurricanes Katrina and Rita.vv
One impact of the final Phase 2 program is that it promotes more efficient use of
transportation fuels in the U.S. The result is that it reduces U.S. oil consumption and imports,
which reduces both financial and strategic risks associated with a potential disruption in supply
or a spike in the cost of a particular energy source. This reduction in risks increases U.S. energy
security. For this rule, an "oil premium" approach is utilized that identifies those energy security
related economic costs which are not reflected in the market price of oil, and which are expected
to change in response to an incremental change in the level of U.S. oil imports.
8.8.2 Impact on U.S. Petroleum Imports
U.S. energy security is generally considered as the continued availability of energy
sources at an acceptable price. Most discussion of U.S. energy security revolves around the topic
of the economic costs of U.S. dependence on oil imports. The U.S.'s energy security problem is
that the U.S. relies on imported oil from potentially unstable sources. In addition, oil exporters
have the ability to raise the price of oil by exerting monopoly power through the formation of a
cartel, the Organization of Petroleum Exporting Countries (OPEC). These factors contribute to
the vulnerability of the U.S. economy to episodic oil supply shocks and price spikes. In 2014,
U.S. expenditures for imports of crude oil and petroleum products, net of revenues for exports,
were $178 billion, and total consumption expenditure was $469 billion (in 2013$) (see Figure 8-
3).62 Recently, as a result of strong growth in domestic oil production mainly from tight shale
formations, U.S. production of oil has increased while U.S. oil imports have decreased. For
example, from 2012 to 2015, domestic oil production increased by 44 percent while oil net
imports and products decreased by 38 percent. While U.S. oil import costs have declined since
2011, total oil expenditures (domestic and imported) remained near historical highs through
2014. Post-2015 oil expenditures are projected (AEO 2015) to remain between double and triple
the inflation-adjusted levels experienced by the U.S. from 1986 to 2002.
vv The events IEA categorized as oil supply disruptions all had a gross peak oil supply loss of at least 1.5 million
barrels a day as a result of wars, revolutions, embargoes or strikes involving major oil exporting nations or from
major storm events or disasters (like the double Hurricane Katrina/Rita) affecting oil producing/processing regions.
IEA 2011 "IEA Response System for Oil Supply Emergencies."

-------
U.S. Expenditures on Crude Oil
800
"00
600
S? ?oo
®

-------
Table 8-25 below. For comparison purposes, Table 8-25 also shows U.S. imports and exports of
crude oil in 2020, 2025, 2030 and 2040 as projected by DOE in the Annual Energy Outlook 2015
(Reference Case). U.S. Gross Domestic Product (GDP) is projected to grow by roughly 48
percent (2009$) between 2020-2040 in the AEO 2015 projections.
Table 8-25 Projected U.S. Exports and Imports of Oil and U.S. Oil Import Reductions
in 2020,2025,2030,2040 and 2050 for the Final Program Relative to the Flat Baseline and using Method B
(Millions of barrels per day (MMBD))a
YEAR
U.S OIL
EXPORTS
U.S. OIL
IMPORTS
U.S. NET
PRODUCT
IMPORTS*
U.S. NET
CRUDE &
PRODUCT
IMPORTS
REDUCTIONS FROM
HD RULES
2020
0.63
6.14
-2.80
2.71
0.007
2025
0.63
6.72
-3.24
2.85
0.162
2030
0.63
7.07
-3.56
2.88
0.405
2040
0.63
8.21
-4.26
3.32
0.721
2050
**
**
**
**
0.861
Notes:
* Negative U.S. Net Product Imports imply positive exports.
** The AEO 2015 only projects energy market and economic trends through 2040.
8.8.3 Methodology Used to Estimate U.S. Energy Security Benefits
In order to understand the energy security implications of reducing U.S. oil imports, EPA
has worked with Oak Ridge National Laboratory (ORNL), which has developed approaches for
evaluating the social costs and energy security implications of oil use. The energy security
estimates provided below are based upon a methodology developed in a peer-reviewed study
entitled, "The Energy Security Benefits of Reduced Oil Use, 2006-2015", completed in March
2008. This ORNL study is an updated version of the approach used for estimating the energy
security benefits of U.S. oil import reductions developed in a 1997 ORNL Report.64 For EPA
and NHTSA rulemakings, the ORNL methodology is updated periodically to account for
forecasts of future energy market and economic trends reported in the U.S. Energy Information
Administration's Annual Energy Outlook.
As part of the process for developing the ORNL energy security estimates, EPA
sponsored an independent, expert peer review of the 2008 ORNL study.65 In addition, EPA
worked with ORNL to address comments raised in the peer review and to develop estimates of
the energy security benefits associated with a reduction in U.S. oil imports. In response to peer
reviewer comments, ORNL modified its model by changing several key parameters involving
OPEC supply behavior, the responsiveness of oil demand and supply to a change in the world oil
price, and the responsiveness of U.S. economic output to a change in the world oil price.
When conducting this analysis, ORNL considered the full cost of importing petroleum
into the U.S. The full economic cost is defined to include two components in addition to the
purchase price of petroleum itself. These are: (1) the higher costs for oil imports resulting from
the effect of U.S. demand on the world oil price (i.e., the "demand" or "monopsony" costs); and

-------
(2) the risk of reductions in U.S. economic output and disruption to the U.S. economy caused by
sudden disruptions in the supply of imported oil to the U.S. (i.e., macroeconomic
disruption/adjustment costs).
The literature on energy security for the last two decades has routinely combined the
monopsony and the macroeconomic disruption components when calculating the total value of
the energy security premium. However, in the context of using a global value for the Social Cost
of Carbon (SCC) the question arises: how should the energy security premium be used when
some benefits from the rule, such as the benefits of reducing greenhouse gas emissions, are
calculated from a global perspective? Monopsony benefits represent avoided payments by U.S.
consumers to oil producers that result from a decrease in the world oil price as the U.S. decreases
its demand for oil. Although there is clearly an overall benefit to the U.S. when considered from
a domestic perspective, the decrease in price due to decreased demand in the U.S. also represents
a loss to oil producing countries, one of which is the U.S. Given the redistributive nature of this
monopsony effect from a global perspective, it is excluded in the energy security benefits
calculations for this final program.
In contrast, the other portion of the energy security premium, the avoided U.S.
macroeconomic disruption and adjustment cost that arises from reductions in U.S. petroleum
imports, does not have offsetting impacts outside of the U.S., and, thus, is included in the energy
security benefits estimated for these final rules. To summarize, the agencies have included only
the avoided macroeconomic disruption portion of the energy security benefits to estimate the
monetary value of the total energy security benefits of these final rules.
For this rulemaking, ORNL updated the energy security premiums by incorporating the
most recent oil price forecast and energy market trends, particularly regional oil supplies and
demands, from the AEO 2015 into its model.66 Table 8-26 provides estimates for energy
security premiums for the years 2020, 2025, 2030 and 2040^, as well as a breakdown of the
components of the energy security premiums for each year. The components of the energy
security premiums and their values are discussed below.
xx AEO 2015 forecasts energy market trends and values only to 2040. The post-2040 energy security premium
values are assumed to be equal to the 2040 estimate.

-------
Table 8-26 Energy Security Premiums in 2020,2025,2030 and 2040 (2013$/Barrel)*
YEAR
MONOPSONY
AVOIDED MACROECONOMIC
TOTAL MID-POINT
(RANGE)
(RANGE)
DISRUPTION/ADJUSTMENT
COSTS
(RANGE)
(RANGE)
2020
$2.21
$5.48
$7.69

($0.65 - $3.59)
($2.51 -$8.92)
($4.54 -$11.14)
2025
$2.59
$6.30
$8.89

($0.76 -$4.14)
($2.92 -$10.22)
($5.22 -$12.83)
2030
$2.83
$7.26
$10.09

($0.83 - $4.56)
($3.40 -$11.73)
($5.90 -$14.59)
2040
$4.09
$9.61
$13.69

($1.19 -$6.67)
($4.54 -$15.39)
($8.12 -$19.64)
Note:
*Top values in each cell are the midpoints, the values in parentheses are the 90 percent confidence intervals.
8.8.3.1 Effect of Oil Use on the Long-Run Oil Price
The first component of the full economic costs of importing petroleum into the U.S.
follows from the effect of U.S. import demand on the world oil price over the long-run. Because
the U.S. is a sufficiently large purchaser of global oil supplies, its purchases can affect the world
oil price. This monopsony power means that increases in U.S. petroleum demand can cause the
world price of crude oil to rise, and conversely, that reduced U.S. petroleum demand can reduce
the world price of crude oil. Thus, one benefit of decreasing U.S. oil purchases, due to
improvements in the fuel efficiency of medium- and heavy-duty vehicles is the potential
decrease in the crude oil price paid for all crude oil purchased.
The demand or monopsony effect can be readily illustrated with an example. If the U.S.
imports 10 million barrels per day at a world oil price of $50 per barrel, its total daily bill for oil
imports is 500 million dollars. If a 10 percent decrease in U.S. imports to 9 million barrels per
day causes the world oil price to drop to $49 per barrel, the daily U.S. oil import bill drops to
$441 million (9 million barrels times $49 per barrel). While the world oil price only declines $1,
the resulting decrease in oil purchase payments of $59 million per day (500 million dollars minus
$441 million) is equivalent to an incremental benefit of $59 per barrel of oil imports reduced
($59 million/1 million barrels per day reduced), or $10 more than the newly-decreased world
price of $49 per barrel. This additional $10 per barrel "import cost premium" represents the
incremental external benefits to the U.S. for avoided import costs beyond the price paid for oil
purchases. This additional benefit from import reduction arises only to the extent that a
reduction in U.S. oil imports affects the world oil price. ORNL estimates this component of the
energy security benefit in 2020 to be $2.21/barrel (2013$), with a range of $0.65/barrel to
$3.59/barrel of imported oil reduced.
There is disagreement in the literature about the magnitude of the monopsony
component, and its relevance for policy analysis. Brown and Huntington (2013)67, for example,
argue that the United States' refusal to exercise its market power to reduce the world oil price
does not represent a proper externality, and that the monopsony component should not be
considered in calculations of the energy security externality. However, they also note in their

-------
earlier discussion paper (Brown and Huntington 2010)68 that this is a departure from the
traditional energy security literature, which includes sustained wealth transfers associated with
stable but higher-price oil markets. On the other hand, Greene (2010)69 and others in prior
literature (e.g., Toman 1993)70 have emphasized that the monopsony cost component is policy-
relevant because the world oil market is non-competitive and strongly influenced by cartelized
and government-controlled supply decisions. Thus, while sometimes couched as an externality,
Greene notes that the monopsony component is best viewed as stemming from a completely
different market failure than an externality (Ledyard 2008)71, yet still implying marginal social
costs to importers.
Recently, the Council on Foreign Relations (i.e., "the Council") (2015)72 released a
discussion paper that assesses NHTSA's analysis of the benefits and costs of CAFE in a lower-
oil-price world. In this paper, the Council notes that while NHTSA cites the monopsony effect of
the CAFE standards for 2017-2025, NHTSA does not include it when calculating the cost-
benefit calculation for the rule. The Council argues that the monopsony benefit should be
included in the CAFE cost-benefit analysis and that including the monopsony benefit is more
consistent with the legislators' intent in mandating CAFE standards in the first place.
The recent National Academy of Science (NAS 2015) Report, "Cost, Effectiveness and
the Deployment of Fuel Economy Technologies for Light-Duty Vehicles,"73 suggests that the
agencies' logic about not accounting for monopsony benefits is inaccurate. According to the
NAS, the fallacy lies in treating the two problems, oil dependence and climate change, similarly.
According to the NAS, "Like national defense, it [oil dependence] is inherently adversarial (i.e.,
oil consumers against producers using monopoly power to raise prices). The problem of climate
change is inherently global and requires global action. If each nation considered only the
benefits to itself in determining what actions to take to mitigate climate change, an adequate
solution could not be achieved. Likewise, if the U.S. considers the economic harm its reduced
petroleum use will do to monopolistic oil producers it will not adequately address its oil
dependence problem. Thus, if the United States is to solve both of these problems it must take
full account of the costs and benefits of each, using the appropriate scope for each problem." At
this point in time, we are continuing to exclude monopsony premiums for the cost benefit
analysis of this final rule, but we will be taking comment on this issue in a near term future
rulemaking.
There is also a question about the ability of gradual, long-term reductions, such as those
resulting from these final rules, to reduce the world oil price in the presence of OPEC's
monopoly power. OPEC is currently the world's marginal petroleum supplier, and could
conceivably respond to gradual reductions in U.S. demand with gradual reductions in supply
over the course of several years as the fuel savings resulting from this rule grow. However, if
OPEC opts for a long-term strategy to preserve its market share, rather than maintain a particular
price level (as they have done recently in response to increasing U.S. petroleum production),
reduced demand would create downward pressure on the global price. The Oak Ridge analysis
assumes that OPEC does respond to demand reductions over the long run, but there is still a price
effect in the model. Under the mid-case behavioral assumption used in the premium
calculations, OPEC responds by gradually reducing supply to maintain market share (consistent
with the long-term self-interested strategy suggested by Gately (2004, 2007)).74

-------
One potential result of the potential decline in the world price of oil as a result of these
rules would be an increase in the consumption of petroleum products, particularly outside the
U.S. In addition, other fuels could be displaced from the increasing use of oil worldwide. For
example, if a decline in the world oil price causes an increase in oil use in China, India, or
another country's industrial sector, this increase in oil consumption may displace natural gas
usage. Alternatively, the increased oil use could result in a decrease in coal used to produce
electricity. An increase in the consumption of petroleum products particularly outside the U.S.,
could lead to a modest increase in emissions of GHGs, criteria air pollutants, and airborne toxics
from their refining and use. However, lower usage of, for example, displaced coal would result
in a decrease in GHG emissions. Therefore, any assessment of the impacts on GHG emissions
and other pollutants from a potential increase in world oil demand would need to take into
account the impacts on all portions of global energy sector.
8.8.3.2 Macroeconomic Disruption Adjustment Costs
The second component of the oil import premium, "avoided macroeconomic
disruption/adjustment costs," arises from the effect of oil imports on the expected cost of supply
disruptions and accompanying price increases. A sudden increase in oil prices triggered by a
disruption in world oil supplies has two main effects: (1) it increases the costs of oil imports in
the short-run, and (2) it can lead to macroeconomic contraction, dislocation and Gross Domestic
Product (GDP) losses. For example, ORNL estimates the combine value of these two factors in
to be $5.48/barrel (2013$) when U.S. oil imports are reduced in 2020, with a range from
$2.51/barrel to $8.92/barrel of imported oil reduced.
There are two main effects of macroeconomic disruption/adjustment costs. The first is
the aggregate effect of the short-run price increase from an oil shock. The oil price shock results
in a combination of real resource shortages, costly short-run shifts in energy supply, behavioral
and demand adjustments by energy users, and other response costs. Unlike pure transfers, the
root cause of the disruption price increase is a real resource supply reduction due, for example, to
disaster or war. Regions where supplies are disrupted, such as the U.S., suffer high costs.
Businesses' and households' emergency responses to supply disruptions and rapid price
increases consume real economic resources.
When households and businesses make decisions related to their oil consumption, such as
whether to invest in fuel-saving technologies or use futures markets, they are unlikely to account
for the effect of their petroleum consumption on the magnitude of costs that supply interruptions
and accompanying price shocks impose on others. As a consequence, the U.S. economy as a
whole will not make sufficient use of these mechanisms to insulate itself from the real costs of
rapid increases in energy prices and outlays that usually accompany oil supply interruptions.
Therefore, the ORNL estimate of avoided macroeconomic disruption/adjustment costs that the
agencies use to value energy security benefits includes the increased oil import costs stemming
from oil price shocks that are unanticipated and not internalized by advance actions of U.S.
consumers and businesses. This aggregate output effect will last as long as the oil price is
elevated. It depends on the extent and duration of any disruption in the world supply of oil, since
these factors determine the magnitude of the resulting increases in prices for petroleum products,
as well as how rapidly these prices return to their pre-disruption level.

-------
The second main effect of macroeconomic disruption/adjustment costs is the
macroeconomic losses due to "allocative" losses. These are the costs of temporary dislocation
and underutilization of available resources due to the oil shock, such as labor unemployment and
idle plant capacity. Because supply disruptions and resulting price increases occur suddenly,
empirical evidence shows they impose additional costs on businesses and households that must
adjust their use of petroleum and other productive factors more rapidly than if the same price
increase had occurred gradually. Dislocational effects include the unemployment of workers and
other resources during the time needed for their intersectoral or interregional reallocation, and
pauses in capital investment due to uncertainty. These adjustments temporarily reduce the level
of economic output that can be achieved even below the "potential" output level that would
ultimately be reached once the economy's adaptation to higher petroleum prices is complete.
The additional costs imposed on businesses and households for making these adjustments reflect
their limited ability to adjust prices, output levels, and their use of energy, labor and other inputs
quickly and smoothly in response to rapid changes in prices for petroleum products.
Since future disruptions in foreign oil supplies are an uncertain prospect, each of the
disruption cost components must be weighted by the probability that the supply of petroleum to
the U.S. will actually be disrupted. Thus, the "expected value" of these costs - the product of the
probability that a supply disruption will occur and the sum of costs from reduced economic
output and the economy's abrupt adjustment to sharply higher petroleum prices - is the relevant
measure of their magnitude. Further, when assessing the energy security value of a policy to
reduce oil use, it is only the change in the expected costs of disruption that results from the
policy that is relevant. The expected costs of disruption may change from lowering the normal
(i.e., pre-disruption) level of domestic petroleum use and imports, from any induced alteration in
the likelihood or size of disruption, or from altering the short-run flexibility (e.g., elasticity) of
petroleum use.
By late 2015/early 2016, world oil prices were sharply lower than in 2014. Future prices
remain uncertain, but sustained markedly lower oil prices can have mixed implications for U.S.
energy security. Under lower prices U.S. expenditures on oil consumption are lower, and they
are a less prominent component of the U.S. economy. This would lessen the issue of imported
oil as an energy security problem for the U.S. On the other hand, sustained lower oil prices
encourage greater oil consumption, and reduce the competitiveness of new U.S. oil supplies and
alternative fuels. The AEO 2015 low-oil price outlook, for example, projects that by 2030 total
U.S. petroleum supply would be 10 percent lower and imports would be 78 percent higher than
the AEO Reference Case. Under the low-price case, 2030 prices are 35 percent lower, so that
import expenditures are 16 percent higher.
A second potential proposed energy security effect of lower oil prices is increased
instability of supply, due to greater global reliance on fewer suppling nations,75 and because
lower prices may increase economic and geopolitical instability in some supplier nations.76'77'78
The International Monetary Fund reported that low oil prices are creating substantial economic
tension in the Middle East oil producers on top of the economic costs of ongoing conflicts, and
noted the risk that Middle East countries including Saudi Arabia could run out of financial assets
without substantial change in policy.79 The concern raised is that oil revenues are essential for
some exporting nations to fund domestic programs and avoid domestic unrest.

-------
The energy security costs estimated here follow the oil security premium framework,
which is well established in the energy economics literature. The oil import premium gained
attention as a guiding concept for energy policy around the time of the second and third major
post-war oil shocks (Bohi and Montgomery 1982, EMF 1982) ,80 Plummer (1982)81 provided
valuable discussion of many of the key issues related to the oil import premium as well as the
analogous oil stockpiling premium. Bohi and Montgomery (1982)82 detailed the theoretical
foundations of the oil import premium established many of the critical analytic relationships
through their thoughtful analysis. Hogan (1981)83 and Broadman and Hogan (1986, 1988)84
revised and extended the established analytical framework to estimate optimal oil import premia
with a more detailed accounting of macroeconomic effects.
Since the original work on energy security was undertaken in the 1980's, there have been
several reviews on this topic. For example, Leiby, Jones, Curlee and Lee (1997)85 provided an
extended review of the literature and issues regarding the estimation of the premium. Parry and
Darmstadter (2004)86 also provided an overview of extant oil security premium estimates and
they estimated some premium components.
The recent economics literature on whether oil shocks are a threat to economic stability
that they once were is mixed. Some of the current literature asserts that the macroeconomic
component of the energy security externality is small. For example, the National Research
Council (2009) argued that the non-environmental externalities associated with dependence on
foreign oil are small, and potentially trivial.87 Analyses by Nordhaus (2007) and Blanchard and
Gali (2010) question the impact of more recent oil price shocks on the economy.88 They were
motivated by attempts to explain why the economy actually expanded immediately after the last
shocks, and why there was no evidence of higher energy prices being passed on through higher
wage inflation. Using different methodologies, they conclude that the economy has largely
gotten over its concern with dramatic swings in oil prices.
One reason, according to Nordhaus, is that monetary policy has become more
accommodating to the price impacts of oil shocks. Another is that consumers have simply
decided that such movements are temporary, and have noted that price impacts are not passed on
as inflation in other parts of the economy. He also notes that real changes to productivity due to
oil price increases are incredibly modest,89 and that the general direction of the economy matters
a great deal regarding how the economy responds to a shock. Estimates of the impact of a price
shock on aggregate demand are insignificantly different from zero.
Blanchard and Gali (2010) contend that improvements in monetary policy (as noted
above), more flexible labor markets, and lessening of energy intensity in the economy, combined
with an absence of concurrent shocks, all contributed to lessen the impact of oil shocks after
1980. They find "... the effects of oil price shocks have changed over time, with steadily smaller
effects on prices and wages, as well as on output and employment."90 In a comment at the
chapter's end, this work is summarized as follows: "The message of this chapter is thus
optimistic in that it suggests a transformation in U.S. institutions has inoculated the economy
against the responses that we saw in the past."

-------
At the same time, the implications of the "Shale Oil Revolution" are now being felt in the
international markets, with current prices at four year lows. Analysts generally attribute this
result in part to the significant increase in supply resulting from U.S. production, which has put
liquid petroleum production on par with Saudi Arabia. The price decline is also attributed to the
sustained reductions in U.S. consumption and global demand growth from fuel efficiency
policies and high oil prices. The resulting decrease in foreign imports, down to about one-third
of domestic consumption (from 60 percent in 2005, for example91), effectively permits U.S.
supply to act as a buffer against artificial or other supply restrictions (the latter due to conflict or
natural disaster, for example).
However, other papers suggest that oil shocks, particularly sudden supply shocks, remain
a concern. Both Blanchard and Gali's and Nordhaus work were based on data and analysis
through 2006, ending with a period of strong global economic growth and growing global oil
demand. The Nordhaus work particularly stressed the effects of the price increase from 2002-
2006 that were comparatively gradual (about half the growth rate of the 1973 event and one-third
that of the 1990 event). The Nordhaus study emphasizes the robustness of the U.S. economy
during a time period through 2006. This time period was just before rapid further increases in
the price of oil and other commodities with oil prices more-than-doubling to over $130/barrel by
mid-2008, only to drop after the onset of the largest recession since the Great Depression.
Hamilton (2012)92 reviewed the empirical literature on oil shocks and suggested that the
results are mixed, noting that some work (e.g. Rasmussen and Roitman (2011) finds less
evidence for economic effects of oil shocks, or declining effects of shocks (Blanchard and Gali
2010), while other work continues to find evidence regarding the economic importance of oil
shocks. For example, Baumeister and Peersman (2011) found that an oil price increase of a
given size seems to have a decreasing effect over time, but noted that the declining price-
elasticity of demand meant that a given physical disruption had a bigger effect on price and
turned out to have a similar effect on output as in the earlier data." Hamilton observes that "a
negative effect of oil prices on real output has also been reported for a number of other countries,
particularly when nonlinear functional forms have been employed". Alternatively, rather than a
declining effect, Ramey and Vine (2010) 93 found "remarkable stability in the response of
aggregate real variables to oil shocks once we account for the extra costs imposed on the
economy in the 1970s by price controls and a complex system of entitlements that led to some
rationing and shortages."
Some of the recent literature on oil price shocks has emphasized that economic impacts
depend on the nature of the oil shock, with differences between price increases caused by sudden
supply loss and those caused by rapidly growing demand. Most recent analyses of oil price
shocks have confirmed that "demand-driven" oil price shocks have greater effects on oil prices
and tend to have positive effects on the economy while "supply-driven" oil shocks still have
negative economic impacts (Baumeister, Peersman and Van Robays, 2010).94 A recent paper by
Kilian and Vigfusson (2014),95 for example, assigned a more prominent role to the effects of
price increases that are unusual, in the sense of being beyond range of recent experience. Kilian
and Vigfussen also conclude that the difference in response to oil shocks may well stem from the

-------
different effects of demand- and supply-based price increases: "One explanation is that oil price
shocks are associated with a range of oil demand and oil supply shocks, some of which stimulate
the U.S. economy in the short run and some of which slow down U.S. growth (see Kilian 2009).
How recessionary the response to an oil price shock is thus depends on the average composition
of oil demand and oil supply shocks over the sample period."
The general conclusion that oil supply-driven shocks reduce economic output is also
reached in a recently published paper by Cashin et al. (2014)96 for 38 countries from 1979-2011.
"The results indicate that the economic consequences of a supply-driven oil-price shock are very
different from those of an oil-demand shock driven by global economic activity, and vary for oil-
importing countries compared to energy exporters," and "oil importers [including the U.S.]
typically face a long-lived fall in economic activity in response to a supply-driven surge in oil
prices" but almost all countries see an increase in real output for an oil-demand disturbance.
Note that the energy security premium calculation in this analysis is based on price shocks from
potential future supply events only.
Despite continuing uncertainty about oil market behavior and outcomes and the
sensitivity of the U.S. economy to oil shocks, it is generally agreed that it is beneficial to reduce
petroleum fuel consumption from an energy security standpoint. Reducing fuel consumption
reduces the amount of domestic economic activity associated with a commodity whose price
depends on volatile international markets. Also, reducing U.S. oil import levels reduces the
likelihood and significance of supply disruptions.
The Competitive Enterprise Institute (CEI) and others argue that there are little, if any,
energy security benefits associated with this rule. In large part CEI argues that oil supplies are
plentiful and that current oil prices are low so that reduced consumption of petroleum products
due to these rules would have no effect on energy security. However, the discussion of current
low oil prices ("lowest Labor Day gasoline prices in a decade") does not assure the absence of
future oil supply shocks or price shocks, or even speak to their reduced likelihood. CEI points
out that the current low oil prices have been observed before as recently as a decade ago, as they
have in more than one instance before that. For example, oil prices were even lower in 1999.
But in the intervening periods, oil supply and price shocks have continued to recur, and the
recent price record only amplifies oil's high historical price volatility.
Also, sharply lower world oil prices do not clearly imply greater energy security for the
U.S. Current low world oil prices may reduce the U.S.'s fracking industry's tight oil production
(as CEI points out), or other sources of oil supplies around the world. Some have hypothesized
that reduction in oil production outside of OPEC may be the objective of some OPEC producers.
With low oil prices, U.S.'s oil import share over time might be larger, increasing the U.S.'s
dependence on imported oil.
Securing America's Future Energy (SAFE), Operation Free and the Investor Network on
Climate Risk agree that this rule does improve America's energy security. SAFE goes on to
state that several policy options should be included in this rule to further enhance energy
security. The agencies agree that these rules enhances America's energy security, but does not
have information to evaluate the policy options that SAFE proposes.

-------
8.8.3.3 Cost of Existing U.S. Energy Security Policies
The last often-identified component of the full economic costs of U.S. oil imports are the
costs to the U.S. taxpayers of existing U.S. energy security policies. The two primary examples
are maintaining the Strategic Petroleum Reserve (SPR) and maintaining a military presence to
help secure a stable oil supply from potentially vulnerable regions of the world. The SPR is the
largest stockpile of government-owned emergency crude oil in the world. Established in the
aftermath of the 1973/74 oil embargo, the SPR provides the U.S. with a response option should a
disruption in commercial oil supplies threaten the U.S. economy. It also allows the U.S. to meet
part of its International Energy Agency obligation to maintain emergency oil stocks, and it
provides a national defense fuel reserve. While the costs for building and maintaining the SPR
are more clearly related to U.S. oil use and imports, historically these costs have not varied in
response to changes in U.S. oil import levels. Thus, while the effect of the SPR in moderating
price shocks is factored into the ORNL analysis, the cost of maintaining the SPR is excluded.
U.S. military costs are excluded from the analysis performed by ORNL because their
attribution to particular missions or activities is difficult, and because it is not clear that these
outlays would decline in response to incremental reductions in U.S. oil imports. Most military
forces serve a broad range of security and foreign policy objectives. Attempts to attribute some
share of U.S. military costs to oil imports are further challenged by the need to estimate how
those costs might vary with incremental variations in U.S. oil imports.
In the proposal to this rule, the agencies solicited comments on quantifying the military
benefits from reduced U.S. imports of oil. The California Air Resources Board (CARB) notes
that the National Research Council (NRC)97 attempted to estimate the military costs associated
with U.S. imports and consumption of petroleum. The NRC cited estimates of the national
defense costs of oil dependence from the literature that range from less than $5 to $50 billion per
year or more. Assuming a range of approximate range of $10 to $50 billion per year, the NRC
divided national defense costs by a projected U.S. consumption rate of approximately 6.4 billion
barrels per year (EIA, 2012). This procedure yielded a range of average national defense cost of
$1.50 - $8.00 per barrel (rounded to the nearest $0.50), with a mid-point of $5/barrel (in 2009$).
The agencies acknowledge this NRC study, but have not included the estimates as part of the
cost-benefit analysis for this rule.
8.8.4 Energy Security Benefits of this Program
Using the ORNL "oil premium" methodology, updating world oil price values and
energy trends using AEO 2015 and using the estimated fuel savings from the final rule estimated
from the MOVES/CAFE models, the agencies have calculated the energy security benefits of
these final rules for different classes for medium- and heavy-duty vehicles for the various years
up to 2050,YY Since the agencies are taking a global perspective with respect to valuing
greenhouse gas benefits from the rule, only the avoided macroeconomic adjustment/disruption
YY In order to determine the energy security benefits beyond 2040, we use the 2040 energy security premium
multiplied by the estimate fuel savings from the final program. Since the AEO 2015 only goes to 2040, we only
calculate energy security premiums to 2040.

-------
portion of the energy security premium is used in the energy security benefits estimates present
below. These results are shown below in Table 8-27, Table 8-28 and Table 8-29 show
discounted model year lifetime energy security benefits for different classes of heavy-duty
vehicles using a three and seven percent discount rate.
Table 8-27 Annual U.S. Energy Security Benefits and Net Present Values at 3% and 7% Discount Rates for
the Final Program (Millions of 2013$)a
CALENDAR
HD
VOCATIONAL
TRACTOR/
SUM
YEAR
PICKUP
& VANS

TRAILER

2018
$0
$0
$4
$4
2019
$0
$0
$9
$9
2020
$0
$0
$14
$14
2021
$2
$9
$44
$55
2022
$8
$18
$83
$109
2023
$18
$27
$125
$171
2024
$32
$43
$193
$268
2025
$51
$58
$263
$372
2026
$74
$74
$335
$482
2027
$101
$95
$431
$627
2028
$129
$117
$528
$775
2029
$157
$140
$626
$923
2030
$186
$162
$726
$1,074
2035
$327
$274
$1,246
$1,847
2040
$438
$370
$1,725
$2,533
2050
$489
$435
$2,101
$3,025
NPV, 3%
$4,166
$3,633
$16,916
$24,716
NPV, 7%
$1,684
$1,485
$6,881
$10,050
Table 8-28 Discounted Model Year Lifetime Energy Security Benefits at a 3% Discount Rate for the Final
Program (Millions of 2013$)a
MODEL
HD PICKUP
VOCATIONAL
TRACTOR/
SUM
YEAR
AND VANS

TRAILER

2018
$0
$0
$30
$30
2019
$0
$0
$29
$29
2020
$0
$0
$28
$28
2021
$21
$85
$379
$485
2022
$56
$85
$380
$520
2023
$90
$84
$378
$552
2024
$124
$130
$595
$849
2025
$157
$131
$598
$886
2026
$190
$131
$596
$917
2027
$221
$174
$788
$1,183
2028
$220
$175
$787
$1,182
2029
$219
$175
$790
$1,184
Sum
$1,296
$1,169
$5,379
$7,844

-------
Table 8-29 Discounted Model Year Lifetime Energy Security Benefits at 7% Discount Rate due to the Final
Program (Millions of 2013$)a
MODEL
HD PICKUP
VOCATIONAL
TRACTOR/
SUM
YEAR
AND VANS

TRAILER

2018
$0
$0
$21
$21
2019
$0
$0
$20
$20
2020
$0
$0
$18
$18
2021
$13
$52
$230
$294
2022
$33
$50
$222
$304
2023
$51
$47
$213
$311
2024
$67
$71
$323
$461
2025
$83
$69
$313
$464
2026
$96
$66
$301
$463
2027
$108
$85
$384
$577
2028
$104
$82
$369
$555
2029
$99
$80
$358
$536
Sum
$653
$602
$2,771
$4,026
8.9 Summary of Benefits and Costs
This section presents the costs, benefits, and other economic impacts of the Phase 2
standards. It is important to note that NHTSA's fuel consumption standards and EPA's GHG
standards will both be in effect, and will jointly lead to increased fuel efficiency and reductions
in GHG and non-GHG emissions. The individual categories of benefits and costs presented in
the tables below include:
•	the vehicle program costs (costs of complying with the vehicle CO2 and fuel
consumption standards),
•	changes in fuel expenditures associated with reduced fuel use by more efficient vehicles
and increased fuel use associated with the "rebound" effect, both of which result from the
program,
•	the global economic value of reductions in GHGs,
•	the economic value of reductions in non-GHG pollutants,
•	costs associated with increases in noise, congestion, and crashes resulting from increased
vehicle use,
•	savings in drivers' time from less frequent refueling,
•	benefits of increased vehicle use associated with the "rebound" effect, and
•	the economic value of improvements in U.S. energy security impacts.
For a discussion of the cost of ownership and the agencies' payback analysis of vehicles
covered by these final rules, please see Chapter 7 of this RIA.
The agencies separate analyses using two analytical methods referred to as Method A and
Method B. For an explanation of these methods, please see Section I.D for the Preamble. And
as discussed in Preamble Section X.A.I, the agencies present estimates of benefits and costs that
are measured against two different assumptions about improvements in fuel efficiency that might
occur in the absence of the Phase 2 standards. The first case (Alternative la) uses a baseline that

-------
projects very little improvement in new vehicles in the absence of new Phase 2 standards, and
the second (Alternative lb) uses a more dynamic baseline that projects more significant
improvements in vehicle fuel efficiency.
Table 8-30 shows benefits and costs for the standards from the perspective of a program
designed to improve the nation's energy security and conserve energy by improving fuel
efficiency. From this viewpoint, technology costs occur when the vehicle is purchased. Fuel
savings are counted as benefits that occur over the lifetimes of the vehicles produced during the
model years subject to the Phase 2 standards as they consume less fuel. The table shows that
benefits far outweigh the costs, and the final program is anticipated to result in large net benefits
to the U.S economy.

-------
Table 8-30 Lifetime Benefits & Costs of the Final Program for Model Years 2018 - 2029 Vehicles Using
Analysis Method A (Billions of 2013$ discounted at 3% and 7%)
CATEGORY
BASELINE 1A
BASELINE IB
3%
7%
3%
7%
Vehicle Program: Technology and
Indirect Costs, Normal Profit on
Additional Investments
24.4
16.6
23.7
16.1
Additional Routine Maintenance
1.7
0.9
1.7
0.9
Congestion, Crashes, Fatalities and Noise
from Increased Vehicle Usea
3.2
1.9
3.1
1.8
Total Costs
29.3
19.4
28.5
18.8
Fuel Savings (valued at pre-tax prices)
163.0
87.0
149.1
79.7
Savings from Less Frequent Refueling
3.2
1.7
3.0
1.6
Economic Benefits from Additional
Vehicle Use
5.5
3.5
5.4
3.4
Reduced Climate Damages from GHG
Emissions b
36.0
33.0
Reduced Health Damages from Non-
GHG Emissions
30.0
16.1
27.2
14.5
Increased U.S. Energy Security
7.9
4.2
7.3
3.9
Total Benefits
246
149
225
136
Net Benefits
216
129
197
117
Note:
Benefits and net benefits use the 3 percent average global SC-CO2, SC-CH4, and SC-N20 value
applied to CO2, CH4, and N20 emissions, respectively; GHG reductions also include HFC
reductions, and include benefits to other nations as well as the U.S. See RIA Chapter 8.5 and
Preamble Section IX.G for further discussion.
b "Congestion, Crashes, Fatalities and Noise from Increased Vehicle Use" includes NHTSA's
monetized value of estimated reductions in the incidence of highway fatalities associated with mass
reduction in HD pickup and vans, but this does not include these reductions from tractor-trailers or
vocational vehicles. This likely results in a conservative overestimate of these costs.
Table 8-30, Table 8-31 and Table 8-33 report benefits and cost from the perspective of
reducing GHG. Table 8-31 shows the annual impacts and net benefits of the final program for
selected future years, together with the net present values of cumulative annual impacts from
2018 through 2050, discounted at 3 percent and 7 percent rates. Table 8-31 and Table 8-33 show
the discounted lifetime costs and benefits for each model year affected by the Phase 2 standards
at 3 percent and 7 percent discount rates, respectively.

-------
Table 8-31 Annual Benefits & Costs and Net Present Values for the Final Program Relative to the Flat
Baseline and using Method B
(Billions of 2013$)abc

2018
2021
2024
2030
2035
2040
2050
NPV,
3%
NPV,
7%
Vehicle
program
-$0.2
-$2.5
-$4.2
-$5.2
-$5.7
-$6.3
-$7.3
-$87.8
-$41.9
Maintenance
$0.0
$0.0
-$0.1
-$0.2
-$0.2
-$0.2
-$0.2
-$3.2
-$1.5
Pre-tax Fuel
$0.1
$1.3
$6.1
$23.4
$38.9
$53.1
$63.4
$523.3
$213.8
Energy
security
$0.0
$0.1
$0.3
$1.1
$1.8
$2.5
$3.0
$24.7
$10.1
Crashes/
Congestion/
Noise
$0.0
-$0.1
-$0.2
-$0.4
-$0.5
-$0.5
-$0.6
-$6.8
-$3.1
Refueling
$0.0
$0.0
$0.1
$0.6
$0.9
$1.1
$1.5
$12.0
$4.9
Travel value
$0.0
$0.3
$0.6
$1.2
$1.7
$2.0
$2.3
$23.4
$10.3
Non-GHG
$0.0 to
$0.0
$0.2 to
$0.5
$0.7 to
$1.8
$2.7 to
$6.8
$4.1 to
$10.1
$5.0 to
$12.5
$6.0 to
$15.0
$58.8 to
$132.0
$22.1 to
$49.7
GHG









SC-GHG;
5% avg
$0.0
$0.1
$0.4
$1.7
$2.8
$3.9
$5.8
$25.1
$25.1
SC-GHG;
3% avg
$0.0
$0.3
$1.4
$5.2
$8.4
$11.1
$15.2
$115.4
$115.4
SC-GHG;
2.5% avg
$0.0
$0.4
$2.0
$7.5
$11.9
$15.5
$20.9
$183.1
$183.1
SC-GHG;
3% 95th
$0.1
$0.9
$4.1
$15.6
$25.5
$33.6
$46.6
$351.0
$351.0
Net benefits









SC-GHG;
5% avg
-$0.1
-$0.6
$4.3
$26.7
$46.6
$64.3
$78.2
$606.2
$253.8
SC-GHG;
3% avg
-$0.1
-$0.4
$5.2
$30.2
$52.2
$71.4
$87.6
$696.4
$344.0
SC-GHG;
2.5% avg
-$0.1
-$0.3
$5.9
$32.6
$55.7
$75.8
$93.3
$764.2
$411.8
SC-GHG;
3% 95th
$0.0
$0.2
$8.0
$40.7
$69.4
$94.0
$119.0
$932.1
$579.7
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Preamble Section X.A.I
h GHG benefit estimates include reductions in CO2, CH4, and N20 but do not include the HFC reductions. 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 (SC-CO2, SC-CH4, and SC-N20, each discounted at
rates of at 5, 3, 2.5 percent) is used to calculate net present value of SC-C02, SC-CH4, and SC-N20, respectively,
for internal consistency. Refer to the SC-C02 TSD for more detail.
Chapter 8.5 of the RIA notes that SC-GHGs increases over time. Corresponding to the years in this table (2020-
2050), the SC-CO2 estimates range as follows: for Average SC-CO2 at 5%: $7-$16; for Average SC-CO2 at 3%:
$27-$46; for Average SC-CO2 at 2.5%: $43-$67; and for 95th percentile SC-CO2 at 3%: $83-$140. For the years
2012-2050, the SC-CH4 estimates range as follows: for Average SC-CH4 at 5%: $440-$l,400; for Average SC-
OW at 3%: $l,000-$2,700; for Average SC-CH4 at 2.5%: $l,400-$3,400; and for 95th percentile SC-CH4 at 3%:
$2,800-$7,400. For the years 2012-2050, the SC-N20 estimates range as follows: for Average SC-N20 at 5%:

-------
$4,000-$ 12,000; for Average SC-N20 at 3%: $14,000-$30,000; for Average SC-N20 at 2.5%: $21,000-$41,000;
and for 95th percentile SC-N20 at 3%: $36,000-$79,000. Chapter 8.5 also presents these SC-GHG estimates.
The table shows the benefits of reduced CO2, CH4, and N2O emissions—and
consequently the annual quantified benefits {i.e., total benefits)—for each of four SC-CO2, SC-
CH4, and SC-N2O values, respectively. As discussed in Chapter 8.5, there are some limitations
to the SC-CO2, SC-CH4, and SC-N2O analysis, including the incomplete way in which the
integrated assessment models capture catastrophic and non-catastrophic impacts, their
incomplete treatment of adaptation and technological change, uncertainty in the extrapolation of
damages to high temperatures, and assumptions regarding risk aversion.
In addition, these monetized GHG benefits exclude the value of reductions in HFC
emissions expected under this program. Although EPA has not included monetized estimates of
benefits of reductions in HFC emissions in this Chapter 8.9, the value of these reductions should
not be interpreted as zero. The reader is referred to Chapter 8.5.2.2 of this RIA to see the
sensitivity analysis that approximates the value of HFC benefits.
The agencies have also conducted a separate analysis of the total benefits over the model
year lifetimes of 2018 through 2029 model year vehicles. In contrast to the calendar year
analysis presented in Table 8-31, the model year lifetime analysis shows the impacts of the
program on vehicles produced during each of the affected model years over the course of their
expected lifetimes. The net societal benefits over the full lifetimes of vehicles produced during
each of the model years are shown in Table 8-31 and Table 8-32 at both 3 percent and 7 percent
discount rates, respectively.

-------
Table 8-32 Discounted Model Year Lifetime Impacts for the Final Program Relative to the Flat Baseline and using Method B
(Billions of 2013$; 3% Discount Rate) a'b'c

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
SUM
Vehicle Program
-$0.2
-$0.2
-$0.2
-$2.1
-$2.0
-$2.1
-$3.1
-$3.0
-$3.0
-$3.6
-$3.5
-$3.4
-$26.5
Maintenance
-
-
-











$0.01
$0.01
$0.01
-$0.15
-$0.16
-$0.16
-$0.18
-$0.18
-$0.17
-$0.30
-$0.29
-$0.29
-$1.9
Pre-tax Fuel
$0.7
$0.7
$0.6
$10.7
$11.4
$12.0
$18.5
$19.1
$19.7
$25.3
$25.2
$25.1
$169.1
Energy Security
$0.0
$0.0
$0.0
$0.5
$0.5
$0.6
$0.8
$0.9
$0.9
$1.2
$1.2
$1.2
$7.8
Crashes, Noise, Congestion
-$0.1
-$0.1
-$0.2
-$0.3
-$0.3
-$0.3
-$0.3
-$0.3
-$0.3
-$0.3
-$0.3
-$0.3
-$3.2
Refueling
$0.0
$0.0
$0.0
$0.2
$0.3
$0.3
$0.4
$0.5
$0.5
$0.6
$0.6
$0.6
$4.0
Travel value
$0.5
$0.5
$0.6
$1.1
$1.0
$1.0
$1.0
$1.0
$1.0
$1.0
$0.9
$0.9
$10.5
Non-GHG
$0.1
$0.1
$0.1
$1.4
$1.4
$1.5
$2.3
$2.3
$2.2

$2.7
$2.7
$19.6

to
to
to
to
to
to
to
to
to
$2.8 to
to
to
to

$0.3
$0.2
$0.2
$3.2
$3.2
$3.3
$5.2
$5.3
$4.8
$6.2
$6.1
$6.0
$44.1
GHG













SC-GHG; 5% avg
$0.0
$0.0
$0.0
$0.6
$0.6
$0.6
$1.0
$1.0
$1.0
$1.3
$1.2
$1.2
$8.6
SC-GHG; 3% avg
$0.2
$0.1
$0.1
$2.4
$2.6
$2.7
$4.1
$4.2
$4.3
$5.5
$5.5
$5.5
$37.2
SC-GHG; 2.5% avg
$0.2
$0.2
$0.2
$3.7
$4.0
$4.2
$6.4
$6.6
$6.8
$8.7
$8.6
$8.6
$58.3
SC-GHG; 3% 95th
$0.5
$0.4
$0.4
$7.2
$7.7
$8.0
$12.3
$12.7
$13.1
$16.8
$16.7
$16.6
$112.5
Net benefits













SC-GHG; 5% avg
$1.1
$1.1
$1.1
$12.8
$13.7
$14.3
$21.8
$22.7
$23.1
$29.6
$29.5
$29.5
$200.2
SC-GHG; 3% avg
$1.2
$1.2
$1.2
$14.6
$15.6
$16.3
$24.9
$26.0
$26.4
$33.9
$33.8
$33.7
$228.8
SC-GHG; 2.5% avg
$1.3
$1.3
$1.3
$16.0
$17.1
$17.8
$27.2
$28.4
$28.9
$37.0
$36.9
$36.9
$249.9
SC-GHG; 3% 95th
$1.5
$1.5
$1.5
$19.5
$20.8
$21.7
$33.2
$34.5
$35.2
$45.1
$44.9
$44.9
$304.1
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the flat baseline, la, and dynamic baseline, lb, please
see Preamble Section X.A.I
b The monetized GHG benefits presented in this analysis exclude the value of changes in HFC emissions expected under this program (see RIA Chapter 8.5).
Although EPA has not monetized changes in HFCs in the main benefits analysis, the value of any increases or reductions should not be interpreted as zero.
0 GHG benefit estimates include reductions in CO2, CH4, and N20 but do not include the HFC reductions. Note that net present value of reduced CO2 GHG
emissions is calculated differently than other benefits. The same discount rate used to discount the value of damages from future emissions (SC-CO2, SC-CH4,
and SC-N2O, each discounted at rates of at 5, 3, 2.5 percent) is used to calculate net present value of SC-CO2, SC-CH4, and SC-N20, respectively, SC-CO2 for
internal consistency. Refer to the SC-CO2 TSD for more detail.

-------
Table 8-33 Discounted Model Year Lifetime Impacts for the Final Program Relative to the Flat Baseline and using Method B (Billions of 2013$; 7%
Discount Rate) a b'c

2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
SUM
Vehicle Program
-$0.2
-$0.2
-$0.2
-$1.6
-$1.5
-$1.5
-$2.2
-$2.0
-$1.9
-$2.2
-$2.1
-$2.0
-$17.6
Maintenance
$0.00
$0.00
$0.00
-$0.10
-$0.09
-$0.09
-$0.10
-$0.10
-$0.09
-$0.15
-$0.14
-$0.13
-$1.0
Pre-tax Fuel
$0.5
$0.4
$0.4
$6.6
$6.7
$6.8
$10.1
$10.1
$10.0
$12.4
$11.9
$11.4
$87.2
Energy Security
$0.0
$0.0
$0.0
$0.3
$0.3
$0.3
$0.5
$0.5
$0.5
$0.6
$0.6
$0.5
$4.0
Crashes, Noise,
Congestion
-$0.1
-$0.1
-$0.1
-$0.2
-$0.2
-$0.2
-$0.2
-$0.2
-$0.2
-$0.1
-$0.1
-$0.1
-$1.8
Refueling
$0.0
$0.0
$0.0
$0.1
$0.2
$0.2
$0.2
$0.2
$0.3
$0.3
$0.3
$0.3
$2.1
Travel value
$0.3
$0.3
$0.4
$0.6
$0.6
$0.6
$0.5
$0.5
$0.5
$0.5
$0.4
$0.4
$5.7
Non-GHG
$0.1 to
$0.2
$0.1 to
$0.1
$0.1 to
$0.1
$0.8
to
$1.8
$0.8
to
$1.7
$0.8
to
$1.7
$1.1
to
$2.6
$1.1
to
$2.5
$1.0
to
$2.2
$1.2
to
$2.7
$1.2
to
$2.6
$1.1
to
$2.5
$9.2
to
$20.8
GHG













SC-GHG; 5% avg
$0.0
$0.0
$0.0
$0.6
$0.6
$0.6
$1.0
$1.0
$1.0
$1.3
$1.2
$1.2
$8.6
SC-GHG; 3% avg
$0.2
$0.1
$0.1
$2.4
$2.6
$2.7
$4.1
$4.2
$4.3
$5.5
$5.5
$5.5
$37.2
SC-GHG; 2.5%
avg
$0.2
$0.2
$0.2
$3.7
$4.0
$4.2
$6.4
$6.6
$6.8
$8.7
$8.6
$8.6
$58.3
SC-GHG; 3% 95th
$0.5
$0.4
$0.4
$7.2
$7.7
$8.0
$12.3
$12.7
$13.1
$16.8
$16.7
$16.6
$112.5
Net benefits













SC-GHG; 5% avg
$0.7
$0.7
$0.6
$7.6
$7.9
$7.9
$11.7
$11.8
$11.6
$14.4
$13.9
$13.5
$102.3
SC-GHG; 3% avg
$0.8
$0.8
$0.8
$9.4
$9.8
$10.0
$14.8
$15.1
$15.0
$18.7
$18.2
$17.7
$130.9
SC-GHG; 2.5%
avg
$0.9
$0.9
$0.8
$10.7
$11.2
$11.4
$17.1
$17.4
$17.4
$21.9
$21.3
$20.9
$151.9
SC-GHG; 3% 95th
$1.1
$1.1
$1.0
$14.2
$14.9
$15.3
$23.0
$23.6
$23.7
$29.9
$29.3
$28.9
$206.1
Notes:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of the less dynamic baseline, la, and more dynamic
baseline, lb, please see Preamble Section X.A.I
b The monetized GHG benefits presented in this analysis exclude the value of changes in HFC emissions expected under this program (see RIA Chapter 8.5).
Although EPA has not monetized changes in HFC s in the main benefits analysis, the value of any increases or reductions should not be interpreted as zero.
0 GHG benefit estimates include reductions in CO2, CH4, and N20 but do not include the HFC reductions. Note that net present value of reduced CO2 GHG
emissions is calculated differently than other benefits. The same discount rate used to discount the value of damages from future emissions (SC-CO2, SC-CH4,
and SC-N2O, each discounted at rates of at 5, 3, 2.5 percent) is used to calculate net present value of SC-CO2, SC-CH4, and SC-N20, respectively, SC-CO2 for
internal consistency. Refer to the SC-CO2 TSD for more detail.

-------
8.10 Employment Impacts
8.10.1 Introduction
Executive Order 13563 (January 18, 2011) directs federal agencies to consider regulatory
impacts on, among other criteria, job creation.98 According to the Executive Order "Our
regulatory system must protect public health, welfare, safety, and our environment while
promoting economic growth, innovation, competitiveness, and job creation. It must be based on
the best available science" (emphasis added). Analysis of employment impacts of a regulation is
not part of a standard benefit-cost analysis (except to the extent that labor costs contribute to
costs). Employment impacts of federal rules are of general interest, however, and have been
particularly so, historically, in the auto sector during periods of challenging labor market
conditions. For this reason, we are describing the connections of these standards to employment
in the regulated sector, the motor vehicle manufacturing sector, as well as the motor vehicle body
and trailer and motor vehicle parts manufacturing sectors.zz
The overall effect of the final rules on motor vehicle sector employment depends on the
relative magnitude of output and substitution effects, described below. Because we do not have
quantitative estimates of the output effect, and only a partial estimate of the substitution effect,
we cannot reach a quantitative estimate of the overall employment effects of the final rules on
motor vehicle sector employment or even whether the total effect will be positive or negative.
According to the U.S. Bureau of Labor Statistics, in 2014, about 850,000 people in the
U.S. were employed in the Motor Vehicle and Parts Manufacturing Sector (NAICS 3361, 3362,
and 3363)," the directly regulated sector. The employment effects of these final rules are
expected to expand beyond the regulated sector. Though some of the parts used to achieve the
standards are likely to be built by motor vehicle manufacturers (including trailer manufacturers)
themselves, the motor vehicle parts manufacturing sector also plays a significant role in
providing those parts, and will also be affected by changes in vehicle sales. Changes in truck
sales, discussed in Chapter 8.4.2, could also affect employment for truck and trailer vendors. As
discussed in Chapter 7, the final rules are expected to reduce the amount of fuel these vehicles
use, and thus affect the petroleum refinery and supply industries as well. Finally, since the net
reduction in cost associated with these final rules is expected to lead to lower transportation and
shipping costs, in a competitive market a substantial portion of those cost savings will be passed
along to consumers, who then will have additional discretionary income (how much of the cost is
passed along to consumers depends on market structure and the relative price elasticities). The
final rules are not expected to have any notable inflationary or recessionary effect.
The employment effects of environmental regulation are difficult to disentangle from
other economic changes and business decisions that affect employment, over time and across
zz The employment analysis in this RIA is part of EPA's ongoing effort to "conduct continuing evaluations of
potential loss or shifts of employment which may result from the administration or enforcement of [the Act]"
pursuant to CAA section 321(a).

-------
regions and industries. In light of these difficulties, we lean on economic theory to provide a
constructive framework for approaching these assessments and for better understanding the
inherent complexities in such assessments. Neoclassical microeconomic theory describes how
profit-maximizing firms adjust their use of productive inputs in response to changes in their
economic conditions.100 Berman and Bui (2001, pp. 274-75) model two components that drive
changes in firm-level labor demand: output effects and substitution effects. 101>AAA Regulation
can affect the profit-maximizing quantity of output by changing the marginal cost of production.
If regulation causes marginal cost to increase, it will place upward pressure on output prices,
leading to a decrease in the quantity demanded, and resulting in a decrease in production. The
output effect describes how, holding labor intensity constant, a decrease in production causes a
decrease in labor demand. As noted by Berman and Bui, although many assume that regulation
increases marginal cost, it need not be the case. A regulation could induce a firm to upgrade to
less polluting and more efficient equipment that lowers marginal production costs, or it may
induce use of technologies that may prove popular with buyers or provide positive network
externalities (see Chapter 8.2 for discussion of this effect). In such a case, output could increase.
The substitution effect describes how, holding output constant, regulation affects labor
intensity of production. Although increased environmental regulation may increase use of
pollution control equipment and energy to operate that equipment, the impact on labor demand is
ambiguous. For example, equipment inspection requirements, specialized waste handling, or
pollution technologies that alter the production process may affect the number of workers
necessary to produce a unit of output. Berman and Bui (2001) model the substitution effect as
the effect of regulation on pollution control equipment and expenditures required by the
regulation and the corresponding change in labor intensity of production.
In summary, as output and substitution effects may be positive or negative, theory alone
cannot predict the direction of the net effect of regulation on labor demand at the level of the
regulated firm. Operating within the bounds of standard economic theory, empirical estimation
of net employment effects on regulated firms is possible when data and methods of sufficient
detail and quality are available. The literature, however, illustrates difficulties with empirical
estimation. For example, studies sometimes rely on confidential plant-level employment data
from the U.S. Census Bureau, possibly combined with pollution abatement expenditure data that
are too dated to be reliably informative. In addition, the most commonly used empirical methods
do not permit estimation of net effects.
The conceptual framework described thus far focused on regulatory effects on plant-level
decisions within a regulated industry. Employment impacts at an individual plant do not
necessarily represent impacts for the sector as a whole. The approach must be modified when
applied at the industry level.
AAA Berman and Bui also discuss a third component, the impact of regulation on factor prices, but conclude that this
effect is unlikely to be important for large competitive factor markets, such as labor and capital. Morgenstern, Pizer
and Shih (2002) use a very similar model, but they break the employment effect into three parts: 1) a demand effect;
2) a cost effect; and 3) a factor-shift effect.

-------
At the industry level, labor demand is more responsive if: (1) the price elasticity of
demand for the product is high, (2) other factors of production can be easily substituted for labor,
(3) the supply of other factors is highly elastic, or (4) labor costs are a large share of total
production costs.102 For example, if all firms in an industry are faced with the same regulatory
compliance costs and product demand is inelastic, then industry output may not change much,
and output of individual firms may change slightly.103 In this case, the output effect may be
small, while the substitution effect depends on input substitutability. Suppose, for example, that
new equipment for fuel efficiency improvements requires labor to install and operate. In this
case, the substitution effect may be positive, and with a small output effect, the total effect may
be positive. As with potential effects for an individual firm, theory cannot determine the sign or
magnitude of industry-level regulatory effects on labor demand. Determining these signs and
magnitudes requires additional sector-specific empirical study. For environmental rules, much
of the data needed for these empirical studies is not publicly available, would require significant
time and resources in order to access confidential U.S. Census data for research, and also would
not be necessary for other components of a typical RIA.
In addition to changes to labor demand in the regulated industry, net employment impacts
encompass changes in other related sectors. For example, the standards are expected to increase
demand for fuel-saving technologies. This increased demand may increase revenue and
employment in the firms providing these technologies. At the same time, the regulated industry
is purchasing the equipment, and these costs may impact labor demand at regulated firms.
Therefore, it is important to consider the net effect of compliance actions on employment across
multiple sectors or industries.
If the U.S. economy is at full employment, even a large-scale environmental regulation is
unlikely to have a noticeable impact on aggregate net employment.BBB Instead, labor would
primarily be reallocated from one productive use to another, and net national employment effects
from environmental regulation will be small and transitory (e.g., as workers move from one job
to another).104
Affected sectors may experience transitory effects as workers change jobs. Some
workers may retrain or relocate in anticipation of new requirements or require time to search for
new jobs, while shortages in some sectors or regions could bid up wages to attract workers.
These adjustment costs can lead to local labor disruptions. Although the net change in the
national workforce is expected to be small, localized reductions in employment may adversely
impact individuals and communities just as localized increases may have positive impacts.
If the economy is operating at less than full employment, economic theory does not
clearly indicate the direction or magnitude of the net impact of environmental regulation on
employment; it could cause either a short-run net increase or short-run net decrease.105 An
important research question is how to accommodate unemployment as a structural feature in
BBB Full employment is a conceptual target for the economy where everyone who wants to work and is available to
do so at prevailing wages is actively employed. The unemployment rate at Ml employment is not zero.

-------
economic models. This feature may be important in assessing large-scale regulatory impacts on
employment.106
Environmental regulation may also affect labor supply. In particular, pollution and other
environmental risks may impact labor productivity or employees' ability to work.107 While the
theoretical framework for analyzing labor supply effects is analogous to that for labor demand, it
is more difficult to study empirically. There is a small emerging literature described in the next
section that uses detailed labor and environmental data to assess these impacts.
To summarize, economic theory provides a framework for analyzing the impacts of
environmental regulation on employment. The net employment effect incorporates expected
employment changes (both positive and negative) in the regulated sector and elsewhere. Labor
demand impacts for regulated firms, and also for the regulated industry, can be decomposed into
output and substitution effects which may be either negative or positive. Estimation of net
employment effects for regulated sectors is possible when data of sufficient detail and quality are
available. Finally, economic theory suggests that labor supply effects are also possible. In the
next section, we discuss the empirical literature.
8.10.1.1 Current State of Knowledge Based on the Peer-Reviewed Literature
In the labor economics literature there is an extensive body of peer-reviewed empirical
work analyzing various aspects of labor demand, relying on the above theoretical framework.108
This work focuses primarily on the effects of employment policies, e.g. labor taxes, minimum
wage, etc.109 In contrast, the peer-reviewed empirical literature specifically estimating
employment effects of environmental regulations is very limited. Several empirical studies,
including Berman and Bui (2001),110 Morgenstern, Pizer and Shih (2002),111 Gray et al
(2014),112 and Ferris, Shadbegian and Wolverton (2014)113 suggest that net employment impacts
may be zero or slightly positive but small even in the regulated sector. Other research suggests
that more highly regulated counties may generate fewer jobs than less regulated ones.114
However, since these latter studies compare more regulated to less regulated counties, they
overstate the net national impact of regulation to the extent that regulation causes plants to locate
in one area of the country rather than another. List et al. (2003)115 find some evidence that this
type of geographic relocation may be occurring. Overall, the peer-reviewed literature does not
contain evidence that environmental regulation has a large impact on net employment (either
negative or positive) in the long run across the whole economy.
Analytic challenges make it very difficult to accurately produce net employment
estimates for the whole economy that would appropriately capture the way in which costs,
compliance spending, and environmental benefits propagate through the macro-economy.
Quantitative estimates are further complicated by the fact that macroeconomic models often have
very little sectoral detail and usually assume that the economy is at full employment. EPA is
currently in the process of seeking input from an independent expert panel on modeling
economy-wide impacts, including employment effects. For more information,
see: https://federalregister.gOv/a/2014-02471.

-------
8.10,2 Employment Impacts in the Motor Vehicle and Parts Manufacturing Sector
This chapter describes changes in employment in the motor vehicle, trailer, and parts
(hence, motor vehicle) manufacturing sectors due to these final rules. We focus on the motor
vehicle manufacturing sector because it is directly regulated, and because it is likely to bear a
substantial share of changes in employment due to these final rules. We include discussion of
effects on the parts manufacturing sector, because the motor vehicle manufacturing sector can
either produce parts internally or buy them from an external supplier, and we do not have
estimates of the likely breakdown of effort between the two sectors.
We follow the theoretical structure of Berman and Bui 116 of the impacts of regulation in
employment in the regulated sectors. In Berman and Bui's (2001, p. 274-75) theoretical model,
as described above, the change in a firm's labor demand arising from a change in regulation is
decomposed into two main components: output and substitution effects.ccc As the output and
substitution effects may be both positive, both negative, or some combination, standard
neoclassical theory alone does not point to a definitive net effect of regulation on labor demand
at regulated firms.
Following the Berman and Bui framework for the impacts of regulation on employment
in the regulated sector, we consider two effects for the motor vehicle sector: the output effect
and the substitution effect.
8.10.2.1 The Output Effect
If truck or trailer sales increase, then more people will be required to assemble trucks,
trailers, and their components. If truck or trailer sales decrease, employment associated with
these activities will decrease. The effects of this final rulemaking on HD vehicle sales thus
depend on the perceived desirability of the new vehicles. On one hand, this final rulemaking will
increase truck and trailer costs; by itself, this effect would reduce truck and trailer sales. In
addition, while decreases in truck performance would also decrease sales, this program is not
expected to have any negative effect on truck performance. On the other hand, this final
rulemaking will reduce the fuel costs of operating the trucks; by itself, this effect would increase
truck sales, especially if potential buyers have an expectation of higher fuel prices. The agencies
have not made an estimate of the potential change in truck or trailer sales. However, as
discussed in Chapter 8.3, the agencies have estimated an increase in vehicle miles traveled {i.e.,
VMT rebound) due to the reduced operating costs of trucks meeting these standards. Since
increased VMT is most likely to be met with more drivers and more trucks, our projection of
VMT rebound is suggestive of an increase in vehicle sales and truck driver employment
ccc The authors also discuss a third component, the impact of regulation on factor prices, but conclude that this
effect is unlikely to be important for large competitive factor markets, such as labor and capital. Morgenstern, Pizer
and Shih (2002) use a very similar model, but they break the employment effect into three parts: 1) the demand
effect; 2) the cost effect; and 3) the factor-shift effect. See Morgenstern, Richard D., William A. Pizer, and Jhih-
Shyang Shih. "Jobs Versus the Environment: An Industry-Level Perspective." Journal of Environmental Economics
and Management 43 (2002): 412-436 (Docket EPA-HQ-OAR).

-------
(recognizing that these increases may be partially offset by a decrease in manufacturing and sales
for equipment of other modes of transportation such as rail cars or barges).
8.10.2.2 The Substitution Effect
The output effect, above, measures the effect due to new truck and trailer sales only. The
substitution effect includes the impacts due to the changes in technologies needed for vehicles to
meet these standards, separate from the effect on output (that is, as though holding output
constant). This effect includes both changes in employment due to incorporation of abatement
technologies and overall changes in the labor intensity of manufacturing. We present estimates
for this effect to provide a sense of the order of magnitude of expected impacts on employment,
which we expect to be small in the automotive sector, and to repeat that regulations may have
positive as well as negative effects on employment.
One way to estimate this effect, given the cost estimates for complying with the final
rules, is to use the ratio of workers to each $1 million of expenditures in that sector. The use of
these ratios has both advantages and limitations. It is often possible to estimate these ratios for
quite specific sectors of the economy: for instance, it is possible to estimate the average number
of workers in the motor vehicle body and trailer manufacturing sector per $1 million spent in the
sector, rather than use the ratio from another, more aggregated sector, such as motor vehicle
manufacturing. As a result, it is not necessary to extrapolate employment ratios from possibly
unrelated sectors. On the other hand, these estimates are averages for the sectors, covering all
the activities in those sectors; they may not be representative of the labor required when
expenditures are required on specific activities, or when manufacturing processes change
sufficiently that labor intensity changes. For instance, the ratio for the motor vehicle
manufacturing sector represents the ratio for all vehicle manufacturing, not just for emissions
reductions associated with compliance activities. In addition, these estimates do not include
changes in sectors that supply these sectors, such as steel or electronics producers. They thus
may best be viewed as the effects on employment in the motor vehicle sector due to the changes
in expenditures in that sector, rather than as an assessment of all employment changes due to
these changes in expenditures. In addition, this approach estimates the effects of increased
expenditures while holding constant the labor intensity of manufacturing; it does not take into
account changes in labor intensity due to changes in the nature of production. This latter effect
could either increase or decrease the employment impacts estimated here.DDD
Some of the costs of these final rules will be spent directly in the motor vehicle
manufacturing sector, but it is also likely that some of the costs will be spent in the motor vehicle
body and trailer and motor vehicle parts manufacturing sectors. The analysis here draws on
estimates of workers per $1 million of expenditures for each of these sectors.
There are several public sources for estimates of employment per $1 million
expenditures. The U.S. Bureau of Labor Statistics (BLS) provides its Employment
DDD As noted above, Morgenstern et al. (2002) separate the effect of holding output constant into two effects: the
cost effect, which holds labor intensity constant, and the factor shift effect, which estimates those changes in labor
intensity.

-------
Requirements Matrix (ERM),117 which provides direct estimates of the employment per $1
million in sales of goods in 202 sectors. The values considered here are for Motor Vehicle
Manufacturing (NAICS 3361), Motor Vehicle Body and Trailer Manufacturing (NAICS 3362),
and Motor Vehicle Parts Manufacturing (NAICS 3363) for 2014.
The Census Bureau provides the Annual Survey of Manufacturers118 (ASM), a subset of
the Economic Census (EC), based on a sample of establishments; though the EC itself is more
complete, it is conducted only every 5 years, while the ASM is annual. Both include more
sectoral detail than the BLS ERM: for instance, while the ERM includes the Motor Vehicle
Manufacturing sector, the ASM and EC have detail at the 6-digit NAICS code level (e.g., light
truck and utility vehicle manufacturing). While the ERM provides direct estimates of
employees/$l million in expenditures, the ASM and EC separately provide number of employees
and value of shipments; the direct employment estimates here are the ratio of those values. The
values reported are for Motor Vehicle Manufacturing (NAICS 3361), Light Truck and Utility
Vehicle Manufacturing (NAICS 336112), Heavy Duty Truck Manufacturing (33612), Motor
Vehicle Body and Trailer Manufacturing (3362), and Motor Vehicle Parts Manufacturing
(NAICS 3363). The values used here are adjusted to remove the employment effects of imports
through use of a ratio of domestic production to domestic sales of 0.78.EEE
Table 8-34 provides the values, either given (BLS) or calculated (ASM and EC) for
employment per $1 million of expenditures in 2014 (2012 for EC), all adjusted to 2013 dollars
using the Bureau of Economic Analysis's Implicit GDP Price Deflators.™ The different data
sources provide similar patterns for the estimates for the sectors. Body and trailer manufacturing
and parts manufacturing appear to be more labor-intensive than vehicle manufacturing; light
truck and utility vehicle manufacturing appears to be less, and heavy duty truck manufacturing
appears to be more, labor-intensive than motor vehicle manufacturing as a whole.
EEE To estimate the proportion of domestic production affected by the change in sales, we use data from Ward's
Automotive Group for total truck production in the U.S. compared to total truck sales in the U.S. For the period
2006-2015, the proportion is 78 percent (Docket EPA-HQ-OAR-), ranging from 68 percent (2009) to 83 percent
(2012) over that time.
FFF At the time of access, the EC data was only available by 2-, 3-, or 6-digit NAICS industry code. To construct the
4- and 5-digit numbers, we separately summed total employees and total expenditure for each 6-digit subcategory.

-------
Table 8-34 Employment per $1 Million Expenditures (2013$) in the Motor Vehicle Manufacturing Sector3
SOURCE
SECTOR
RATIO OF
WORKERS PER
$1 MILLION
EXPENDITURE
S
RATIO OF WORKERS
PER $1 MILLION
EXPENDITURES,
ADJUSTED FOR
DOMESTIC VS.
FOREIGN
PRODUCTION
BLS ERM
Motor vehicle mfg (3361)
0.393
0.306
BLS ERM
Motor vehicle body & trailer mfg (3362)
0.991
0.773
BLS ERM
Motor vehicle parts mfg (3363)
1.709
1.334
ASM
Motor vehicle mfg (3361)
0.582
0.454
ASM
Light truck & utility vehicle mfg
(336112)
0.468
0.365
ASM
Heavy duty truck mfg (33612)
1.018
0.794
ASM
Motor vehicle body & trailer mfg (3362)
3.189
2.489
ASM
Motor vehicle parts mfg (3363)
2.081
1.624
EC
Motor vehicle mfg (3361)
0.594
0.463
EC
Light truck & utility vehicle mfg
(336112)
0.472
0.369
EC
Heavy duty truck mfg (33612)
0.975
0.760
EC
Motor vehicle body & trailer mfg (3362)
3.502
2.733
EC
Motor vehicle parts mfg (3363)
2.126
1.659
Note:
" BLS ERM refers to the U.S. Bureau of Labor Statistics' Employment Requirement Matrix, 2014 values.
ASM refers to the U.S. Census Bureau's Annual Survey of Manufactures, 2014 values. EC refers to the U.S.
Census Bureau's Economic Census, 2012 values.
Over time, the amount of labor needed in the motor vehicle industry has changed:
automation and improved methods have led to significant productivity increases. The BLS
ERM, for instance, provided estimates that, in 1997, 1.09 workers in the Motor Vehicle
Manufacturing sector were needed per $1 million, but only 0.39 workers by 2014 (in 2013$).119
Because the ERM is available annually for 1997-2014, we used these data to estimate
productivity improvements over time. We regressed logged ERM values on a year trend for the
Motor Vehicle Manufacturing, Motor Vehicle Body and Trailer Manufacturing, and Motor
Vehicle Parts Manufacturing sectors. We used this approach because the coefficient describing
the relationship between time and productivity is a direct measure of the average percent change
in productivity per year. The results suggest a 6.6 percent per year productivity improvement in
the Motor Vehicle Manufacturing Sector, and a 4.9 percent per year improvement in the Motor
Vehicle Parts Manufacturing Sector. The Motor Vehicle Body and Trailer Manufacturing Sector
results were more complex: the workers/$l million values before 2010 are substantially higher
(averaging 3.98 in 2013$) than those in 2010 and after (averaging 1.28 in 2013$); we used
dummy variables to account for this shift, and estimate productivity gains of 1 percent per year
before 2010, and 14 percent after. This dramatic difference may suggest taking care when
relying on the data for this sector. As discussed further below, we only report maximum and
minimum employment impacts, and the Motor Vehicle Body and Trailer Manufacturing
estimates provide the maximum values in all but one year; they may therefore create greater
uncertainty about the upper bound of the substitution-effect employment.

-------
We then used the regression results to project the number of workers per $1 million
through 2027. We calculated separate sets of projections (adjusted to 2013$) for the BLS ERM
data as well as the EC and ASM for all three sectors discussed above. The BLS ERM
projections were calculated directly from the fitted regression equations since the regressions
themselves used ERM data. For the ASM and EC projections, we used the ERM's ratio of the
projected value in each future year to the projected value in 2014 for ASM and 2012 for EC (the
base years in our data) to determine how many workers will be needed per $1 million of 2013$.
In other words, we apply the projected productivity growth estimated using the ERM data to the
ASM and EC numbers.
Finally, to simplify the presentation and give a range of estimates, we compared the
projected employment among the 3 sectors for the ERM, EC, and ASM, and we provide only the
maximum and minimum employment effects estimated for the three data sources. We provide
the range rather than a point estimate because of the inherent difficulties in estimating
employment impacts; the range gives an estimate of the expected magnitude. The details of the
calculations may be found in the docket. The ERM estimates in the Motor Vehicle
Manufacturing Sector are consistently the minimum values. The ASM estimates in the Motor
Vehicle Body and Trailer Manufacturing Sector are the maximum values for all years but 2027,
where the ASM value for Motor Vehicle Parts Manufacturing provides the maximum value.
Chapter 7 of the RIA discusses the vehicle cost estimates developed for these final rules.
The final step in estimating employment impacts is to multiply costs (in $ millions) by workers
per $1 million in costs, to estimate employment impacts in the regulated and parts manufacturing
sectors. Increased costs of vehicles and parts would, by itself, and holding labor intensity
constant, be expected to increase employment between 2018 and 2027 between zero and 4.5
thousand each year.
While we estimate employment impacts, measured in job-years, beginning with program
implementation, some of these employment gains may occur earlier as motor vehicle
manufacturers and parts suppliers hire staff in anticipation of compliance with the standards. A
job-year is a way to calculate the amount of work needed to complete a specific task. For
example, a job-year is one year of work for one person.

-------
Table 8-35 Employment Effects due to Increased Costs of Vehicles and Parts (Substitution Effect), in Job-
years
YEAR
COSTS
(MILLIONS OF
2012$)
MINIMUM EMPLOYMENT
DUE TO SUBSTITUTION
EFFECT (ERM ESTIMATES,
EXPENDITURES IN THE
MOTOR VEHICLES MFG
SECTOR)
MAXIMUM EMPLOYMENT DUE
TO SUBSTITUTION EFFECT (ASM
ESTIMATES, EXPENDITURES IN
THE BODY AND TRAILER MFG
SECTORA)
2018
$227
0
400
2019
$215
0
400
2020
$220
0
300
2021
$2,270
300
3,100
2022
$2,243
300
2,900
2023
$2,485
300
2,900
2024
$3,890
400
4,200
2025
$4,146
400
4,100
2026
$4,203
400
3,800
2027
$5,219
500
4,500
Note:
" For 2027, the maximum employment effects are associated with ASM's Motor Vehicle Parts Manufacturing
sector.
8.10.2.3 Summary of Employment Effects in the Motor Vehicle Sector
The overall effect of these final rules on motor vehicle sector employment depends on the
relative magnitude of the output effect and the substitution effect. Because we do not have
quantitative estimates of the output effect, and only a partial estimate of the substitution effect,
we cannot reach a quantitative estimate of the overall employment effects of these final rules on
motor vehicle sector employment or even whether the total effect will be positive or negative.
The standards are not expected to provide incentives for manufacturers to shift
employment between domestic and foreign production. This is because the standards will apply
to vehicles sold in the U.S. regardless of where they are produced. If foreign manufacturers
already have increased expertise in satisfying the requirements of the standards, there may be
some initial incentive for foreign production, but the opportunity for domestic manufacturers to
sell in other markets might increase. To the extent that the requirements of these final rules
might lead to installation and use of technologies that other countries may seek now or in the
future, developing this capacity for domestic production now may provide some additional
ability to serve those markets.
Some vehicle parts are made in-house and would be included directly in the regulated
sector. Others are made by independent suppliers and are not directly regulated, but they will be
affected by the rules as well. The parts manufacturing sector will be involved primarily in
providing "add-on" parts, or components for replacement parts built internally. If demand for
these parts increases due to the increased use of these parts, employment effects in this sector are
expected to be positive. If the demand effect in the regulated sectors is significantly negative
enough, it is possible that demand for other parts may decrease. As noted, the agencies do not
predict a direction for the demand effect.

-------
8,10,3 Employment Impacts in Other Affected Sectors
8.10.3.1	Transport and Shipping Sectors
Although not directly regulated by these final rules, employment effects in the transport
and shipping sector are likely to result from these regulations. If the overall cost of shipping a
ton of freight decreases because of increased fuel efficiency (taking into account the increase in
upfront purchasing costs), in a perfectly competitive industry these costs savings, depending on
the relative elasticities of supply and demand, will be passed along to customers. With lower
prices, demand for shipping would lead to an increase in demand for truck shipping services
(consistent with the VMT rebound effect analysis) and therefore an increase in employment in
the truck shipping sector. In addition, if the relative cost of shipping freight via trucks becomes
cheaper than shipping by other modes (e.g., rail or barge), then employment in the truck
transport industry is likely to increase. If the trucking industry is more labor intensive than other
modes, we would expect this effect to lead to an overall increase in employment in the transport
and shipping sectors.I20-121 Such a shift would, however, be at the expense of employment in the
sectors that are losing business to trucking. The first effect - a gain due to lower shipping costs -
is likely to lead to a net increase in employment. The second effect, due to mode-shifting, may
increase employment in trucking, but decrease employment in other shipping sectors (e.g., rail or
barge), with the net effects dependent on the labor-intensity of the sectors and the volumes.
8.10.3.2	Fuel Suppliers
In addition to the effects on the trucking industry and related truck parts sector, these
final rules will result in reductions in fuel use that lower GHG emissions. Fuel saving,
principally reductions in liquid fuels such as diesel and gasoline, will affect employment in the
fuel suppliers industry sectors, principally the Petroleum Refinery sector.
Chapter 7 of this RIA provides estimates of the effects of these standards on expected
fuel consumption. While reduced fuel consumption represents savings for purchasers of fuel, it
also represents a loss in value of output for the petroleum refinery industry, which will result in
reduced sectoral employment. Because this sector is material-intensive, the employment effect is
not expected to be large.000
8.10.3.3	Fuel Savings
As a result of this rulemaking, it is anticipated that trucking firms will experience fuel
savings. Fuel savings lower the costs of transportation goods and services. In a competitive
market, some of the fuel savings that initially accrue to trucking firms are likely to be passed
along as lower transportation costs that, in turn, could result in lower prices for final goods and
services. Some of the savings might also be retained by firms for investments or for distributions
to firm owners. Again, how much accrues to customers versus firm owners will depend on the
relative elasticities of supply and demand. Regardless, the savings will accrue to some segment
GGG In the 2014 BLS ERM cited above, the Petroleum and Coal Products Manufacturing sector has a ratio of
workers per $1 million of 0.215, lower than all but two of the 181 sectors with non-zero employment per $1 million.

-------
of consumers: either owners of trucking firms or the general public, and the effect will be
increased spending by consumers in other sectors of the economy, creating jobs in a diverse set
of sectors, including retail and service industries.
As described in Preamble Section IX.C.(2), the value of fuel savings from this
rulemaking is projected to be $15.8 billion (2013$) in 2027, according to Table IX-6. If all those
savings are spent, the fuel savings will stimulate increased employment in the economy through
those expenditures. If the fuel savings accrue primarily to firm owners, they may either reinvest
the money or take it as profit. Reinvesting the money in firm operations could increase
employment directly. If they take the money as profit, to the extent that these owners are
wealthier than the general public, they may spend less of the savings, and the resulting
employment impacts would be smaller than if the savings went to the public. Thus, while fuel
savings are expected to decrease employment in the refinery sector, they are expected to increase
employment through increased consumer expenditures.
8.10.4 Summary of Employment Impacts
The primary employment effects of these rules are expected to be found throughout
several key sectors: truck and engine manufacturers, the trucking industry, truck parts
manufacturing, fuel production, and consumers. These rules initially take effect in model year
2018; the unemployment rate at that time is unknowable. In an economy with full employment,
the primary employment effect of a rulemaking is likely to be to move employment from one
sector to another, rather than to increase or decrease employment. For that reason, we focus our
partial quantitative analysis on employment in the regulated sector, to examine the impacts on
that sector directly. We discuss the likely direction of other impacts in the regulated sector as
well as in other directly related sectors, but we do not quantify those impacts, because they are
more difficult to quantify with reasonable accuracy, particularly so far into the future.
For the regulated sector, we have not quantified the output effect. The substitution effect
is associated with potential increased employment between zero and 4.5 thousand jobs per year
between 2018 and 2027, depending on the share of employment impacts in the affected sectors
(Motor Vehicle Manufacturing, Motor Vehicle Body and Trailer Manufacturing, and Motor
Vehicle Parts Manufacturing). These estimates do not include potential changes, either greater
or less, in labor intensity of production. As mentioned above, some of these job gains may occur
earlier as auto manufacturers and parts suppliers hire staff to prepare to comply with the
standard.
Lower prices for shipping are expected to lead to an increase in demand for truck
shipping services and, therefore, an increase in employment in that sector, though this effect may
be offset somewhat by changes in employment in other shipping sectors. Reduced fuel
production implies less employment in the fuel provision sectors. Finally, any net cost savings
are expected to be passed along to some segment of consumers: either the general public or the
owners of trucking firms, who are expected then to increase employment through their
expenditures. Under conditions of full employment, any changes in employment levels in the
regulated sector due to this program are mostly expected to be offset by changes in employment
in other sectors.

-------
8.11 Oil Price Sensitivity Analysis using Method B
In this section, EPA presents a sensitivity analysis examining the impact on net benefits
using AEO's "low oil price" and "high oil price" cases. The sensitivity analysis is based on the
final program relative to the flat baseline as the "primary" case using Method B. Fuel price
changes were not used as an input to technology application rates (i.e., a constant $/vehicle has
been used throughout this sensitivity analysis). The primary analysis (presented earlier in this
chapter) uses the AEO reference case oil prices. The primary case and both high and low oil
price case $/gallon values are shown in Table 8-36.
Table 8-36 AEO2015 Fuel Prices in the Low Oil Price Case, our Primary Analysis Case, and the AEO High
Oil Price Case (2013$)
YEAR
RETAIL
UNTAXED
Diesel
Gasoline
Diesel
Gasoline
Low
Primary
High
Low
Primary
High
Low
Primary
High
Low
Primary
High
2018
$2.50
$3.08
$4.79
$2.21
$2.70
$4.04
$2.04
$2.62
$4.33
$1.81
$2.30
$3.62
2019
$2.55
$3.12
$4.88
$2.28
$2.70
$4.11
$2.09
$2.66
$4.42
$1.88
$2.30
$3.69
2020
$2.61
$3.17
$4.97
$2.33
$2.74
$4.17
$2.16
$2.72
$4.52
$1.94
$2.35
$3.75
2021
$2.65
$3.23
$5.07
$2.35
$2.78
$4.26
$2.20
$2.78
$4.62
$1.96
$2.39
$3.85
2022
$2.70
$3.31
$5.18
$2.37
$2.82
$4.33
$2.25
$2.86
$4.73
$1.98
$2.43
$3.92
2023
$2.73
$3.37
$5.27
$2.39
$2.86
$4.41
$2.29
$2.93
$4.83
$2.00
$2.47
$4.00
2024
$2.76
$3.43
$5.42
$2.40
$2.90
$4.49
$2.32
$2.99
$4.98
$2.01
$2.52
$4.08
2025
$2.82
$3.49
$5.54
$2.40
$2.95
$4.56
$2.38
$3.05
$5.10
$2.02
$2.57
$4.16
2026
$2.86
$3.56
$5.66
$2.42
$3.00
$4.65
$2.43
$3.13
$5.23
$2.04
$2.62
$4.25
2027
$2.89
$3.63
$5.78
$2.43
$3.04
$4.74
$2.46
$3.20
$5.35
$2.05
$2.66
$4.34
2028
$2.90
$3.70
$5.92
$2.44
$3.09
$4.85
$2.47
$3.27
$5.49
$2.06
$2.71
$4.46
2029
$2.91
$3.77
$6.05
$2.45
$3.14
$4.96
$2.48
$3.35
$5.63
$2.08
$2.76
$4.57
2030
$2.91
$3.84
$6.17
$2.45
$3.20
$5.05
$2.49
$3.42
$5.75
$2.08
$2.82
$4.66
2031
$2.93
$3.92
$6.30
$2.47
$3.26
$5.16
$2.51
$3.50
$5.89
$2.10
$2.88
$4.76
2032
$2.93
$4.00
$6.43
$2.47
$3.33
$5.27
$2.51
$3.59
$6.02
$2.10
$2.95
$4.88
2033
$2.94
$4.09
$6.60
$2.49
$3.39
$5.40
$2.53
$3.68
$6.19
$2.12
$3.02
$5.01
2034
$2.95
$4.17
$6.74
$2.50
$3.46
$5.53
$2.54
$3.76
$6.34
$2.14
$3.09
$5.14
2035
$2.96
$4.26
$6.84
$2.52
$3.53
$5.64
$2.55
$3.86
$6.44
$2.16
$3.16
$5.25
2036
$2.98
$4.35
$6.99
$2.54
$3.60
$5.78
$2.58
$3.95
$6.59
$2.18
$3.23
$5.40
2037
$2.99
$4.45
$7.14
$2.55
$3.66
$5.89
$2.59
$4.05
$6.75
$2.19
$3.30
$5.51
2038
$3.00
$4.55
$7.29
$2.56
$3.74
$6.04
$2.60
$4.16
$6.90
$2.21
$3.38
$5.66
2039
$3.02
$4.65
$7.44
$2.58
$3.83
$6.17
$2.63
$4.26
$7.05
$2.23
$3.47
$5.79
2040
$3.03
$4.75
$7.61
$2.60
$3.90
$6.33
$2.64
$4.36
$7.22
$2.25
$3.54
$5.94
Note:
Our Primary case values are the AEO reference fuel price case values and are taken from AEO2015.
The impacts of using the low and high oil price cases on our estimated fuel savings and
net benefits are shown in Table 8-37.

-------
Table 8-37 MY2018-2029 Lifetime Sensitivity on Net Benefits using AEO2014 Low and High Oil Price Cases
for the Final Program Relative to the Flat Baseline and using Method B (Billions of 2013$; 3% Discounting)a

LOW OIL
PRIMARY
HIGH OIL

PRICE
CASE
PRICE

CASE

CASE
Vehicle program
-$27
-$27
-$27
Maintenance
-$1.9
-$1.9
-$1.9
Fuel
$119
$169
$282
Benefits
$86
$88
$94
Net benefits
$176
$229
$348
Note:
a For an explanation of analytical Methods A and B, please see
Preamble Section I.D; for an explanation of the flat baseline, la,
and dynamic baseline, lb, please see Preamble Section X. A. 1

-------
Appendix 8.A to Chapter 8 - Supplemental Analysis of
Quantified and Monetized Non-GHG Health and
Environmental Impacts
This appendix presents the results of our quantified and monetized criteria pollutant
health impacts analysis due to the Phase 2 standards compared to a future-year reference scenario
without the standards in place. Specifically, we present PM2.5- and ozone-related health benefits
of the standards for calendar year (CY) 2040.
As described in Chapter 8.6, we consider this analysis to be supplemental to the primary
analysis because, out of necessity, the air quality modeling was based on emissions inventories
that reflected the form of the standards as they were proposed, not finalized (air quality modeling
results are presented in Appendix 6A). The length of time needed to prepare the inventories and
run the air quality model requires EPA to make air quality modeling input decisions early in the
analytical process, and therefore made it impossible to base the health impacts analysis on the
emissions changes associated with the final rulemaking.
The chief limitation when using air quality inventories based on emissions from the
proposal is that they can diverge from the estimated emissions of the final rulemaking. How
much the emissions might diverge and how that difference would impact the air quality modeling
and health benefit results is difficult to anticipate. For the FRM, EPA concluded that when
comparing the proposal and final rule inventories, the differences were enough to justify the
move of the typical CY benefits analysis (based on air quality modeling) from the primary
estimate of costs and benefits to a supplemental analysis presented in this Chapter.HHH While we
believe this supplemental analysis is still illustrative of the standard's potential benefits, EPA has
instead chosen to characterize the CY benefits in the primary analysis in a manner consistent
with the MY lifetime analysis. That is, we apply PM-related "benefits per-ton" values to the CY
final rule emission reductions to estimate the PM-related benefits of the final rule.
8A.1 Quantified and Monetized Non-GHG Human Health Benefits of the 2040 Calendar
Year (CY) Analysis
8 A. 1.1	Overview
This section presents EPA's analysis of the criteria pollutant related health impacts
resulting from non-GHG emission reductions that can be expected to occur as a result of the
Phase 2 standards. CO2 emissions are predominantly the byproduct of fossil fuel combustion
processes that also produce criteria and hazardous air pollutant emissions. The vehicles that are
subject to the Phase 2 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 and will also affect emissions from upstream sources that occur during
111111 See Chapter 5 for a presentation and discussion of the differences between the proposal inventories used to
conduct the air quality modeling and the final rule inventories.

-------
the refining and distribution of fuel. Changes in ambient concentrations of ozone, PM2.5, and air
toxics that will result from the Phase 2 standards are expected to affect human health by reducing
premature deaths and other serious human health effects, as well as other important
improvements in public health and welfare. Children especially benefit from reduced exposures
to criteria and toxic pollutants, because they tend to be more sensitive to the effects of these
respiratory pollutants. Ozone and particulate matter have been associated with increased
incidence of asthma and other respiratory effects in children, and particulate matter has been
associated with a decrease in lung maturation. Some minority groups and children living under
the poverty line are even more vulnerable with higher prevalence of asthma.
The analysis in this section aims to characterize the benefits of the standards by
answering two key questions:
1.	What are the health and welfare effects of changes in ambient particulate matter
(PM2.5) and ozone air quality resulting from reductions in precursors including NOx and SO2?
2.	What is the economic value of these effects?
For the supplemental health benefits analysis, we have quantified and monetized the
health and environmental impacts in 2040, representing projected impacts associated with a year
when most of the fleet is turned over. Overall, we estimate that the standards will lead to a net
decrease in PM2.5- and ozone-related health impacts in 2040. The estimated decrease in
population-weighted national average PM2.5 exposure results in a net decrease in adverse PM-
related human health impacts (the decrease in national population-weighted annual average
PM2.5 is 0.01 [j,g/m3 in 2040).111 The estimated decrease in population-weighted national average
ozone exposure results in a net decrease in ozone-related health impacts (population-weighted
maximum 8-hour average ozone decreases by 0.21 ppb in 2040).
Using the lower end of EPA's range of preferred premature mortality estimates (Krewski
et al., 2009 for PM2.5 and Smith et al., 2009 for ozone),122'123 we estimate that by 2040,
implementation of the standards will reduce approximately 310 premature mortalities annually
and will yield between $2.8 and $3.0 billion in total annual benefits, depending on the discount
rate used.JJJ The upper end of the range of avoided premature mortality estimates associated
with the standards (based on Lepeule et al., 2012 for PM2.5 and Zanobetti and Schwartz, 2008 for
ozone)124'125 results in approximately 640 premature mortalities avoided in 2040 and will yield
between $5.9 and $6.4 billion in total benefits. Thus, even using the lower end of the range of
m Note that the national, population-weighted PM2 5 and ozone air quality metrics presented in this Chapter
represent an average for the entire, gridded U.S. CMAQ domain. These are different than the population-weighted
PM2 5 and ozone design value metrics presented in Chapter 7, which represent the average for areas with a current
air quality monitor.
111 The monetized value of PM2 5-related mortality accounts for a twenty-year segmented cessation lag. To discount
the value of premature mortality that occurs at different points in the future, we apply both a 3 and 7 percent
discount rate. We also use both a 3 and 7 percent discount rate to value PM-related nonfatal heart attacks
(myocardial infarctions). 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.

-------
premature mortality estimates, the criteria pollutant-related health benefits of the standards
presented in this rule are projected to be substantial.
We base our analysis of the rule's impact on human health and the environment on peer-
reviewed studies of air quality and human health effects.126"127 To model the ozone and PM air
quality impacts of the standards, we used the Community Multiscale Air Quality (CMAQ) model
(see Appendix 6A). The modeled ambient air quality data serves as an input to the
Environmental Benefits Mapping and Analysis Program - Community Edition version 1.1
(BenMAP-CE),IvIvK BenMAP-CE 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 in 2040 is presented
in Table 8A.1. We present total benefits (the sum of morbidity-related benefits and mortality-
related benefits) based on the PM- and ozone-related premature mortality function used. These
estimates represent EPA's preferred approach to characterizing a best estimate of benefits.
Table 8A-1 Estimated 2040 Monetized PM-and Ozone-Related Health Benefits (billions, 2013$)a

Discount
Rate
Benefits
Ozone Benefits 0
b
$1.0 to $1.8
PM2.5 Benefits d
3%
$2.0 to $4.6

7%
$1.8 to $4.2
Total Benefits
3%
$3.0 to $6.4e

7%
$2.8 to $5.9e
Notes:
a Rounded to two significant figures. These estimates reflect the economic value of avoided morbidities and
premature deaths using risk coefficients from the studies noted in Table 8A-8.
b Ozone-only benefits reflect short-term exposure impacts and as such are assumed to occur in the same year as
ambient ozone reductions. Consequently, social discounting is not applied to the benefits for this category.
0 Range reflects application of effect estimates from Smith et al. (2009) and Zanobetti and Schwartz (2008).
d Range reflects application of effect estimates from Krewski et al. (2009) and Lepeule et al. (2012).
e Excludes additional health and welfare benefits which could not be quantified (see Table 8A-2).
The benefits in Table 8A-1 include all of the human health impacts we are able to
quantify and monetize at this time. However, the full complement of human health and welfare
effects associated with PM, ozone, and other criteria pollutants 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, PM, and other criteria pollutants 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
KKK Information on BenMAP, including downloads of the software, can be found at https://www3 .epa.gov/benmap.

-------
cultural monuments and other materials, and environmental benefits due to reductions of impacts
of eutrophication in coastal areas. These are listed in Table 8A-2. As a result, the health benefits
quantified in this section are likely underestimates of the total benefits attributable to the
standards.
Table 8A-2 lists the PM- and ozone-related benefits categories we will use to quantify the
non-GHG incidence impacts associated with the standards. Table 8A-2 also lists non-GHG-related
endpoints we are currently unable to quantify and/or monetize.
Table 8A-2 Estimated Quantified and Unquantified Health Effects
Benefits Category
Specific Effect
Effect Has Been
Quantified
Effect Has Been
Monetized
More Information
Improved Human Health
Reduced
incidence of
premature
mortality and
Adult premature mortality based on
cohort study estimates and expert
elicitation estimates (age >25 or age
>30)
S
S
PM NAAQS
RIA,128 Section
5.6
morbidity from
exposure to PM2 5
Infant mortality (age <1)
V
V
PM NAAQS RIA,
Section 5.6

Non-fatal heart attacks (age >18)
V
V
PM NAAQS RIA,
Section 5.6

Hospital admissions—respiratory (all
ages)
V
V
PM NAAQS RIA,
Section 5.6

Hospital admissions—cardiovascular
(age >20)
V
V
PM NAAQS RIA,
Section 5.6

Emergency department visits for
asthma (all ages)
V
V
PM NAAQS RIA,
Section 5.6

Acute bronchitis (age 8-12)
V
V
PM NAAQS RIA,
Section 5.6

Lower respiratory symptoms (age 7-
14)
~
~
PM NAAQS RIA,
Section 5.6

Upper respiratory symptoms
(asthmatics age 9-11)
~
~
PM NAAQS RIA,
Section 5.6

Asthma exacerbation (asthmatics age
6-18)
~
~
PM NAAQS RIA,
Section 5.6

Lost work days (age 18-65)
~
~
PM NAAQS RIA,
Section 5.6

Minor restricted-activity days (age
18-65)
~
~
PM NAAQS RIA,
Section 5.6

Chronic Bronchitis (age >26)
—
—
PM NAAQS RIA,
Section 5.6°

Emergency department visits for
cardiovascular effects (all ages)
—
—
PM NAAQS RIA,
Section 5.6°

Strokes and cerebrovascular disease
(age 50-79)
—
—
PM NAAQS RIA,
Section 5.6°

Other cardiovascular effects (e.g.,
other ages)
—
—
PM ISA3,129

Other respiratory effects (e.g.,
pulmonary function, non-asthma ER
visits, non-bronchitis chronic
diseases, other ages and populations)


PM ISA3

-------
Benefits Category
Specific Effect
Effect Has Been
Quantified
Effect Has Been
Monetized
More Information

Reproductive and developmental
effects (e.g., low birth weight, pre-
term births, etc.)


PM ISAa'b

Cancer, mutagenicity, and
genotoxicity effects
—
—
PM ISAa,b
Reduced
incidence of
premature
Premature mortality based on short-
term study estimates (all ages)
V
V
Ozone NAAQS
RIA,130 Section
6.6
mortality and
morbidity from
Premature mortality based on long-
term study estimates (age 30-99)
d
d
Ozone NAAQS
RIA, Section 6.6
exposure to
ozone
Hospital admissions—respiratory
causes (age > 65)
V
V
Ozone NAAQS
RIA, Section 6.6

Emergency department visits for
asthma (all ages)
V
V
Ozone NAAQS
RIA, Section 6.6

Asthma exacerbation (age 6-18)
V
V
Ozone NAAQS
RIA, Section 6.6

Minor restricted-activity days (age
18-65)
V
V
Ozone NAAQS
RIA, Section 6.6

School absence days (age 5-17)
V
V
Ozone NAAQS
RIA, Section 6.6

Decreased outdoor worker
productivity (age 18-65)
d
d
Ozone NAAQS
RIA, Section 6.6

Other respiratory effects (e.g.,
premature aging of lungs)
—
—
Ozone ISAa131

Cardiovascular and nervous system
effects
—
—
Ozone ISAb

Reproductive and developmental
effects
	
—
Ozone ISAb

-------
Benefits Category
Specific Effect
Effect Has Been
Quantified
Effect Has Been
Monetized
More Information
Reduced
incidence of
morbidity from
exposure to air
toxics
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)


IRI§a,b,132
Notes:
a We assess these benefits qualitatively because we do not have sufficient confidence in available data or methods.
b We assess these benefits qualitatively because current evidence is only suggestive of causality or there are other
significant concerns over the strength of the association.
0 We assess these benefits qualitatively due to time and resource limitations for this analysis.
d We quantified these benefits, but they are not part of the core monetized benefits.
While there will be impacts associated with air toxic pollutant emission changes that
result from the standards, 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.133 While EPA has since
improved these tools, there remain critical limitations for estimating incidence and assessing
benefits of reducing mobile source air toxics.
As part of the second prospective analysis of the benefits and costs of the Clean Air
Act,134 EPA conducted a case study analysis of the health effects associated with reducing
exposure to benzene in Houston from implementation of the Clean Air Act. While reviewing the
draft report, EPA's Advisory Council on Clean Air Compliance Analysis concluded that "the
challenges for assessing progress in health improvement as a result of reductions in emissions of

-------
hazardous air pollutants (HAPs) are daunting...due to a lack of exposure-response functions,
uncertainties in emissions inventories and background levels, the difficulty of extrapolating risk
estimates to low doses and the challenges of tracking health progress for diseases, such as
cancer, that have long latency periods."135 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 standards.LLL
The reduction in air pollution emissions that will result from the standards is projected to
have "welfare" co-benefits in addition to human health benefits, including changes in visibility,
materials damage, ecological effects from PM deposition, ecological effects from nitrogen and
sulfur emissions, vegetation effects from ozone exposure, and climate effects. Despite our goal
to quantify and monetize as many of the benefits as possible for the standards, the welfare co-
benefits of the standards remain unquantified and non-monetized in this RIA due to data,
methodology, and resource limitations. As a result, the benefits quantified in this analysis are
likely underestimates of the total benefits attributable to the standards. We refer the reader to
Chapter 6 of the PM NAAQS RIA for a complete discussion of these welfare co-benefits.136
8A. 1.2	Human Health Impacts
Table 8A-3 and Table 8A-4 present the annual PM2.5 and ozone health impacts in the 48
contiguous U.S. states associated with the standards. For each endpoint presented in the tables,
we provide both the point estimate and the 90 percent confidence interval.
Using EPA's preferred estimates, based on the American Cancer Society (ACS) and Six-
Cities studies and a no threshold assumption in the model of mortality, we estimate that the
standards will result in between 210 and 480 cases of avoided PM2.5-related premature deaths
annually in 2040. For ozone, changes in mortality risk are estimated using two ozone-related
short-term effect estimates, Smith et al., 2009 and Zanobetti and Schwartz, 2008, consistent with
the 2015 Ozone NAAQS RIA. We estimate that the standards will result in between 99 and 210
cases of avoided ozone-related premature mortalities.
LLL In April, 2009, EPA hosted a workshop on estimating the benefits or reducing hazardous air pollutants. This
workshop built upon the work accomplished in the June 2000 Science Advisory Board/EPA Workshop on the
Benefits of Reductions in Exposure to Hazardous Air Pollutants, which generated thoughtful discussion on
approaches to estimating human health benefits from reductions in air toxics exposure, but no consensus was
reached on methods that could be implemented in the near term for a broad selection of air toxics. Please visit
http://epa.gov/air/toxicair/2009workshop.html for more information about the workshop and its associated materials.

-------
Table 8A-3 Estimated PIVh.s-Related Health Impacts3
Health Effect
2040 Annual Reduction in
Incidence (5th - 95th percentile)
Premature Mortality - Derived from epidemiology literature15
Adult, age 30+, ACS Cohort Study (Krewski et al., 2009)
Adult, age 25+, Six-Cities Study (Lepeule et al., 2012)
Infant, age <1 year (Woodruff et al., 1997)
210
(150-270)
480
(280 - 680)
0.29
(0.14-0.44)
Non-fatal myocardial infarction (adult, age 18 and over)
Peters et al. (2001)
Pooled estimate of 4 studies
260
(93 - 420)
27
(13-60)
Hospital admissions - respiratory (all ages)0, e
66
(-15 - 120)
Hospital admissions - cardiovascular (adults, age >18)d
56
(25-110)
Emergency room visits for asthma (age 18 years and younger)6
96
(-17-190)
Acute bronchitis, (children, age 8-12)e
280
(-10 - 580)
Lower respiratory symptoms (children, age 7-14)
3,600
(1,700 - 5,500)
Upper respiratory symptoms (asthmatic children, age 9-18)
5,200
(1,600 - 8,700)
Asthma exacerbation (asthmatic children, age 6-18)
5,400
(680- 11,000)
Work loss days
23,000
(20,000 - 26,000)
Minor restricted activity days (adults age 18-65)
140,000
(120,000 - 160,000)
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 most recent American Cancer Society (ACS) Cohort Study
(Krewski et al., 2009) and the most recent Six-Cities Study (Lepeule et al., 2012). 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).137
0 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.
e The negative estimates at the 5th percentile confidence estimates for these morbidity endpoints reflect the
statistical power of the study used to calculate these health impacts. These results do not suggest that reducing
air pollution results in additional health impacts.

-------
Table 8A-4 Estimated Ozone-Related Health Impactsa b
Health Effect
2040 Annual Reduction in Incidence
(5th - 95th percentile)
Short-Term Premature Mortality, All agesb
Multi-City Analyses
Smith et al. (2009)
Zanobetti and Schwartz (2008)
99
(55 - 140)
160
(98 - 230)
Hospital admissions- respiratory causes (adult, 65 and older)0
200
(-14-410)
Emergency room visit for asthma (all ages)
510
(84 - 1,200)
Asthma exacerbation (age 6-18)°
170,000
(-96,000 - 390,000)
Minor restricted activity days (adults, age 18-65)
410,000
(200,000 - 620,000)
School absence days
140,000
(62,000 - 270,000)
Notes:
a All incidence estimates are rounded to whole numbers with a maximum of two significant digits.
b All incidence estimates are based on ozone-only models unless otherwise noted.
0 The negative estimates at the 5th percentile confidence estimates for these morbidity endpoints reflect the statistical
power of the studies used to calculate these health impacts. These results do not suggest that reducing air pollution
results will adversely affect health, but rather, that we are less confident in the magnitude of the expected benefits
for this endpoint.
8A.1.3	Monetized Estimates of Human Health and Environmental Impacts
Table 8A-5 presents the estimated monetary value of changes in the incidence of ozone
and PM2.5-related health and environmental effects. Total aggregate monetized benefits are
presented in Table 8A-6. All monetized estimates are presented in 2013 dollars. Where
appropriate, estimates account for growth in real gross domestic product (GDP) per capita
between 2000 and 2040.MMM The monetized value of PM2.5-related mortality also accounts for a
twenty-year segmented cessation lag.™™ To discount the value of premature mortality that
MMM 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 2040, 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. 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.
NNN 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

-------
occurs at different points in the future, we apply both a 3 and 7 percent discount rate. We also
use both a 3 and 7 percent discount rate to value PM-related nonfatal heart attacks (myocardial
infarctions).000
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 standards 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 non-monetized health and welfare benefits. Our estimate of total monetized benefits
in 2040 for the standards, using the ACS and Six-Cities PM mortality studies and the two ozone
mortality studies, is between $3.0 and $6.4 billion, assuming a 3 percent discount rate, or
between $2.8 and $5.9 billion, assuming a 7 percent discount rate. As the results indicate, total
benefits are driven primarily by the reduction in PM2.5- and ozone-related premature fatalities
each year.
The next largest benefit is for reductions in 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 many more work loss days than PM-related premature mortalities, 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 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.
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.
000 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.

-------
Table 8A-5 Estimated Monetary Value of Changes in Incidence of Health and Welfare Effects (millions of
2013$) a-b

2040
PM2.5-Related Health Effect
(5 th and 95 th Percentile)
Premature Mortality -
Derived from Epidemiology
Studies°'d
Adult, age 30+ - ACS study
(Krewski et al., 2009)
3% discount rate
7% discount rate
$2,000
($300 - $4,700)
$1,800
($270 - $4,200)
Adult, age 25+ - Six-Cities study
(Lepeule et al., 2012)
3% discount rate
7% discount rate
$4,500
($650 -$11,000)
$4,100
($580 - $9,900)
Infant Mortality, <1 year -
(Woodruff et al. 1997)
$3.1
($0.41 - $7.6)
Non-fatal myocardial infarction (adult, age 18 and over)
Peters et al. (2001)
3% Discount Rate
7% Discount Rate
Pooled estimate of 4 studies
3% Discount Rate
7% Discount Rate
$31
($6.6 - $74)
$31
($6.1 -$74)
$3.4
($0.83 - $9.0)
$3.3
($0.77 - $8.9)
Hospital admissions for respiratory causes'1
$2.0
(-$0.61 -$3.8)
Hospital admissions for cardiovascular causes
$3.2
($1.6 -$5.7)
Emergency room visits for asthmad
$0.04
(-$0,007 - $0.09)
Acute bronchitis (children, age 8-12)d
$0.15
(-$0,005 - $0.36)
Lower respiratory symptoms (children, 7-14)
$0.08
($0.03 -$0.15)
Upper respiratory symptoms (asthma, 9-11)
$0.19
($0.05 - $0.41)
Asthma exacerbations
$0.34
($0.04 - $0.80)
Work loss days
$4.0
($3.5 -$4.6)
Minor restricted-activity days (MRADs)
$10
($5.9 -$15)
Ozone-Related Health Effects
Premature Mortality, All ages -
Derived from Multi-city analyses
Smith etal., 2009
$940
($47 - $2,500)
Zanobetti &
Schwartz, 2008
$1,700
($250 - $4,200)
Hospital admissions- respiratory causes (adult, 65 and older)d
$6.7
(-$0.48 -$14)

-------
Emergency room visit for asthma (all ages)
$0.23
($0.04 - $0.53)
Asthma exacerbation (age 6-18)d
$11
(-$4.2 - $29)
Minor restricted activity days (adults, age 18-65)
$30
($13 -$52)
School absence days
$15
($6.5 - $28)
Notes:
a 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 (2040).
0 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.
dThe negative estimate at the 5th percentile confidence estimate for this morbidity endpoint reflects the
statistical power of the study used to calculate this health impact. This result does not suggest that reducing
air pollution results in additional health impacts.
Table 8A-6 Estimated 2040 Monetized PM-and Ozone-Related Health Benefits3

Discount
Rate
Benefits
Ozone Benefits 0
b
$1.0 to $1.8
PM2.5 Benefits d
3%
$2.0 to $4.6

7%
$1.8 to $4.2
Total Benefits
3%
$3.0 to $6.4e

7%
$2.8 to $5.9e
Notes:
a Rounded to two significant figures. These estimates reflect the economic value of avoided morbidities and
premature deaths using risk coefficients from the studies noted in Table 8A-8.
b Ozone-only benefits reflect short-term exposure impacts and as such are assumed to occur in the same year as
ambient ozone reductions. Consequently, social discounting is not applied to the benefits for this category.
0 Range reflects application of effect estimates from Smith et al. (2009) and Zanobetti and Schwartz (2008).
d Range reflects application of effect estimates from Krewski et al. (2009) and Lepeule et al. (2012).
e Excludes additional health and welfare benefits which could not be quantified (see Table 8A-2).
8A.1.4	Methodology
We follow a "damage-function" approach in calculating total benefits of the modeled
changes in environmental quality. This approach estimates changes in individual health
endpoints (specific effects that can be associated with changes in air quality) and assigns values
to those changes assuming independence of the values for those individual endpoints. Total
benefits are calculated simply as the sum of the values for all non-overlapping health endpoints.
The "damage-function" approach is the standard method for assessing costs and benefits of
environmental quality programs and has been used in several recent published analyses.138'139'140
To assess economic value in a damage-function framework, the changes in environmental
quality must be translated into effects on people or on the things that people value. In some

-------
cases, the changes in environmental quality can be directly valued. In other cases, such as for
changes in ozone and PM, an impact analysis must first be conducted to convert air quality
changes into effects that can be assigned dollar values. For the purposes of this RIA, the health
impacts analysis (HIA) includes those health effects that are directly linked to ambient levels of
air pollution and specifically to those linked to ozone and PM2.5.
We note at the outset that the EPA rarely has the time or resources to perform extensive
new research to measure directly either the health outcomes or their values for regulatory
analyses. Thus, similar to Kunzli et al. (2000)141 and other, more recent health impact analyses,
our estimates are based on the best available methods of benefits transfer. Benefits transfer is the
science and art of adapting primary research from similar contexts to obtain the most accurate
measure of benefits for the environmental quality change under analysis. Adjustments are made
for the level of environmental quality change, the socio-demographic and economic
characteristics of the affected population, and other factors to improve the accuracy and
robustness of benefits estimates.
8A. 1.4.1 Human Health Impact Assessment
The health impact assessment (HIA) quantifies the changes in the incidence of adverse
health impacts resulting from changes in human exposure to PM2.5 and ozone air quality. HIAs
are a well-established approach for estimating the retrospective or prospective change in adverse
health impacts expected to result from population-level changes in exposure to pollutants.142
PC-based tools such as the environmental Bent fits Mapping and Analysis Program (BenMAP)
can systematize health impact analyses by applying a database of key input parameters, including
health impact functions and population projections—provided that key input data are available,
including air quality estimates and risk coefficients.143 Analysts have applied the HIA approach
to estimate human health impacts resulting from hypothetical changes in pollutant
levels.144'145,146 The EPA and others have relied upon this method to predict future changes in
health impacts expected to result from the implementation of regulations affecting air quality.147
For this assessment, the HIA is limited to those health effects that are directly linked to ambient
ozone and PM2.5 concentrations.
The HIA approach used in this analysis involves three basic steps: (1) utilizing
projections of PM2.5 air qualityppp and estimating the change in the spatial distribution of the
ambient air quality; (2) determining the subsequent change in population-level exposure; (3)
calculating health impacts by applying concentration-response relationships drawn from the
epidemiological literature to this change in population exposure.
A typical health impact function might look like:
Ay = y0-(e^-l),
ppp Projections of ambient PM2 5 concentrations for this analysis were generated using the Community Multiscale Air
Quality model (CMAQ). See Chapter 7 of this RIA for more information on the air quality modeling.

-------
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 in Chapter 8.1.2.5.3. Chapter 8.1.2.4.6 describes the ozone and PM air quality inputs to
the health impact functions.
Potentially Affected Populations
Quantified and monetized human health impacts depend on the demographic
characteristics of the population, including age, location, and income. We use population
projections based on economic forecasting models developed by Woods and Poole, Inc.148 The
Woods and Poole (WP) database contains county-level projections of population by age, sex, and
race out to 2040, relative to a baseline using the 2010 Census data. Projections in each county
are determined simultaneously with every other county in the United States to take into account
patterns of economic growth and migration. The sum of growth in county-level populations is
constrained to equal a previously determined national population growth, based on Bureau of
Census estimates.149 According to WP, linking county-level growth projections together and
constraining to a national-level total growth avoids potential errors introduced by forecasting
each county independently. County projections are developed in a four-stage process:
¦	First, national-level variables such as income, employment, and populations are
forecasted.
Second, employment projections are made for 179 economic areas defined by the
Bureau of Economic Analysis,150 using an "export-base" approach, which relies on
linking industrial-sector production of non-locally consumed production items, such
as outputs from mining, agriculture, and manufacturing with the national economy.
The export-based approach requires estimation of demand equations or calculation of
historical growth rates for output and employment by sector.
Third, population is projected for each economic area based on net migration rates
derived from employment opportunities and following a cohort-component method
based on fertility and mortality in each area.
¦	Fourth, employment and population projections are repeated for counties, using the
economic region totals as bounds. The age, sex, and race distributions for each
region or county are determined by aging the population by single year of age by sex
and race for each year through 2040 based on historical rates of mortality, fertility,
and migration.
Effect Estimate Sources
The first step in selecting effect coefficients is to identify the health endpoints to be
quantified. We base our selection of health endpoints on consistency with the EPA's Integrated
Science Assessments (which replace previous Criteria Documents), with input and advice from

-------
the SAB-HES, a scientific review panel specifically established to provide advice on the use of
the scientific literature in developing benefits analyses for the EPA's Report to Congress on The
Benefits and Costs of the Clean Air Act 1990 to 2020.151 In addition, we included more recent
epidemiology studies from the ozone ISA, PM ISA, and the PM Provisional
Assessment. 15Z151154 In general, we follow a weight of evidence approach, based on the
biological plausibility of effects, availability of concentration-response functions from well
conducted peer-reviewed epidemiological studies, cohesiveness of results across studies, and a
focus on endpoints reflecting public health impacts (like hospital admissions) rather than
physiological responses (such as changes in clinical measures like Forced Expiratory Volume
[FEV1]).
There are several types of data that can support the determination of types and magnitude
of health effects associated with air pollution exposures. These sources of data include
toxicological studies (including animal and cellular studies), human clinical trials, and
observational epidemiology studies. All of these data sources provide important contributions to
the weight of evidence surrounding a particular health impact. However, only epidemiology
studies provide direct concentration-response relationships that can be used to evaluate
population-level impacts of reductions in ambient pollution levels in a health impact assessment.
For the data-derived estimates, we relied on the published scientific literature to ascertain
the relationship between PM2.5, ozone, and adverse human health effects. We evaluated
epidemiological studies using the selection criteria summarized in Table 8A-7. These criteria
include consideration of whether the study was peer-reviewed, the match between the pollutant
studied and the pollutant of interest, the study design and location, and characteristics of the
study population, among other considerations. In general, the use of concentration-response
functions from more than a single study can provide a more representative distribution of the
effect estimate. However, there are often differences between studies examining the same
endpoint, making it difficult to pool the results in a consistent manner. For example, studies may
examine different pollutants or different age groups. For this reason, we consider very carefully
the set of studies available examining each endpoint and select a consistent subset that provides a
good balance of population coverage and match with the pollutant of interest. In many cases,
either because of a lack of multiple studies, consistency problems, or clear superiority in the
quality or comprehensiveness of one study over others, a single published study is selected as the
basis of the effect estimate.
When several effect estimates for a pollutant and a given health endpoint have been
selected (with the exception of mortality), they are quantitatively combined or pooled to derive a
more robust estimate of the relationship. The BenMAP Manual Technical Appendices provides
details of the procedures used to combine multiple impact functions.155 In general, we used
fixed or random effects models to pool estimates from different single city studies of the same
endpoint. Fixed effects pooling simply weights each study's estimate by the inverse variance,
giving more weight to studies with greater statistical power (lower variance). Random effects
pooling accounts for both within-study variance and between-study variability, due, for example,
to differences in population susceptibility. We used the fixed effects model as our null
hypothesis and then determined whether the data suggest that we should reject this null

-------
hypothesis, in which case we would use the random effects model. QQQ Pooled impact functions
are used to estimate hospital admissions and asthma exacerbations. When combining evidence
across multi-city studies (e.g., cardiovascular hospital admission studies), we use equal weights
pooling. The effect estimates drawn from each multi-city study are themselves pooled across a
large number of urban areas. For this reason, we elected to give each study an equal weight
rather than weighting by the inverse of the variance reported in each study. For more details on
methods used to pool incidence estimates, see the BenMAP Manual Appendices.
Effect estimates selected for a given health endpoint were applied consistently across all
locations nationwide. This applies to both impact functions defined by a single effect estimate
and those defined by a pooling of multiple effect estimates. Although the effect estimate may, in
fact, vary from one location to another (e.g., because of differences in population susceptibilities
or differences in the composition of PM), location-specific effect estimates are generally not
available.
QQQ EPA recently changed the algorithm BenMAP uses to calculate study variance, which is used in the pooling
process. Prior versions of the model calculated population variance, while the version used here calculated sample
variance. This change did not affect the selection of random or fixed effects for the pooled incidence estimates
between the proposal and final RIA.

-------
Table 8A-7 Criteria Used When Selecting C-R Function
Consideration
Comments
Peer-Reviewed
Research
Peer-reviewed research is preferred to research that has not undergone the peer-review
process.
Study Type
Among studies that consider chronic exposure (e.g., over a year or longer), prospective
cohort studies are preferred over ecological studies because they control for important
individual-level confounding variables that cannot be controlled for in ecological studies.
Study Period
Studies examining a relatively longer period of time (and therefore having more data) are
preferred, because they have greater statistical power to detect effects. Studies that are
more recent are also preferred because of possible changes in pollution mixes, medical
care, and lifestyle over time. However, when there are only a few studies available,
studies from all years will be included.
Population Attributes
The most technically appropriate measures of benefits would be based on impact
functions that cover the entire sensitive population but allow for heterogeneity across age
or other relevant demographic factors. In the absence of effect estimates specific to age,
sex, preexisting condition status, or other relevant factors, it may be appropriate to select
effect estimates that cover the broadest population to match with the desired outcome of
the analysis, which is total national-level health impacts. When available, multi-city
studies are preferred to single city studies because they provide a more generalizable
representation of the concentration-response function.
Study Size
Studies examining a relatively large sample are preferred because they generally have
more power to detect small magnitude effects. A large sample can be obtained in several
ways, including through a large population or through repeated observations on a smaller
population (e.g., through a symptom diary recorded for a panel of asthmatic children).
Study Location
U.S. studies are more desirable than non-U. S. studies because of potential differences in
pollution characteristics, exposure patterns, medical care system, population behavior,
and lifestyle. National estimates are most appropriate when benefits are nationally
distributed; the impact of regional differences may be important when benefits only
accrue to a single area.
Pollutants Included in
Model
When modeling the effects of ozone and PM (or other pollutant combinations) jointly, it
is important to use properly specified impact functions that include both pollutants.
Using single-pollutant models in cases where both pollutants are expected to affect a
health outcome can lead to double-counting when pollutants are correlated.
Measure of PM
For this analysis, impact functions based on PM2 5 are preferred to PM10 because of the
focus on reducing emissions of PM2 5 precursors, and because air quality modeling was
conducted for this size fraction of PM. Where PM2.5 functions are not available, PM10
functions are used as surrogates, recognizing that there will be potential downward
(upward) biases if the fine fraction of PM10 is more (less) toxic than the coarse fraction.
Economically
Valuable Health
Effects
Some health effects, such as forced expiratory volume and other technical measurements
of lung function, are difficult to value in monetary terms. These health effects are not
quantified in this analysis.
Non-overlapping
Endpoints
Although the benefits associated with each individual health endpoint may be analyzed
separately, care must be exercised in selecting health endpoints to include in the overall
benefits analysis because of the possibility of double-counting of benefits.
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 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 8A-8 lists the health
endpoints included in this analysis.

-------
Table 8A-8 Health Impact Functions Used in BenMAP to Estimate Impacts of PM2.5 and Ozone Reductions



Relative Risk or Effect Estimate (P)


Study
(with 95th Percentile Confidence
Endpoint
Study
Population
Interval orSE)
PM-related Health Impacts
Premature Mortality
Premature mortality-
Krewski et al. (2009)156
> 29 years
RR = 1.06 (1.04-1.06) per 10 pg/m3
cohort study, all-cause
Lepeule et al. (2012)157
> 24 years
RR = 1.14 (1.07-1.22) per 10 pg/m3
Premature mortality—
Woodruff et al. (1997)158
Infant (< 1
OR = 1.04 (1.02-1.07) per 10 pg/m3
all-cause

year)

Chronic Illness
Nonfatal heart attacks
Peters et al. (2001)159
Adults (> 18
OR = 1.62 (1.13-2.34) per 20 pg/m3

Pooled estimate:
years)


Pope et al. (2006)160

P = 0.00481 (0.00199)

Sullivan et al. (2005)161

P = 0.00198 (0.00224)

Zanobetti et al. (2009)

P = 0.00225 (0.000591)

Zanobetti and Schwartz

P = 0.0053 (0.00221)

(2006)162


Hospital Admissions
Respiratory
Zanobetti et al. (2009)—ICD
> 64 years
P=0.00207 (0.00446)

460-519 (All respiratory)



Kloog et al. (2012)163-ICD 460-

P=0.0007 (0.000961)

519 (All Respiratory



Moolgavkar (2000)164—ICD
18-64 years
1.02 (1.01-1.03) per 36 pg/m3

490-496 (Chronic lung disease)



Babin et al. (2007)165-ICD 493
< 19 years
P=0.002 (0.004337)

(asthma)



Sheppard (2003)166-ICD 493
(asthma)
< 18
RR = 1.04 (1.01-1.06) per 11.8 pg/m3
Cardiovascular
Pooled estimate:
> 64 years


Zanobetti et al. (2009)— ICD

P=0.00189 (0.000283)

390-459 (all cardiovascular)



Peng et al. (2009)167-ICD 426-

P=0.00068

427; 428; 430-438; 410-414;

(0.000214)

429; 440-449 (Cardio-, cerebro-



and peripheral vascular disease)



Peng et al. (2008)168— ICD 426-

P=0.00071

427; 428; 430-438; 410-414;

(0.00013)

429; 440-449 (Cardio-, cerebro-



and peripheral vascular disease)



Bell et al. (2008)169— ICD 426-

P=0.0008

427; 428; 430-438; 410-414;

(0.000107)

429; 440-449 (Cardio-, cerebro-



and peripheral vascular disease)



Moolgavkar (2000)170—ICD
20-64 years
RR=1.04 (t statistic: 4.1) per 10 pg/m3

390-429 (all cardiovascular)


Asthma-related
Pooled estimate:
All ages

emergency department
Mar et al. (2010)171
RR = 1.04 (1.01-1.07) per 7 pg/m3
visits
Slaughter et al. (2005)172

RR = 1.03 (0.98-1.09) per 10 pg/m3

Glad et al. (2012)173

P=0.00392 (0.002843)
Other Health Endpoints

-------
Acute bronchitis
Dockery et al. (1996)1
8-12 years OR = 1.50 (0.91-2.47) per 14.9 ng/m3
Asthma exacerbations
Work loss days
Acute respiratory
symptoms (MRAD)
Upper respiratory
symptoms
Lower respiratory
symptoms
Pooled estimate:
Ostro et al. (2001)175 (cough,
wheeze, shortness of breath)b
Mar et al. (2004)176 (cough,
shortness of breath)
Ostro (1987)177
Ostro and Rothschild (1989)178
(Minor restricted activity days)
Pope et al. (1991)179
Schwartz and Neas (2000)180
6-18 years b OR = 1.03 (0.98-1.07)
OR = 1.06 (1.01-1.11)
OR = 1.08 (1.00-1.17) per 30 ng/m3
RR= 1.21 (1-1.47) per
RR = 1.13 (0.86-1.48) per 10 ng/m3
18-65 years p=0.0046 (0.00036)
18-65 years P=0.00220 (0.000658)
1.003 (1-1.006) per 10 ng/m3
7-14 years OR = 1.33 (1.11-1.58) per 15 ng/m3
Asthmatics,
9-11 years
Ozone-related Health Impacts
Premature Mortality
Premature
mortality—short-term
Smith et al. (2009)181
Zanobetti and Schwartz
(2008)182
All ages
P = 0.00032 (0.00008)
P = 0.00051 (0.00012)
Hospital Admissions
Respiratory
Asthma-related
emergency
department visits
Pooled estimate:
Katsouyarwi et al.
(2009)183
Pooled estimate:
Glad et al. (2012)184
Ito et al. (2007)185
Mar and Koenig (2009)186
Peel et al. (2005)187
Sarnat et al. (2013)188
Wilson et al. (2005)189
> 65 years
0-99 years
P = 0.00064 (0.00040) penalized splines
P = 0.00306 (0.00117)
P = 0.00521 (0.00091)
P = 0.01044 (0.00436) (0-17 yr olds)
P = 0.00770 (0.00284) (18-99 yr olds)
P = 0.00087 (0.00053)
P = 0.00111 (0.00028)
RR = 1.022 (0.996 - 1.049) per 25
Other Health Endpoints
Asthma exacerbation
School loss days
Acute respiratory
symptoms (MRAD)
Pooled estimate:b
Mortimer et al. (2002)190
Schildcrout et al.
(2006)191
Pooled estimate:
Chen et al. (2000)192
Gilliland et al. (2001)193
Ostro and Rothschild
(1989)194
6-18 years
5-17 years
18-65
years
P = 0.00929 (0.00387)
P = 0.00222 (0.00282)
P = 0.015763 (0.004985)
P = 0.007824 (0.004445)
P = 0.002596 (0.000776)
Notes:
a For PM, studies highlighted in red represent updates incorporated since the ozone NAAQS RIA (U.S. EPA, 2008).
These updates were introduced in the PM NAAQS RIA (U.S. EPA, 2012). For ozone, studies highlighted in red
represent updates incorporated since the 2008 ozone NAAQS RIA (U.S. EPA, 2008).

-------
b The original study populations were 8 to 13 years for the Ostro et al. (2001) study and 7 to 12 years for the Mar et
al. (2004) study. Based on advice from the SAB-HES, we extended the applied population to 6-18 years, reflecting
the common biological basis for the effect in children in the broader age group. See: U.S. EPA-SAB (2004) and
NRC (2002).
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.195'196
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 8A-9 summarizes the sources of baseline incidence rates and provides average
incidence rates for the endpoints included in the analysis. Table 8A-10 presents the asthma
prevalence rates used in this 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.197

-------
Table 8A-9 Baseline Incidence Rates and Population Prevalence Rates for Use in Impact Functions, General
Population
Endpoint
Parameter

Rates
Value
Source
Mortality
Daily or annual mortality
Age-, cause-, and
CDC WONDER (2004-2006)

rate projected to 2025 a
county-specific rate
U.S. Census bureau, 2000
Hospitalizations
Daily hospitalization rate
Age-, region-, state-
2007 HCUP data files b


, county- and cause-



specific rate

E R Visits
Daily ER visit rate for asthma
Age-, region-, state-
2007 HCUP data files b

and cardiovascular events
, county- and cause-



specific rate

Nonfatal Myocardial
Daily nonfatal myocardial
Age-, region-, state-
2007 HCUP data filesb adjusted by
Infarction (heart
infarction incidence rate per
, and county-
0.93 for probability of surviving
attacks)
person, 18+
specific rate
after 28 days (Rosamond et al.,



1999)
Asthma Exacerbations0
Incidence among asthmatic

Ostro et al. (2001)

African-American children



daily wheeze
0.173


daily cough
0.145


daily shortness of breath
0.074

Acute Bronchitis
Annual bronchitis incidence
0.043
American Lung Association (2002,

rate, children

Table ll)198
Lower Respiratory
Daily lower respiratory
0.0012
Schwartz et al. (1994, Table 2)
Symptoms
symptom incidence among



children d


Upper Respiratory
Daily upper respiratory
0.3419
Pope et al. (1991, Table 2)
Symptoms
symptom incidence among



asthmatic children


Work Loss Days
Daily WLD incidence rate per

1996 HIS (Adams, Hendershot, and

person (18-65)

Marano, 1999, Table 41)199; U.S.

Aged 18-24
0.00540
Census Bureau (2000)2°°

Aged 25-44
0.00678


Aged 45-64
0.00492

School Loss Days
Rate per person per year,
9.9
National Center for Education

assuming 180 school days

Statistics (1996) and 1996 HIS

per year

(Adams et al., 1999, Table 47);
Minor Restricted-
Daily MRAD incidence rate
0.02137
Ostro and Rothschild (1989,
Activity Days
per person

p. 243)
Notes:
a Mortality rates are only available at 5-year increments.
b Healthcare Cost and Utilization Program (HCUP) database contains individual level, state and regional-level
hospital and emergency department discharges for a variety of International Classification of Diseases (ICD) codes
(AHRQ, 2007). 201
0 The incidence of exacerbated asthma was quantified among children of all races, using the baseline incidence rate
reported in Ostro et al. (2001).
d Lower respiratory symptoms are defined as two or more of the following: cough, chest pain, phlegm, and wheeze.

-------
Table 8A-10 Asthma Prevalence Rates Used for this Analysis


Asthma Prevalence Rates
Population Group
Value
Source
All Ages
0.0780
American Lung Association (2010, Table 7)
< 18
0.0941

5-17
0.1070

18-44
0.0719

45-64
0.0745

65+
0.0716

African American, 5-17
0.1776
American Lung Association (2010, Table 9)
African American, <18
0.1553
American Lung Association a
Note:
" Calculated by ALA for U.S. EPA, based on NHIS data (CDC, 2008).202,203
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).204 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, 19 8 7).205:206
We provide unit values for health endpoints (along with information on the distribution of the
unit value) in Table 8A-11. All values are in constant year 2013 dollars, adjusted for growth in
real income out to 2024 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 2012 PM
NAAQS RIA.207 For details on valuation estimates for ozone-related endpoints, see the 2015
Ozone NAAQSRIA.208

-------
Table 8A-11 Unit Values for Economic Valuation of Health Endpoints (2011$)a
Health Endpoint
Central Estimate of
Value Per Statistical
Incidence


2000
Income
Level
2024
Income
Level
Derivation of Distributions of Estimates
Premature Mortality
(Value of a Statistical
Life)
$8,300,000
$10,000,000
EPA currently recommends a central VSL of $6.3m (2000$) based on
a Weibull distribution fitted to 26 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 (U.S. EPA, 2010).209
Nonfatal Myocardial
Infarction (heart
attack)
3% discount rate
Age 0-24
Age 25-44
Age 45-54
Age 55-65
Age 66 and over
7% discount rate
Age 0-24
Age 25-44
Age 45-54
Age 55-65
Age 66 and over


No distributional information available. Age-specific cost-of-illness
values reflect lost earnings and direct medical costs over a 5-year
period following a nonfatal MI. Lost earnings estimates are based on
Cropper and Krupnick (1990).210 Direct medical costs are based on
simple average of estimates from Russell et al. (1998)211 and Wittels et
al. (1990).212
Lost earnings:
Cropper and Krupnick (1990). Present discounted value of 5 years of
lost earnings:
age of onset: at 3% at 7%
25-44 $8,774 $7,855
45-54 $12,932 $11,578
55-65 $74,746 $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)
$100,000
$110,000
$120,000
$210,000
$100,000
$100,000
$110,000
$120,000
$210,000
$100,000
$100,000
$110,000
$120,000
$190,000
$100,000
$100,000
$110,000
$120,000
$190,000
$100,000
Hospital Admissions



Chronic Lung
Disease (18-64)
$22,000
$22,000
No distributional information available. 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 chronic lung illnesses)
reported in Agency for Healthcare Research and Quality (2007)
(www.ahrq.gov).213
Asthma Admissions
(0-64)
$16,000
$16,000
No distributional information available. 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 (2007)
(www.ahrq.gov).
All Cardiovascular
Age 18-64
Age 65-99
$44,000
$42,000
$44,000
$42,000
No distributional information available. 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
(2007) (www.ahrq.gov).

-------
All respiratory (ages
65+)
$37,000
$37,000
No distributions available. The COI point 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 respiratory category illnesses) reported in
Agency for Healthcare Research and Quality, 2007 (www.ahrq.gov).
Emergency
Department Visits for
Asthma
$440
$440
No distributional information available. Simple average of two unit
COI values (2000$):
(1)	$310, from Smith et al. (1997)214 and
(2)	$260, from Stanford et al. (1999).215
Respiratory Ailments Not Requiring Hospitalization
Upper Respiratory
Symptoms (URS)
$35
$32
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
estimates of WTP (IEc, 1994) to avoid each symptom in the cluster
and assuming additivity of WTPs. In the absence of information
surrounding the frequency with which each of the seven types of URS
occurs within the URS symptom complex, we assumed a uniform
distribution between $9.2 and $43 (2000$).
Lower Respiratory
Symptoms (LRS)
$22
$21
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 (IEc, 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. In the
absence of information surrounding the frequency with which each of
the 11 types of LRS occurs within the LRS symptom complex, we
assumed a uniform distribution between $6.9 and $25 (2000$).
Asthma
Exacerbations
$56
$60
Asthma exacerbations are valued at $45 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).216 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 exacerbation is assumed to be equivalent to a day in which
asthma is moderate or worse as reported in the Rowe and Chestnut
(1986) study. The value is assumed to have a uniform distribution
between $16 and $71 (2000$).
Acute Bronchitis
$460
$500
Assumes a 6-day episode, with the distribution of the daily value
specified as uniform with the low and high values based on those
recommended for related respiratory symptoms in Neumann et al.
(1994). The low daily estimate of $10 is the sum of the mid-range
values recommended by IEc (1994) for two symptoms believed to be
associated with acute bronchitis: coughing and chest tightness. The
high daily estimate was taken to be twice the value of a minor
respiratory restricted-activity day, or $110 (2000$).
Work Loss Days
(WLDs)
Variable
(U.S.
median =
$150)
Variable
(U.S.
median =
$150)
No distribution available. Point estimate is based on county-specific
median annual wages divided by 52 and then by 5—to get median
daily wage. U.S. Year 2000 Census, compiled by Geolytics, Inc.
(Geolytics, 2002)217

-------
Minor Restricted
Activity Days
(MRADs)
$64
$68
Median WTP estimate to avoid one MRAD from Tolley et al.
(1986).218 Distribution is assumed to be triangular with a minimum of
$22 and a maximum of $83, with a most likely value of $52 (2000$).
Range is based on assumption that value should exceed WTP for a
single mild symptom (the highest estimate for a single symptom—for
eye irritation—is $ 16) and be less than that for a WLD. The triangular
distribution acknowledges that the actual value is likely to be closer to
the point estimate than either extreme.
School Loss Days
$98
$98
No distribution available. Based on (1) the probability that, if a school
child stays home from school, a parent will have to stay home from
work to care for the child, and (2) the value of the parent's lost
productivity.
Notes:
a All estimates are rounded to two significant digits. Unrounded estimates in 2000$ are available in the Appendix J
of the BenMAP user manual (U.S. EPA, 2015).219 Income growth projections are only currently available in
BenMAP through 2024, so the 2040 estimates use income growth through 2024 and are therefore likely
underestimates. Currently, BenMAP does not have an inflation adjustment to 2013$. We ran BenMAP for a
currency year of 2010$ and then adjusted the resulting benefit-per-ton estimates to 2013$ using the Consumer Price
Index (CPI-U, all items). This approach slightly underestimates the inflation for medical index and wage index
between 2010 and 2013, which affects COI estimates and wage-based estimates.
8A. 1.5 Processing Air Quality Modeling Data for Health Impacts Analysis
In the Appendix to Chapter 6, we summarized the methods for and results of estimating
air quality for the standards. 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 standards. 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.
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.RRR'SSS
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).
111111 The ozone season for this analysis is defined as the 5-month period from May to September.
sss 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).

-------
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
TTT TJIJU
maximum. '
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 2005 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 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, 2001).220 The guidance recommends that model predictions
be used in a relative sense to estimate changes expected to occur in each major PM2.5 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
standards.
Table 8A-12 provides those ozone and PM2.5 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.
TTT The 12-km grid squares contain the population data used in the health benefits analysis model, BenMAP.
1X111 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://www3.epa.gov/air/benmap.

-------
Table 8A-12 Summary of CMAQ-Derived Population-Weighted Ozone and PM2.5 Air Quality Metrics for
Health Benefits Endpoints Associated with the Standards

2040
Statistic51
Baseline
Changeb
Ozone Metric: National Population-Weightec
Average (ppb)°
Daily Maximum 8-Hour Average
Concentration
41.67
-0.21
PM2.5 Metric: National Population-Weighted Average (|ig/m3)
Annual Average Concentration
7.32
-0.01
Notes:
" 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). Note that the national, population-weighted
PM2.5 and ozone air quality metrics presented in this chapter represent an average for the entire,
gridded U.S. CMAQ domain. These are different than the population-weighted PM2 5 and ozone
design value metrics presented in Chapter 7, which represent the average for areas with a current
air quality monitor.
h The change is defined as the control-case value minus the base-case value; a negative value
therefore indicates a reduction and a positive value an increase.
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
different than the final emission inventories estimated for the standards. Please refer to Chapter
6 for more information about the inventories used in the air quality modeling that supports the
health impacts analysis.
8 A. 1.6 Methods for Describing Uncertainty
In any complex analysis using estimated parameters and inputs from numerous models,
there are likely to be many sources of uncertainty and this analysis is no exception. As outlined
both in this and preceding chapters, many inputs were used to derive the estimate of benefits for
the standards, including emission inventories, air quality models (with their associated
parameters and inputs), epidemiological health effect estimates, estimates of values (both from
WTP and COI studies), population estimates, income estimates, and estimates of the future state
of the world {i.e., regulations, technology, and human behavior). Each of these inputs may be
uncertain and, depending on its role in the benefits analysis, may have a disproportionately large
impact on estimates of total benefits. For example, emissions estimates are used in the first stage
of the analysis. As such, any uncertainty in emissions estimates will be propagated through the
entire analysis. When compounded with uncertainty in later stages, small uncertainties in
emission levels can lead to large impacts on total benefits.
The National Research Council (NRC) (2002, 2008)221'222 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 uncertainty. In general, the NRC concluded that EPA's general methodology for
calculating the benefits of reducing air pollution is reasonable and informative in spite of

-------
inherent uncertainties. Since the publication of these reports, EPA's Office of Air and Radiation
(OAR) continues to make progress toward the goal of characterizing the aggregate impact of
uncertainty in key modeling elements on both health incidence and benefits estimates in two key
ways: Monte Carlo analysis and expert-derived concentration-response functions. In this
analysis, we use both of these two methods to assess uncertainty quantitatively, as well as
provide a qualitative assessment for those aspects that we are unable to address quantitatively.
First, we used Monte Carlo methods for characterizing random sampling error associated
with the concentration response functions from epidemiological studies and random effects
modeling to characterize both sampling error and variability across the 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. The reported standard errors in the
epidemiological studies determined the distributions for individual effect estimates.
In benefit analyses of air pollution regulations conducted to date, the estimated impact of
reductions in premature mortality has accounted for 85 to 95 percent of total monetized benefits.
Therefore, it is particularly important to attempt 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. 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
used in the RIA for 2012 PM NAAQS, presenting two empirical estimates of premature deaths
avoided. This characterization, including confidence intervals, omit the contribution 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.
Some key sources of uncertainty in each stage of both the PM and ozone health impact
assessment are the following:
•	gaps in scientific data and inquiry;
•	variability in estimated relationships, such as epidemiological effect estimates,
introduced through differences in study design and statistical modeling;
•	errors in measurement and projection for variables such as population growth rates;

-------
•	errors due to misspecification of model structures, including the use of surrogate
variables, such as using PMio when PM2.5 is not available, excluded variables, and
simplification of complex functions; and
•	biases due to omissions or other research limitations.
In Table 8A-13 we summarize some of the key uncertainties in the benefits analysis.
Table 8A-13 Primary Sources of Uncertainty in the Benefits Analysis
1.	Uncertainties Associated with Impact Functions	
-	The value of the ozone or PM effect estimate in each impact function.
-	Application of a single impact function to pollutant changes and populations in all locations.
-	Similarity of future-year impact functions to current impact functions.
-	Correct functional form of each impact function.
-	Extrapolation of effect estimates beyond the range of ozone or PM concentrations observed in the source
epidemiological study.
-	Application of impact functions only to those subpopulations matching the original study population.
2.	Uncertainties Associated with CMAQ-Modeled Ozone and PM Concentrations	
-	Responsiveness of the models to changes in precursor emissions from the control policy.
-	Projections of future levels of precursor emissions, especially ammonia and crustal materials.
-	Lack of ozone and PM2.5 monitors in all rural areas requires extrapolation of observed ozone data from
urban to rural areas.	
3.	Uncertainties Associated with PM Mortality Risk	
-	Limited scientific literature supporting a direct biological mechanism for observed epidemiological
evidence.
-	Direct causal agents within the complex mixture of PM have not been identified.
-	The extent to which adverse health effects are associated with low-level exposures that occur many times
in the year versus peak exposures.
-	The extent to which effects reported in the long-term exposure studies are associated with historically
higher levels of PM rather than the levels occurring during the period of study.
-	Reliability of the PM2 5 monitoring data in reflecting actual PM2.5 exposures.	
4.	Uncertainties Associated with Possible Lagged Effects	
-	The portion of the PM-related long-term exposure mortality effects associated with changes in annual PM
levels that would occur in a single year is uncertain as well as the portion that might occur in subsequent
years.	
5.	Uncertainties Associated with Baseline Incidence Rates	
-	Some baseline incidence rates are not location specific (e.g., those taken from studies) and therefore may
not accurately represent the actual location-specific rates.
-	Current baseline incidence rates may not approximate well baseline incidence rates in 2040.
-	Projected population and demographics may not represent well future-year population and demographics.
6.	Uncertainties Associated with Economic Valuation	
-	Unit dollar values associated with health and welfare endpoints are only estimates of mean WTP and
therefore have uncertainty surrounding them.
-	Mean WTP (in constant dollars) for each type of risk reduction may differ from current estimates because
of differences in income or other factors.	
7.	Uncertainties Associated with Aggregation of Monetized Benefits	
-	Health and welfare benefits estimates are limited to the available impact functions. Thus, unqualified or
un-monetized benefits are not included.

-------
References
1	Environmental Protection Agency and Department of Transportation, "Light-Duty Vehicle Greenhouse Gas
Emissions Standards and Corporate Average Fuel Economy Standards; Final Rule," Federal Register 75(88) (May 7,
2010), especially Sections III.H.l (pp. 25510-25513) and IV.G.6 (pp. 25651-25657); Environmental Protection
Agency and Department of Transportation, "2017 and Later Model Year Light-Duty Vehicle Greenhouse Gas
Emissions and Corporate Average Fuel Economy Standards; Final Rule," Federal Register 77(199) (October 15,
2012), especially Sections III.H.l (pp. 62913-62919) and IV.G.5.a (pp. 63102-63104).
2	State of Massachusetts v. EPA, 549 U.S. at 533.
3	Klemick, Heather, Elizabeth Kopits, Keith Sargent, and Ann Wolverton (2015). "Heavy-Duty Trucking and the
Energy Efficiency Paradox: Evidence form Focus Groups and Interviews." Transportation Research Pari A 77: 154-
166. Docket EPA-HQ-OAR-2014-0827; Roeth, Mike, Dave Kircher, Joel Smith, and Rob Swim (2013). "Barriers to
the Increased Adoption of Fuel Efficiency Technologies in the North American On-Road Freight Sector." NACFE
report for the International Council on Clean Transportation, Docket EPA-HQ-OAR-2014-0827-0084; Aarnink,
Sanne, Jasper Faber, and Eelco den Boer (2012). "Market Barriers to Increased Efficiency in the European On-road
Freight Sector." CE Delft report for the International Council on Clean Transportation, Docket EPA-HQ-OAR-
2014-0827-0076.
4	Committee to Assess Fuel Economy Technologies for Medium- and Heavy-Duty Vehicles; National Research
Council; Transportation Research Board (2010). "Technologies and Approaches to Reducing the Fuel Consumption
of Medium- and Heavy-Duty Vehicles," (hereafter, "NAS 2010"). Washington, D.C. The National Academies
Press. Available electronically from the National Academies Press Website at
http://www.nap.edu/catalog.php?record_id=12845 (accessed September 10, 2010), Docket EPA-HQ-OAR-2014-
0827-0122.
5	Vernon, David and Alan Meier (2012). "Identification and quantification of principal-agent problems affecting
energy efficiency investments and use decisions in the trucking industry." Energy Policy, 49(C), pp. 266-273,
Docket EPA-HQ-OAR-2014-0827-0090.
6	Blumstein, Carl and Margaret Taylor (2013). "Rethinking the Energy-Efficiency Gap: Producers, Intermediaries,
and Innovation," Energy Institute at Haas Working Paper 243, University of California at Berkeley, Docket EPA-
HQ-OAR-2014-0827-0075; Tirole, Jean (1998). The Theory of Industrial Organization. Cambridge, MA: MIT
Press, pp.400, 402, Docket EPA-HQ-OAR-2014-0827-0089.
7	American Transportation Research Institute, An Analysis of the Operational Costs of Trucking, September 2013),
Docket EPA-HQ-OAR-2014-0827-0512.
8	Transport Canada, Operating Cost of Trucks, 2005. See http://www.tc.gc.ca/eng/policy/report-acg-
operatingcost2005-2005-e-2-1727.htm, accessed on July 16, 2010, Docket EPA-HQ-OAR-2014-0827-0070.
9	Winebrake, J.J., Green, E.H., Comer, B., Corbett, J.J., Froman, S., 2012. Estimating the direct rebound effect for
on-road freight transportation. Energy Policy 48, 252-259.
10	Greene, D.L., Kahn, J.R., Gibson, R.C., 1999, "Fuel economy rebound effect for U.S. household vehicles," The
Energy Journal, 20.
11	For a discussion of the wide range of definitions found in the literature, see Appendix D: Discrepancy in Rebound
Effect Definitions, in EERA (2014), "Research to Inform Analysis of the Heavy-Duty vehicle Rebound Effect," and
Excerpts of Draft Final Report of Phase 1 under EPA contract EP-C-13-025. (Docket ID: EPA-HQ-OAR-2014-
0827). See also Greening, L.A., Greene, D.L., Difiglio, C., 2000, "Energy efficiency and consumption — the
rebound effect — a survey," Energy Policy, 28, 389-401.
12	Committee to Assess Fuel Economy Technologies for Medium- and Heavy-Duty Vehicles; National Research
Council; Transportation Research Board (2010). "Technologies and Approaches to Reducing the Fuel Consumption
of Medium- and Heavy-Duty Vehicles," Washington, D.C. The National Academies Press. Available electronically
from the National Academies Press Website at http://www.nap.edu/catalog.php?record_id= 12845 (last accessed
September 10, 2010).
13	American Transportation Research Institute, An Analysis of the Operational Costs of Trucking, September 2013.
14	Transport Canada, Operating Cost of Trucks, 2005. See http://www.tc.gc.ca/eng/policv/report-acg-
operatingcost2005-2005-e-2-1727.htm. accessed on July 16, 2010.

-------
15	American Transportation Research Institute (ATRI). "An Analysis of the Operational Costs of Trucking: 2015
Update." Page 17. September 2015.
16	These factors are discussed more fully in a report to EPA from EERA, which illustrates in a series of diagrams the
complex system of decisions and decision-makers that could influence the magnitude and timing of the rebound
effect. See Sections 2.2.2, 2.2.3, 2.2.4, and 2.3 in EERA (2014), "Research to Inform Analysis of the Heavy-Duty
Vehicle Rebound Effect," Excerpts of Draft Final Report of Phase 1 under EPA contract EP-C-13-025.
17	See Chapter 8, Section 3.2 of the "Draft Regulatory Impact Analysis" for the proposal, accessible at
http://www3.epa.gov/otaq/climate/regs-heavy-duty.htm (Last Accessed: March 11, 2016).
18	See Chapter 8, Section 3.2 of the "Draft Regulatory Impact Analysis" for the proposal, accessible at
http://www3.epa.gov/otaq/climate/regs-heavy-duty.htm (Last Accessed: March 11, 2016).
19	EERA (2014), "Research to Inform Analysis of the Heavy-Duty vehicle Rebound Effect," Excerpts of Draft Final
Report of Phase 1 under EPA contract EP-C-13-025.
20	EERA (2015), "Working Paper on Fuel Price Elasticities for Heavy Duty Vehicles," Draft Final Report of Phase 2
under EPA contract EP-C-11-046.
21	Winebrake, J. J., et al., Fuel price elasticities in the U.S. combination trucking sector. Transportation Research
PartD: Transport and Environment, 2015. 38: p. 166-177.
22	Leard, B., et al., Fuel Costs, Economic Activity, and the Rebound Effect for Heavy-Duty Trucks. September 2015,
Resources for the Future: RF DP 15-43, Washington, DC. EPA-HQ-OAR-2014- 0827-1200-A1.
23	American Transportation Research Institute. An Analysis of the Operational Costs of Trucking: A 2015 Update.
Page 17. September 2015.
24	The new vocational vehicle cost assumed in the VMT rebound analysis is $100,000 based on the average of the
vocational vehicle prices listed in the ICF cost report, July 2010, pages 16-17. The combination tractor-trailer costs
are $100,000 based on the average price of day cabs and sleeper cabs and $25,000 for each of the three trailers
(assumed a trailer-to-tractor ratio of 3:1) as listed in the ICF cost report pages 3 and 16. EPA-HQ-OAR-2010-0162-
0044.
25	U.S. EPA. Motor Vehicle Emission Simulator (MOVES2014a). Last accessed on March 15, 2016 at
http ://www3. epa.gov/otaq/models/moves/.
26	U.S. EPA and NHTSA. Draft Regulatory Impact Analysis: Greenhouse Gas Emissions and Fuel Efficiency
Standards for Medium- and Heavy-Duty Engines and Vehicles - Phase 2. EPA-420-D-15-002. June 2015.
Technology package costs for Alternatives lb, 2, 3, and 4 from Tables 11-21 and 11-22 on pages 11-27 and 11-28.
27	80 FR July 13, 2015 at 40494-40495.
28	U.S. Energy Information Administration. Annual Energy Outlook 2015. Table 12: Petroleum Product Prices.
29	U.S. EPA and NHTSA. Draft Regulatory Impact Analysis: Greenhouse Gas Emissions and Fuel Efficiency
Standards for Medium- and Heavy-Duty Engines and Vehicles - Phase 2. EPA-420-D-15-002. June 2015.
Technology package effectiveness as discussed in Chapter 11.4, pages 11-73 through 11-85.
30	Memo from Energy and Environmental Research Associates, LLC Regarding HDV Rebound Effect, dated June 8,
2011.
31	Committee to Assess Fuel Economy Technologies for Medium- and Heavy-Duty Vehicles; National Research
Council; Transportation Research Board (2010). "Technologies and Approaches to Reducing the Fuel Consumption
of Medium- and Heavy-Duty Vehicles," Washington, D.C., The National Academies Press. Available electronically
from the National Academies Press Website at http://www.nap.edu/catalog.php?record_id=12845, page 152.
32	A baseline tractor price of a new day cab is $89,500 versus $113,000 for a new sleeper cab based on information
gathered by ICF in the "Investigation of Costs for Strategies to Reduce Greenhouse Gas Emissions for Heavy-Duty
On-Road Vehicles," July 2010. Page 3. Docket Identification Number EPA-HQ-OAR-2010-0162-0044.
33	See NAS 2010 Report, pp. 150-151, EPA-HQ-OAR-2014-0827-0276.
34	See NAS 2010 Report, page 151, Docket EPA-HQ-OAR-2014-0827-0276.
35	Source: Table 2.14 (Errata). Lifetimes, radiative efficiencies and direct (except for CH4) GWPs relative to
CO2. IPCC Fourth Assessment Report "Climate Change 2007: Working Group I: The Physical Science Basis."
36	West JJ, Fiore AM, Horowitz LW, Mauzerall DL (2006) Global health benefits of mitigating ozone pollution with
methane emission controls. Proc Natl Acad Sci USA 103 (ll):3988-3993. doi:10.1073/pnas.0600201103
37	Anenberg, S., Schwartz, J., Shindell, D., Amann, M., Faluvegi, G., Klimont, Z., Vignati, E., (2012) Global air
quality and health co-benefits of mitigating near-term climate change through methane and black carbon emission
controls. Environ Health Perspect 120 (6):831. doi: 10.1289/ehp. 1104301.
38	Shindell, D., Kuylenstierna, JCI., Vignati, E., vanDingenen, R., Amann, M., Klimont, Z., Fowler D (2012)
Simultaneously Mitigating Near-Term Climate Change and Improving Human Health and Food Security. Science
335 (6065): 183-189. doi:10.1126/science,1210026.

-------
39	Sarofim, M.C., Waldhoff, S.T., Anenberg, S.C., 2015. "Valuing the Ozone-Related Health Benefits of Methane
Emission Controls," Environ. Resource Econ. DOI 10.1007/sl0640-015-9937-6.
40	Fann, N., Baker, K.R., and Fulcher, C.M. (2012). Characterizing the PM2.5-related health benefits of emission
reductions for 17 industrial, area and mobile emission sectors across the U.S., Environment International, 49, 241-
151, published online September 28, 2012.
41	U.S. Environmental Protection Agency. (2014). Control of Air Pollution from Motor Vehicles: Tier 3 Motor
Vehicle Emission and Fuel Standards Final Rule: Regulatory Impact Analysis, Assessment and Standards Division,
Office of Transportation and Air Quality, EPA-420-R-14-005, March 2014. Available on the internet:
http://www3.epa.gov/otaq/documents/tier3/420rl4005.pdf.
42	U.S. Environmental Protection Agency. (2012). Regulatory Impact Analysis for the Final Revisions to the
National Ambient Air Quality Standards for Particulate Matter, Health and Environmental Impacts Division, Office
of Air Quality Planning and Standards, EPA-452-R-12-005, December 2012. Available on the internet:
http://www3.epa.gov/ttnecasl/regdata/RIAs/finalria.pdf.
43	U.S. Environmental Protection Agency (U.S. EPA). (2012). Regulatory Impact Analysis: Final Rulemaking for
2017-2025 Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy
Standards, Assessment and Standards Division, Office of Transportation and Air Quality, EPA-420-R-12-016,
August 2012. Available on the Internet at: http://www3.epa.gov/otaq/climate/documents/420rl2016.pdf.
44	U.S. Environmental Protection Agency (U.S. EPA). (2012/ Regulatory Impact Analysis: Final Rulemaking for
2017-2025 Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy
Standards, Assessment and Standards Division, Office of Transportation and Air Quality, EPA-420-R-12-016,
August 2012. Available on the Internet at: http://www3.epa.gov/otaq/climate/documents/420rl2016.pdf.
45	U.S. Environmental Protection Agency (U.S. EPA). (2013). Regulatory Impact Analysis for the Reconsideration
of the Existing Stationary Compression Ignition (CI) Engines NESHAP, Office of Air Quality Planning and
Standards, Research Triangle Park, NC. January. EPA-452/R-13-001. Available at
http://www3.epa.gov/ttnecasl/regdata/RIAs/RICE_NESHAPreconsideration_Compression_Ignition_Engines_RIA_
final2013_EPA.pdf.
46	U.S. Environmental Protection Agency (U.S. EPA). (2013). Regulatory Impact Analysis for Reconsideration of
Existing Stationary Spark Ignition (SI) RICE NESHAP, Office of Air Quality Planning and Standards, Research
Triangle Park, NC. January. EPA-452/R-13-002. Available at
http://www3.epa.gov/ttnecasl/regdata/RIAs/NESHAP_RICE_Spark_Ignition_RIA_finalreconsideration2013_EPA.
pdf.
47	U.S. Environmental Protection Agency (U.S. EPA). (2015). Regulatory Impact Analysis for Residential Wood
Heaters NSPSRevision. Office of Air Quality Planning and Standards, Research Triangle Park, NC. February.
EPA-452/R-15-001. Available at http://www2.epa.gOv/sites/production/files/2015-02/documents/20150204-
residential-wood-heaters-ria.pdf.
48	U.S. Environmental Protection Agency. (2012). Regulatory Impact Analysis for the Final Revisions to the
National Ambient Air Quality Standards for Particulate Matter, Health and Environmental Impacts Division, Office
of Air Quality Planning and Standards, EPA-452-R-12-005, December 2012. Available on the internet:
http://www3.epa.gov/ttnecasl/regdata/RIAs/finalria.pdf.
49	Fann, N., Baker, K.R., and Fulcher, C.M. (2012). Characterizing the PM2.5-related health benefits of emission
reductions for 17 industrial, area and mobile emission sectors across the U.S., Environment International, 49, 241-
151, Published online September 28, 2012.
50	Fann, N., Baker, K.R., and Fulcher, C.M. (2012). Characterizing the PM2.5-related health benefits of emission
reductions for 17 industrial, area and mobile emission sectors across the U.S., Environment International, 49, 241-
151, Published online September 28, 2012.
51	U.S. Environmental Protection Agency (U.S. EPA). (2009). Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available at .
52	U.S. Environmental Protection Agency. (2012). Regulatory Impact Analysis for the Final Revisions to the
National Ambient Air Quality Standards for Particulate Matter, Health and Environmental Impacts Division, Office
of Air Quality Planning and Standards, EPA-452-R-12-005, December 2012. Available on the internet:
http://www3.epa.gov/ttnecasl/regdata/RIAs/finalria.pdf.
53	Science Advisory Board. 2001. NATA - Evaluating the National-Scale Air Toxics Assessment for 1996 - an
SAB Advisory, http://www3.epa.gov/ttn/atw/sab/sabrev.html.

-------
54	U.S. Environmental Protection Agency (U.S. EPA). 2011. The Benefits and Costs of the Clean Air Act from 1990
to 2020. Office of Air and Radiation, Washington, DC. March. Available on the Internet at
.
56	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 21, 2010).
57	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 8, 2012).
58	Passenger vehicle fuel dispensing rate per EPA regulations, last viewed on August 4, 2010 at
http://www3. epa.gov/oms/regs/ld-hwy/evap/spitback. txt.
59	Survey data results on refueling times for light-duty trucks derived from "Evaluation of the Effectiveness of
TPMS in Proper Tire Pressure Maintenance," November 2012. Report No. DOT HS 811 681. http://www-
nrd.nhtsa.dot.gov/Pubs/811681 .pdf.
60	U.S. Department of Transportation, Valuation of Travel Guidance, July 9, 2014, at page 14.
61	Based on data from the CIA, combining various recent years, https://www.cia.gov/librarv/publications/the-world-
factbook/rankorder/2242rank.html.
62	See EIA Annual Energy Review, various editions. For data 2011-2013, and projected data: EIA Annual Energy
Outlook (AEO) 2014 (Reference Case). See Table 11, file "aeotab_ll.xls."
63	For historical data: EIA Annual Energy Review, various editions. For data 2011-2013, and projected data: EIA
Annual Energy Outlook (AEO) 2014 (Reference Case). See Table 11, file "aeotab_ll.xls."
64	Leiby, Paul N., Donald W. Jones, T. Randall Curlee, and Russell Lee, Oil Imports: An Assessment of Benefits and
Costs, ORNL-6851, Oak Ridge National Laboratory, November, 1997.
65	Peer Review Report Summary: Estimating the Energy Security Benefits of Reduced U.S. Oil Imports, ICF, Inc.,
September 2007.
66	Leiby, P., Factors Influencing Estimate of Energy Security Premium for Heavy-Duty Phase II Final Rule,
11/18/2014, Oak Ridge National Laboratory.
67	Brown, Stephen P.A. andHillard G. Huntington. 2013. Assessing the U.S. Oil Security Premium,
Energy Economics, vol. 38, pp 118-127.
68	Reassessing the Oil Security Premium. RFF Discussion Paper Series, (RFF DP 10-05). doi: RFF DP 10-05
69	Greene, D. L. 2010. Measuring energy security: Can the United States achieve oil independence? Energy Policy,
38(4), 1614-1621. doi:10.1016/j.enpol.2009.01.041.
70	Reassessing the Oil Security Premium. RFF Discussion Paper Series, (RFF DP 10-05). doi:RFF DP 10-05
71	Ledyard, John O. "Market Failure." The New Palgrave Dictionary of Economics. Second Edition. Eds. Steven N.
Durlauf and Lawrence E. Blume. Palgrave Macmillan, 2008.
72	Council on Foreign Relations, "Automobile Fuel Economy Standards in a Lower-Oil-Price World," Sivarm &
Levi, November 2015.
73	Transitions to Alternative Vehicles and Fuels," Committee on Transitions to Alternative Vehicles and Fuels,
National Research Council, 2013.
74	Gately, Dermot 2004. "OPEC's Incentives for Faster Output Growth," The Energy Journal,25(2):75-96; Gately,
Dermot 2007. "What Oil Export Levels Should We Expect From OPEC?," The Energy Journal, 28(2): 151-173.
75
Fatih Birol, Executive Director of the International Energy Agency, warns that prolonged lower oil prices would
trigger energy-security concerns by increasing reliance on a small number of low-cost producers "or risk a sharp
rebound in price if investment falls short." "It would be a grave mistake to index our attention to energy security to
changes in the oil price," Birol said. "Now is not the time to relax. Quite the opposite: a period of low oil prices is
the moment to reinforce our capacity to deal with future energy security threats." Hussain, Y. (2015). "Grave
mistake" to be complacent on energy security, International Energy Agency warns. Financial Post, (November 10).
Retrieved from http://business.financialpost.com/news/energy/grave-mistake-to-be-complacent-on-energy-security-
international-energy-agency-warns.
76	Batovic, A. (2015). Low oil prices fuel political and economic instability. Global Risk Insights, 18-19. Retrieved
from http://globalriskinsights.eom/2015/09/low-oil-prices-fuel-political-and-economic-instability/.
77
Monaldi, F. (2015). The Impact of the Decline in Oil Prices on the Economics, Politics and Oil Industry of
Venezuela. Columbia Center on Global Energy Policy Discussion Papers, (September). Retrieved from

-------
http://energypolicy.columbia.edu/sites/default/files/energy/Impact of the Decline in Oil Prices on Venezuela,
September 2015.pdf.
no
Even, S., & Guzansky, Y. (2015). Falling oil prices and Saudi stability - Opinion. Jerusalem Post, (September
30). Retrieved from http://www.jpost.com/Opinion/Falling-oil-prices-and-Saudi-stability-419534.
79
International Monetary Fund (IMF). (2015). IMF Regional Economic Outlook - Middle East and Central Asia.
Regional Economic Outlook (Vol. 33). Tomkiw, L. (2015). Oil Rich Saudi Arabia Running Out Of Assets? IMF
Report Says It's Possible In Next 5 Years. International Business Times, October 21, 19-22. Retrieved from
http://www.ibtimes.com/oil-rich-saudi-arabia-running-out-assets-imf-report-says-its-possible-next-5-years-215017.
80	Bohi, Douglas R. And W. David Montgomery 1982. "Social Cost of Imported and U.S. Import Policy," Annual
Review of Energy, 7:37-60. Energy Modeling Forum, 1981. World Oil, EMF Report 6 (Stanford University Press:
Stanford 39 CA. https//emf.stanford.edu/publications/emf-6-world-oil.
81	Plummer, James L. (Ed.) 1982. Energy Vulnerability, "Basic Concepts, Assumptions and Numerical Results," pp.
13-36, (Cambridge MA: Ballinger Publishing Co.).
82	Bohi, Douglas R. And W. David Montgomery 1982 "Social Cost of Imported and U.S. Import Policy," Annual
Review of Energy, 7:37-60.
83	Hogan, William W., 1981. "Import Management and Oil Emergencies," Chapter 9 in Deese, 5 David and Joseph
Nye, eds. Energy and Security. Cambridge, MA: Ballinger Publishing Co.
84	Broadman, H. G. 1986. "The Social Cost of Imported Oil," Energy Policy 14(3):242-252. Broadman H.G. and
W. W. Hogan, 1988. "Is an Oil Import Tariff Justified? An American Debate: The Numbers Say 'Yes'," The Energy
Journal 9: 7-29.
85	Leiby, Paul N., Donald W. Jones, T. Randall Curlee, and Russell Lee, Oil Imports: An Assessment of Benefits
and Costs, ORNL-6851, Oak Ridge National Laboratory, November 1, 1997.
86	Parry, Ian W.H. and Joel Darmstadter 2004. "The Costs of U.S. Oil Dependency," Resources for the Future,
November 17, 2004 (also published as NCEP Technical Appendix Chapter 1: Enhancing Oil Security, the National
Commission on Energy Policy 2004 Ending the Energy Stalemate - A Bipartisan Strategy to Meet America's
Energy Challenges).
87	National Research Council, 2009. Hidden Costs of Energy: Unpriced Consequences of Energy Production and
Use. National Academy of Science, Washington, DC.
88	See, William Nordhaus, "Who's Afraid of a Big Bad Oil Shock?," available at
http://aida.econ.vale.edu/~nordhaus/homepage/Big Bad Oil Shock Meeting.pdf, and Olivier Blanchard and Jordi
Gali, "The macroeconomic Effects of Oil price Shocks: Why are the 2000s so different from the 1970s?," pp. 373-
421, in The International Dimensions of Monetary Policy. Jordi Gali and Mark Gertler, editors, University of
Chicago Press, February 2010, available at http://www.nber.org/chapters/c0517.pdf
89	In fact, "... energy-price changes have no effect on multifactor productivity and very little effect on labor
productivity." Page 19. He calculates the productivity effect of a doubling of oil prices as a decrease of 0.11 percent
for one year and 0.04 percent a year for ten years. Page 5. (The doubling reflects the historical experience of the
post-war shocks, as described in Table 7.1 in Blanchard and Gali, p. 380).
90	Blanchard and Gali, p. 414.
91	See, Oil price Drops on Oversupply, http://www.oil-price.net/en/articles/oil-price-drops-on-oversupplv.php.
10/6/2014.
92	Hamilton, J. D., 2012, Oil Prices, Exhaustible Resources, and Economic Growth. In Handbook of Energy and
Climate Change. Retrieved from http://econweb.ucsd.edu/~jhamilto/handbook_climate.pdf [EPA-HQ-OAR-2014-
0827-0577],
93	Ramey, V. A., & Vine, D. J. (2010). "Oil, Automobiles, and the U.S. Economy: How Much have Things Really
Changed?," National Bureau of Economic Research Working Papers, WP 16067(June). Retrieved from
http://www.nber.Org/papers/w 16067.pdf.
94	Baumeister, C.; Peersman, G.; VanRobays, I., 2010, "The Economic Consequences of Oil Shocks: Differences
across Countries and Time", Workshop and Conference on Inflation Challenges in the Era of Relative Price Shocks.
95	Kilian, L.; Vigfusson, R.J., 2014,; "The Role of Oil Price Shocks in Causing U.S. Recessions", Board of
Governors of the Federal Reserve System. International Finance Discussion Papers.
96	Cashin, P.,; Mohaddes, K.,; Raissi, Maziar, and Raissi, M.,e2014," The differential effects of oil demand and
supply shocks on the global economy". Energy Economics.
97	National Research Council, "Transitions to alternative vehicles and fuels," 2013.
98	http://www.whitehouse.gov/sites/default/files/omb/inforeg/eol2866/eol3563_0118201 l.pdf.

-------
99	U.S. Department of Labor, Bureau of Labor Statistics. "Automotive Industry; Employment, Earnings, and
Hours," http://www.bls.gov/iag/tgs/iagauto.htm, accessed 8/18/14.
100	Layard, P.R.G., and A. A. Walters (1978), Microeconomic Theory (McGraw-Hill, Inc.), Chapter 9 (Docket EPA-
HQ-OAR-2014-0827-0070).
101	Berman, E. and L. T. M. Bui (2001). "Environmental Regulation and Labor Demand: Evidence from the South
Coast Air Basin." Journal of Public Economics 79(2): 265-295 (Docket EPA-HQ-OAR-2014-0827-0074).
102	Ehrenberg, Ronald G., and Robert S. Smith (2000), Modern Labor Economics: Theory and Public Policy
(Addison Wesley Longman, Inc.), p. 108, Docket EPA-HQ-OAR-2014-0827-0077.
103	This discussion draws fromBerman, E. and L. T. M. Bui (2001). "Environmental Regulation and Labor Demand:
Evidence from the South Coast Air Basin." Journal of Public Economics 79(2): 265-295 (Docket EPA-HQ-0 AR-
2014-0827-0074), p. 293.
104	Arrow et al. (1996). "Benefit-Cost Analysis in Environmental, Health, and Safety Regulation: A Statement of
Principles." American Enterprise Institute, The Annapolis Center, and Resources for the Future, Docket EPA-HQ-
OAR-2014-0827-0073. See discussion on bottom of p. 6. In practice, distributional impacts on individual workers
can be important, as discussed later in this section.
105	Schmalensee, Richard, and Robert N. Stavins. "A Guide to Economic and Policy Analysis of EPA's Transport
Rule." White paper commissioned by Excelon Corporation, March 2011, Docket EPA-HQ-OAR-2014-0827-0071.
106	Klaiber, H. Allen, and V. Kerry Smith (2012). "Developing General Equilibrium Benefit Analyses for Social
Programs: An Introduction and Example." Journal of Benefit-Cost Analysis 3(2), Docket EPA-HQ-OAR-2014-
0827-0085.
107	Graff Zivin, J., and M. Neidell (2012). "The Impact of Pollution on Worker Productivity." American Economic
Review 102: 3652-3673), Docket EPA-HQ-OAR-2014-0827-0092.
108	Hamermesh (1993), Labor Demand (Princeton, NJ: Princeton University Press), Chapter 2 (Docket EPA-HQ-
OAR-2011-0135), Docket EPA-HQ-OAR-2014-0827-0082.
109	Ehrenberg, Ronald G., and Robert S. Smith (2000), Modern Labor Economics: Theory and Public Policy
(Addison Wesley Longman, Inc.), Chapter 4, Docket EPA-HQ-OAR-2014-0827-0077.
110	Berman, E. and L. T. M. Bui (2001). "Environmental Regulation and Labor Demand: Evidence from the South
Coast Air Basin." Journal of Public Economics 79(2): 265-295 (Docket EPA-HQ-OAR-2014-0827-0074).
111	Morgenstern, Richard D., William A. Pizer, and Jhih-Shyang Shih. "Jobs Versus the Environment: An Industry-
Level Perspective." Journal of Environmental Economics and Management 43 (2002): 412-436 (Docket EPA-HQ-
OAR-2014-0827-0088).
112	Gray, Wayne B., Ronald J. Shadbegian, Chunbei Wang, and Merve Meral. "Do EPA Regulations Affect Labor
Demand? Evidence from the Pulp and Paper Industry." Journal of Environmental Economics and Management: 68,
2014, 188-202, Docket EPA-HQ-OAR-2014-0827-0080.
113	Ferris, Ann, Ronald J. Shadbegian and Ann Wolverton. "The Effect of Environmental Regulation on Power
Sector Employment: Phase I of the Title IV S02 Trading Program." Journal of the Association of Environmental
and Resource Economists. (Forthcoming 2014), Docket EPA-HQ-OAR-2014-0827-0078.
114	Greenstone, M. (2002). "The Impacts of Environmental Regulations on Industrial Activity: Evidence from the
1970 and 1977 Clean Air Act Amendments and the Census of Manufactures," Journal of Political Economy 110(6):
1175-1219 (Docket EPA-HQ-OAR- 2014-0827-0081); Walker, Reed. (2011)."Environmental Regulation and Labor
Reallocation." American Economic Review: Papers and Proceedings 101(3): 442-447 (Docket EPA-HQ-0 AR-
2014-0827-0091).
115	List, J. A., D. L. Millimet, P. G. Fredriksson, and W. W. McHone (2003). "Effects of Environmental Regulations
on Manufacturing Plant Births: Evidence from a Propensity Score Matching Estimator." The Review of Economics
and Statistics 85(4): 944-952 (Docket EPA-HQ-OAR- 2014-0827-0087).
116	Berman, E. and L. T. M. Bui (2001). "Environmental Regulation and Labor Demand: Evidence from the South
Coast Air Basin." Journal of Public Economics 79(2): 265-295 (Docket EPA-HQ-0 AR-0827-0074).
117	http://www.bls.gov/emp/ep_data_emp_requirements.htm; see "HD Substitution Effect Employment
Impacts,"Docket EPA-HQ-OAR-2014-0827.
118	http://www.census.gov/manufacturing/asm/index.html; see "HD Substitution Effect Employment
Impacts,"Docket EPA-HQ-OAR-2014-0827.
119	http://www.bls.gov/emp/ep_data_emp_requirements.htm; see "HD Substitution Effect Employment
Impacts,"Docket EPA-HQ-OAR-2014-0827. This analysis used data for sectors 80 (Motor Vehicle Manufacturing),
81 (Motor Vehicle Body and Trailer Manufacturing), and 82 (Motor Vehicle Parts Manufacturing) from "Chain-
weighted (2009 dollars) real domestic employment requirements tables."

-------
120	American Transportation Research Institute, "An Analysis of the Operational Costs of Trucking: 2011 Update."
See http://www.atri-online.org/research/results/Op_Costs_201 l_Update_one_page_summary.pdf, Docket EPA-HQ-
OAR-2014-0827-0512.
121	Association of American Railroads, "All Inclusive Index and Rail Adjustment Factor." June 3, 2011. See
http://www.aar.Org/~/media/aar/RailCostIndexes/AAR-RCAF-2011-Q3.ashx, Docket EPA-HQ-OAR-2014-0827-
0065.
122	Krewski, D., M. Jerret, R.T. Burnett, R. Ma, E. Hughes, Y. Shi, et al. (2009). Extended follow-up and spatial
analysis of the American Cancer Society study linking particulate air pollution and mortality. HEI Research Report,
140, Health Effects Institute, Boston, MA.
123	Smith, R.L.; B. Xu and P. Switzer. 2009. Reassessing the relationship between ozone and short- term mortality in
U.S. urban communities. Inhalation Toxicology. 21:37-61.
124Lepeule J, LadenF, Dockery D, Schwartz J (2012). Chronic Exposure to Fine Particles and Mortality: An
Extended Follow-Up of the Harvard Six Cities Study from 1974 to 2009. Environ Health Perspect. Jul;120(7):965-
70.
125	Zanobetti, A; J. Schwartz. 2008. Mortality displacement in the association of ozone with mortality: an analysis of
48 cities in the United States. American Journal of Respiratory and Critical Care Medicine. 177:184-189.
126	U.S. Environmental Protection Agency. (2012). Regulatory Impact Analysis for the Final Revisions to the
National Ambient Air Quality Standards for Particulate Matter. Prepared by: Office of Air and Radiation, Office of
Air Quality Planning and Standards. Retrieved April 27, 2016 at http://www3.epa.gov/ttn/ecas/ria.html. EPA-
452/R-12-005
127	U.S. Environmental Protection Agency. (2015). Regulatory Impact Analysis of the Final Revisions to the
National Ambient Air Quality Standards for Ground-Level Ozone. Prepared by: Office of Air and Radiation, Office
of Air Quality Planning and Standards. Retrieved April 27, 2016 at http://www3.epa.gov/ttn/ecas/ria.html. EPA-
452/R-15-007.
128	U.S. Environmental Protection Agency. (2012). Regulatory Impact Analysis for the Final Revisions to the
National Ambient Air Quality Standards for Particulate Matter. Prepared by: Office of Air and Radiation, Office of
Air Quality Planning and Standards. Retrieved April 27, 2016 at http://www3.epa.gov/ttn/ecas/ria.html. EPA-
452/R-12-005.
129	U.S. EPA. 2009 Final Report: Integrated Science Assessment for Particulate Matter. U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009.
130	U.S. Environmental Protection Agency. (2015). Regulatory Impact Analysis of the Final Revisions to the
National Ambient Air Quality Standards for Ground-Level Ozone. Prepared by: Office of Air and Radiation, Office
of Air Quality Planning and Standards. Retrieved April 27, 2016 at http://www3.epa.gov/ttn/ecas/ria.html. EPA-
452/R-15-007.
131	U.S. EPA. 2013 Final Report: Integrated Science Assessment of Ozone and Related Photochemical Oxidants.
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-10/076F, 2013.
132	U.S. Environmental Protection Agency (2016). Integrated Risk Information System, https://www3.epa.gov/iris
133	Science Advisory Board. 2001. NATA - Evaluating the National-Scale Air Toxics Assessment for 1996 - an
SAB Advisory, http://www3.epa.gov/ttn/atw/sab/sabrev.html.
134U.S. Environmental Protection Agency (U.S. EPA). 2011. The Benefits and Costs of the Clean Air Act from 1990
to 2020. Office of Air and Radiation, Washington, DC. March. Available on the Internet at

-------
139	Fann N, Roman HR, Fulcher CF, Gentile M, Wesson K, Hubbell BJ, Levy JI. 2012. "Maximizing Health
Benefits and Minimizing Inequality: Incorporating Local Scale Data in the Design and Evaluation of Air Quality
Policies." Risk Analysis. 32(6): 908—922.
140	Tagaris E, Liao KJ, Delucia AJ, et al. 2009. "Potential impact of climate change on air-pollution related human
health effects." Environ. Sci. Technol. 43: 4979-4988.
141	Kunzli, N., R. Kaiser, S. Medina, M. Studnicka, O. Chanel, P. Filliger, M. Herry, F. Horak Jr., V. Puybonnieux-
Texier, P. Quenel, J. Schneider, R. Seethaler, J-C Vergnaud, and H. Sommer. 2000. "Public-Health Impact of
Outdoor and Traffic-Related Air Pollution: A European Assessment." The Lancet 356:795-801.
142	Levy JI, Baxter LK, Schwartz J. 2009. "Uncertainty and variability in health-related damages from coal-fired
power plants in the United States." Risk Anal. 29(7) 1000-1014.
143	Abt Associates, Inc. 2010. Environmental Benefits and Mapping Program (Version 4.0). Bethesda, MD. Prepared
for U.S. Environmental Protection Agency Office of Air Quality Planning and Standards. Research Triangle Park,
NC. Available on the Internet at .
144	Hubbell, B., A. Hallberg, D.R. McCubbin, and E. Post. 2005. "Health-Related Benefits of Attaining the 8-Hr
Ozone Standard." Environmental Health Perspectives 113:73-82. ICD, 1979.
145	Tagaris E, Liao KJ, Delucia AJ, et al. 2009. "Potential impact of climate change on air-pollution related human
health effects." Environ. Sci. Technol. 43: 4979-4988.
146	Fann N, Roman HR, Fulcher CF, Gentile M, Wesson K, Hubbell BJ, Levy JI. 2012. "Maximizing Health
Benefits and Minimizing Inequality: Incorporating Local Scale Data in the Design and Evaluation of Air Quality
Policies." Risk Analysis. 32(6): 908—922.
147	U.S. Environmental Protection Agency (U.S. EPA). 2011. Regulatory Impact Analysis for the Final Mercury and
Air Toxics Standards. EPA-452/R-11-011. December. Available on the Internet at
.
148	Woods & Poole Economics Inc. 2012. Population by Single Year of Age. CD-ROM. Woods & Poole Economics,
Inc. Washington, D.C.
149	Hollman, F.W., T.J. Mulder, and J.E. Kalian. 2000. Methodology and Assumptions for the Population
Projections of the United States: 1999 to 2100. Population Division Working Paper No. 38, Population Projections
Branch, Population Division, U.S. Census Bureau, Department of Commerce. January. Available on the Internet at
.
150	U.S. Bureau of Economic Analysis (U.S. BEA). 2004. BEA Economic Areas (EAs). U.S. Department of
Commerce. Available on the Internet at .
151	U.S. Environmental Protection Agency (U.S. EPA). 2011. The Benefits and Costs of the Clean Air Act 1990 to
2020: EPA Report to Congress. Office of Air and Radiation, Office of Policy, Washington, DC. March. Available
on the Internet at .
152	U.S. Environmental Protection Agency (U.S. EPA). 2013. Integrated Science Assessment for Ozone and Related
Photochemical Oxidants: Final. Research Triangle Park, NC: U.S. Environmental Protection Agency. (EPA
document number EPA/600/R-10/076F).
153	U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter
(Final Report). EPA-600-R-08-139F. National Center for Environmental Assessment—RTP Division. December.
Available on the Internet at .
154	U.S. Environmental Protection Agency (U.S. EPA). 2012. Provisional Assessment of Recent Studies on Health
Effects of Particulate Matter Exposure. EPA/600/R-12/056A. National Center for Environmental Assessment—RTP
Division. December.
155	Abt Associates, Inc. 2012. BenMAP User's Manual Appendices. Prepared for U.S. Environmental Protection
Agency Office of Air Quality Planning and Standards. Research Triangle Park, NC. September. Available on the
Internet at 
156	Krewski D, Jerrett M, Burnett RT, Ma R, Hughes E, Shi, Y, et al. 2009. Extended follow-up and spatial analysis
of the American Cancer Society study linking particulate air pollution and mortality. HEI Research Report, 140,
Health Effects Institute, Boston, MA.
157	Lepeule J, Laden F, Dockery D, Schwartz J. 2012. "Chronic Exposure to Fine Particles and Mortality: An
Extended Follow-Up of the Harvard Six Cities Study from 1974 to 2009." Environmental Health Perspectives 120
(7):965-70.
158	Woodruff, T.J., J. Grillo, and K.C. Schoendorf. 1997. "The Relationship Between Selected of postneonatal infant
mortality and particulate air pollution in the United States." Environmental Health Perspectives 105(6): 608-612.
159	Peters, A., D. W. Dockery, J. E. Muller and M. A. Mittleman. 2001. "Increased particulate air pollution and the
triggering of myocardial infarction." Circulation 103 (23): 2810-5.

-------
160	Pope, C. A., Ill, J. B. Muhlestein, et al. 2006. "Ischemic heart disease events triggered by short-term exposure to
fine particulate air pollution." Circulation 114(23): 2443-8.
161	Sullivan, J., L. Sheppard, et al. 2005. "Relation between short-term fine-particulate matter exposure and onset of
myocardial infarction." Epidemiology 16(1): 41-8.
162	Zanobetti A. and Schwartz, J. 2006. "Air pollution and emergency admissions in Boston, MA." Journal of
Epidemiology and Community Health 60(10): 890-5.
163	Kloog, I., B.A. Coull, A. Zanobetti, P. Koutrakis, J.D. Schwartz. 2012. "Acute and Chronic Effects of Particles
on Hospital Admissions in New England." PLoS ONE. Vol 7 (4): 1-8.
164	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.
165	Babin, S. M., H. S. Burkom, et al. 2007. "Pediatric Patient Asthma-related Emergency Department Visits and
Admissions in Washington, DC, from 2001-2004, and Associations with Air Quality, Socio-economic Status and
Age Group." Environmental Health 6: 9.
166	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.
167	Peng, R. D., M. L. Bell, et al. 2009. "Emergency admissions for cardiovascular and respiratory diseases and the
chemical composition of fine particle air pollution." Environmental Health Perspectives 117(6): 957-63.
168	Peng, R. D., H. H. Chang, et al. 2008. "Coarse particulate matter air pollution and hospital admissions for
cardiovascular and respiratory diseases among Medicare patients." Journal of the American Medical Association
299(18): 2172-9.
169	Bell ML; Ebisu K; Peng RD; Walker J; Samet JM; Zeger SL; Dominic F. 2008. Seasonal and regional short-term
effects of fine particles on hospital admissions in 202 U.S. counties, 1999-2005. Am J Epidemiol, 168: 1301-1310.
170	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.
171	Mar, T. F., J. Q. Koenig and J. Primomo. 2010. "Associations between asthma emergency visits and particulate
matter sources, including diesel emissions from stationary generators in Tacoma, Washington." Inhalation
Toxicology 22 (6): 445-8.
172	Slaughter, James C et al. 2005. "Association between particulate matter and emergency room visits, hospital
admissions and mortality in Spokane, Washington." Journal of Exposure Analysis and Environmental Epidemiology
15, 153-159.
173	Glad, J.A., L.L. Brink, E.O. Talbott, P.C. Lee, X. Xu, M. Saul, and J. Rager. 2012. "The Relationship of Ambient
ozone and PM2.5 Levels and Asthma Emergency Department Visits: Possible Influence of Gender and Ethnicity."
Archives of Environmental & Occupational Health 62 (2): 103-108.
174	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.
175	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.
176	Mar, T. F., T. V. Larson, et al. 2004. "An analysis of the association between respiratory symptoms in subjects
with asthma and daily air pollution in Spokane, Washington." Inhalation Toxicology 16(13): 809-15.
177	Ostro, B.D. 1987. "Air Pollution and Morbidity Revisited: A Specification Test." Journal of Environmental
Economics Management 14:87-98.
178	Ostro, B.D. and S. Rothschild. 1989. "Air Pollution and Acute Respiratory Morbidity: An Observational Study of
Multiple Pollutants." Environmental Research 50:238-247.
179	Pope, C.A., III, D.W. Dockery, J.D. Spengler, and M.E. Raizenne. 1991. "Respiratory Health and PM10
Pollution: A Daily Time Series Analysis." American Review of Respiratory Diseases 144:668-674.
180	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.
181	Smith, R.L.; B. Xu and P. Switzer. 2009. Reassessing the relationship between ozone and short- term mortality in
U.S. u Zanobetti, A; J. Schwartz. 2008. Mortality displacement in the association of ozone with mortality: an
analysis of 48 cities in the United States. American Journal of Respiratory and Critical Care Medicine. 177:184-
189.rban communities. Inhalation Toxicology. 21:37-61.
182	Zanobetti, A; J. Schwartz. 2008. Mortality displacement in the association of ozone with mortality: an analysis of
48 cities in the United States. American Journal of Respiratory and Critical Care Medicine. 177:184-189.

-------
183	Katsouyanni, K.; J.M. Samet; H.R. Anderson; R. Atkinson; A.L. Tertre; S. Medina, et al. 2009. Air Pollution and
Health: A European and North American Approach (APHENA). Health Effects Institute.
184	Glad, J.A., L.L. Brink, E.O. Talbott, P.C. Lee, X. Xu, M. Saul, and J. Rager. 2012. "The Relationship of Ambient
ozone and PM2.5 Levels and Asthma Emergency Department Visits: Possible Influence of Gender and Ethnicity."
Archives of Environmental & Occupational Health 62 (2): 103-108.
185	Ito, K.; G.D. Thurston and R.A. Silverman. 2007. Characterization of PM2.5, gaseous pollutants, and
meteorological interactions in the context of time-series health effects models. Journal of Exposure Science and
Environmental Epidemiology. 17(S2):S45-60.
186	Mar, TF; Koenig, JQ. 2009. Relationship between visits to emergency departments for asthma and ozone
exposure in greater Seattle, Washington. Ann Allergy Asthma Immunol 103: 474-
479.http://dx.doi.org/10.1016/S1081-1206(10)60263-3.
187	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.
188	Sarnat, J. A., Sarnat, S. E., Flanders, W. D., Chang, H. H., Mulholland, J., Baxter, L., Ozkaynak, H. 2013.
Spatiotemporally resolved air exchange rate as a modifier of acute air pollution-related morbidity in Atlanta. J Expo
Sci Environ Epidemiol, doi: 10.1038/jes.2013.32.
189	Wilson, A.M., C.P. Wake, T. Kelly et al. 2005. Air pollution and weather, and respiratory emergency room visits
in two northern New England cities: an ecological time-series study. Environ Res. Vol. 97 (3): 312-21.
190	Mortimer, KM; Neas, LM; Dockery, DW; Redline, S; Tager, IB. 2002. The effect of air pollution on innercity
children with asthma. EurRespir J 19: 699-705.
191	Schildcrout, J. S., Sheppard, L., Lumley, T., Slaughter, J. C., Koenig, J. Q., & Shapiro, G. G. (2006). Ambient air
pollution and asthma exacerbations in children: an eight-city analysis. Am J Epidemiol, 164(6), 505-517.
192	Chen L, Jennison BL, Yang W, and Omaye ST. 2000. "Elementary school absenteeism and air pollution."
Inhalation Toxicology 12(11):997-1016.
193	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.
194	Ostro, B.D. and S. Rothschild. 1989. "Air Pollution and Acute Respiratory Morbidity: An Observational Study of
Multiple Pollutants." Environmental Research 50:238-247.
195	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.
196	National Research Council (NRC). (2002). Estimating the Public Health Benefits of Proposed Air Pollution
Regulations. Washington, DC: The National Academies Press.
197
Abt Associates, Inc. October 2005. Methodology for County-level Mortality Rate Projections. Memorandum to
BryanHubbell and Zachary Pekar, U.S. EPA.
198	American Lung Association (ALA). 2002. Trends in Asthma Morbidity and Mortality. American Lung
Association, Best Practices and Program Services, Epidemiology and Statistics Unit. Available at
.
199	Adams PF, Hendershot GE, Marano MA. 1999. "Current Estimates from the National Health Interview Survey,
1996." Vital and Health Statistics 10(200): 1-212.
200	U.S. Census Bureau. 2000. Component Assumptions of the Resident Population by Age, Sex, Race, and Hispanic
Origin. Available online at: http://www.census.gov/population/www/projections/natdet-D5.html
201	Agency for Healthcare Research and Quality (AHRQ). 2007. HCUPnet, Healthcare Cost and Utilization Project.
Rockville, MD. Available at .
202	American Lung Association (ALA). 2010. Trends in Asthma Morbidity and Mortality. American Lung
Association Epidemiology and Statistics Unit Research and Program Services Division. February. Table 7. Based on
NHIS data (CDC, 2008). Available at http://www.lungusa.org/finding-cures/our-research/trend-reports/asthma-
trend-report.pdf.
203	Centers for Disease Control and Prevention (CDC). 2008. National Center for Health Statistics. National Health
Interview Survey, 1999-2008.
204	Freeman (III), AM. 1993. The Measurement of Environmental and Resource Values: Theory and Methods.
Washington, DC: Resources for the Future.
205	Harrington, W., and P.R. Portney. 1987. Valuing the Benefits of Health and Safety Regulation. Journal of Urban
Economics 22:101-112.
206	Berger, M.C., G.C. Blomquist, D. Kenkel, and G.S. Tolley. 1987. Valuing Changes in Health Risks: A
Comparison of Alternative Measures. The Southern Economic Journal 53:977-984.

-------
207	U.S. Environmental Protection Agency. (2012). Regulatory Impact Analysis for the Final Revisions to the
National Ambient Air Quality Standards for Particulate Matter. Prepared by: Office of Air and Radiation, Office of
Air Quality Planning and Standards. Retrieved April 27, 2016 at http://www3.epa.gov/ttn/ecas/ria.html. EPA-
452/R-12-005
208	U.S. Environmental Protection Agency. (2015). Regulatory Impact Analysis of the Final Revisions to the
National Ambient Air Quality Standards for Ground-Level Ozone. Prepared by: Office of Air and Radiation, Office
of Air Quality Planning and Standards. Retrieved April 27, 2016 at http://www3.epa.gov/ttn/ecas/ria.html. EPA-
452/R-15-007.
209	U.S. Environmental Protection Agency (U.S. EPA). 2010. Guidelines for Preparing Economic Analyses. EPA
240-R-10-001. National Center for Environmental Economics, Office of Policy Economics and Innovation.
Washington, DC. December. Available at .
214	Smith, D.H., D.C. Malone, K.A. Lawson, L.J. Okamoto, C. Battista, and W.B. Saunders. 1997. "A National
Estimate of the Economic Costs of Asthma." American Journal of Respiratory and Critical Care Medicine 156(3 Pt
l):787-793.
215	Stanford, R., T. McLaughlin and L. J. Okamoto. 1999. "The cost of asthma in the emergency department and
hospital." American Journal of Respiratory and Critical Care Medicine 160 (1): 211-5.
216	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.
217	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],
218	Tolley, G.S. et al. 1986. Valuation of Reductions in Human Health Symptoms and Risks. University of Chicago.
Final Report for the U.S. Environmental Protection Agency, Office of Policy Analysis. January. Available at
.
219	U.S. Environmental Protection Agency (U.S. EPA). 2015. BenMAP-CE User's Manual Appendices. Office of
Air Quality Planning and Standards. Research Triangle Park, NC. Available at
.
220	U.S. EPA, (2001), "Guidance for Demonstrating Attainment of Air Quality Goals for PM2.5 and Regional
Haze," http://www3.epa.gov/ttn/scram/guidance_sip.htm, Modeling Guidance, DRAFT-PM
221	National Research Council (NRC). 2002. Estimating the Public Health Benefits of Proposed Air Pollution
Regulations. Washington, DC: The National Academies Press.
222	National Research Council (NRC). 2008. Estimating Mortality Risk Reduction and Economic Benefits from
Controlling Ozone Air Pollution. National Academies Press. Washington, DC.

-------
Chapter 9. Safety Impacts
9.1 Summary of Supporting HD Vehicle Safety Research
As discussed in the Notice of Proposed Rulemaking, NHTSA and EPA considered the
potential safety impact of technologies that improve Medium- and Heavy-Duty vehicle fuel
efficiency and GHG emissions when determining potential regulatory alternatives. The safety
assessment of the technologies in this rule was informed by two comprehensive NAS reports, an
extensive analysis of safety effects of HD pickups and vans using estimates from the DOT report
on the effect of mass reduction and vehicle size on safety, and focused agency-sponsored safety
testing and research. The following section provides a concise summary of the literature and
work considered by the agencies in development of this final rule.
9.1.1 National Academy of Sciences HD Phase 1 and Phase 2 Reports
As required by EISA, the National Research Council has been conducting continuing
studies of the technologies and approaches for reducing The Fuel Consumption of Medium- and
Heavy-Duty Vehicles. The first was a report issued in 2010, "Technologies and Approaches to
Reducing the Fuel Consumption of Medium- and Heavy-Duty Vehicles" ("NAS Report"). The
second a report issued in 2014, "Reducing the Fuel Consumption and Greenhouse Gas Emissions
of Medium- and Heavy-Duty Vehicles, Phase Two-First Report" ("NAS HD Phase 2 First
Report"). While the reports primarily focused on reducing vehicle fuel consumption and
emissions through technology application, and examined potential regulatory frameworks, both
reports contain findings and recommendations related to safety. In developing this rule, the
agencies carefully considered the reports' findings related to safety.
In particular, NAS indicated that idle reduction strategies can also accommodate for the
safety of the driver in both hot and cold weather conditions. The agencies considered this
potential approach for application of idle reduction technologies by allowing for override
provisions, as defined in 40 CFR 1037.660(b), where operator safety is a primary consideration.
Override is allowed if the external ambient temperature reaches a level below which or above
which the cabin temperature cannot be maintained within reasonable heat or cold exposure
threshold limit values for the health and safety of the operator (not merely comfort).
NAS also reported extensively on the emergence of natural gas (NG) as a viable fuel
option for commercial vehicles, but alluded to the existence of uncertainties regarding its safety.
The committee found that while the public crash databases do not contain information on vehicle
fuel type, the information, at the time of the report, indicates that the crash-related safety risk for
NG storage on vehicles does not appear to be appreciably different from diesel fuel risks. The
committee also found that while there are two existing SAE-recommended practice standards for
NG-powered HD vehicles, the industry will benefit from best practice directives to minimize
crash risks for NG fuel tanks, such as on shielding to prevent punctures during crashes. As a
final point, NAS stated that manufacturers and operators have a great incentive to prevent
possible NG leakage from a vehicle fuel system because it would be a significant safety concern
and reduce vehicle range. No recommendations were made for additional Federal safety

-------
regulations for these vehicles. In response, the agencies reviewed and discuss the existing NG
vehicle standards and best practices cited by NAS in Section XI of the NPRM.
In the NAS Committee's Phase 1 report, the Committee indicated that aerodynamic
fairings detaching from trucks on the road could be a potential safety issue. However, the Phase
2 interim report stated that "Anecdotal information gained during the observations of on-road
trailers indicates a few skirts badly damaged or missing from one side. The skirt manufacturers
report no safety concerns (such as side skirts falling off) and little maintenance needed."
The NAS report also identified the link between tire inflation and condition and vehicle
stopping distance and handling, which impacts overall safety. The committee found that tire
pressure monitoring systems and automatic tire inflation systems are being adopted by fleets at
an increasing rate. However, the committee noted that there are no standards for performance,
display, and system validation. The committee recommended that NHTSA issue a white paper
on the minimum performance of tire pressure systems from a safety perspective.
The agencies considered the safety findings in both NAS reports in developing this rule
and conducted additional research on safety to further examine information and findings of the
reports.
9.1.2 DOT CAFE Model HD Pickup and Van Safety Analysis
This analysis considered the potential crash safety effects on the technologies
manufacturers may apply to HD pickups and vans to meet each of the regulatory alternatives
evaluated in the NPRM. NHTSA research has shown that vehicle mass reduction affects overall
societal fatalities associated with crashes and, most relevant to this rule, that mass reduction in
heavier light- and medium-duty vehicles has an overall beneficial effect on societal fatalities.
Reducing the mass of a heavier vehicle involved in a multiple vehicle crash reduces the
likelihood of fatalities among the occupants of the other vehicle(s). In addition to the effects of
mass reduction, the analysis anticipates that these standards, by reducing the cost of driving HD
pickups and vans, will lead to increased travel by these vehicles and, therefore, more crashes
involving these vehicles. The Method A and B analyses, included in the NPRM, consider overall
impacts from both of these factors, using a methodology similar to NHTSA's analyses for the
MYs 2017 - 2025 CAFE and GHG emission standards.
The Method A analysis includes estimates of the extent to which HD pickups and vans
produced during MYs 2014-2030 may be involved in fatal crashes, considering the mass,
survival, and mileage accumulation of these vehicles, taking into account changes in mass and
mileage accumulation under each regulatory alternative. These calculations make use of the
same coefficients applied to light trucks in the MYs 2017-2025 CAFE rulemaking analysis. As
discussed above, vehicle miles traveled may increase due to the fuel economy rebound effect,
resulting from improvements in vehicle fuel efficiency and cost of fuel, as well as the assumed
future growth in average vehicle use. Increases in total lifetime mileage increase exposure to
vehicle crashes, including those that result in fatalities. Consequently, the modeling system
computes total fatalities attributed to vehicle use for vehicles of a given model year based on
safety class and weight threshold. These calculations also include a term that accounts for the
fact that vehicles involved in future crashes will be certified to more stringent safety standards

-------
than those involved with past crashes upon which the base rates of involvement in fatal crashes
were estimated. Since the use of mass reducing technology is present within the model, safety
impacts may also be observed whenever a vehicle's base weight decreases. Thus, in addition to
computing total fatalities related to vehicle use, the modeling system also estimates changes in
fatalities due to reduction in a vehicle's curb weight.
The total fatalities attributed to vehicle use and vehicle weight change for vehicles of a
given model year are then summed. Lastly, total fatalities occurring within the industry in a
given model year are accumulated across all vehicles. In addition to using inputs to estimate the
future involvement of modeled vehicles in crashes involving fatalities, the model also applies
inputs defining other accident-related externalities estimated on a dollar per mile basis. For
vehicles above 4,594 pounds—i.e., the majority of the HD pickup and van fleet—mass reduction
is estimated to reduce the net incidence of highway fatalities by 0.34 percent per 100 pounds of
removed curb weight. For the few HD pickups and vans below 4,594 pounds, mass reduction is
estimated to increase the net incidence of highway fatalities by 0.52 percent per 100 pounds.
Because there are many more HD pickups and vans above 4,594 pounds than below 4,594
pounds, the overall effect of mass reduction in the segment is estimated to reduce the incidence
of highway fatalities. The estimated increase in vehicle miles traveled due to the fuel economy
rebound effect is estimated to increase exposure to vehicle crashes and offset these reductions.
9.1.3 Volpe Research on MD/HD Fuel Efficiency Technologies
The 2010 NAS Report recommended that NHTSA perform a thorough safety analysis to
identify and evaluate potential safety issues with fuel efficiency-improving technologies. The
Department of Transportation Volpe Center's 2015 report titled "Review and Analysis of
Potential Safety Impacts of and Regulatory Barriers to Fuel Efficiency Technologies and
Alternative Fuels in Medium- and Heavy-Duty Vehicles"1 summarizes research and analysis
findings on potential safety issues associated with both the diverse alternative fuels (natural gas-
CNG and LNG, propane, biodiesel, and power train electrification), and the specific FE
technologies recently adopted by the MD/HDV fleets. These include Intelligent Transportation
Systems (ITS) and telematics, speed limiters, idle reduction devices, tire technologies (single-
wide tires, and tire pressure monitoring systems-TPMS and Automated Tire Inflation Systems-
ATIS), aerodynamic components, vehicle light-weighting materials, and Long Combination
Vehicles (LCVs).
Chapter 1 provides an overview of the study's rationale, background, and key objective,
namely, to identify the technical and operational/behavioral safety benefits and disbenefits of
MD/HDVs equipped with FE technologies and using emerging alternative fuels (AFs). Recent
MD/HDV national fleet crash safety statistical averages are also provided for context, although
no information exists in crash reports relating to specific vehicle FE technologies and fuels.
(NHTSA/FARS and FMCSA/CSA databases do not include detailed information on vehicle fuel
economy technologies, since the state crash report forms are not coded down to an individual
fuel economy technology level).
Chapters 2 and 3 are organized by clusters of functionally-related FE technologies for
vehicles and trailers (e.g., tire systems, ITS, light-weighting materials, and aerodynamic systems)
and alternative fuels, which are described and their respective associated potential safety issues

-------
are discussed. Chapter 2 summarizes the findings from a comprehensive review of available
technical and trade literature and Internet sources regarding the benefits, potential safety hazards,
and the applicable safety regulations and standards for deployed FE technologies and alternative
fuels. Chapter 2 safety-relevant fuel-specific findings include:
Both CNG- and LNG-powered vehicles present potential hazards, and call for well-
known engineering and process controls to assure safe operability and
crashworthiness. However, based on the reported incident rates of NGVs and the
experiences of adopting fleets, it appears that NGVs can be operated at least as safely
as diesel MD/HDVs.
There are no safety contraindications to the large scale fleet adoption of CNG or LNG
fueled heavy duty trucks and buses, and there is ample experience with the safe
operation of large public transit fleets. Voluntary industry standards and best
practices suffice for safety assurance, though improved training of CMV operators
and maintenance staff in natural gas safety of equipment and operating procedures is
needed.
Observing CNG and LNG fuel system and maintenance facility standards, coupled
with sound design, manufacture, and inspection of natural gas storage tanks will
further reduce the potential for leaks, tank ruptures, fires, and explosions.
Biodiesel blends used as drop-in fuels have presented some operational safety
concerns dependent on blending fraction, such as material compatibility, bio-fouling
sludge accumulation, or cold-weather gelling. However, best practices for biodiesel
storage, and improved gaskets and seals that are biodiesel resistant, combined with
regular maintenance and leak inspection schedules for the fuel lines and components
enable the safe use of biodiesel in newer MD/HDVs
Propane (LPG, or autogas) presents well-known hazards including ignition (due to
leaks or crash) that are preventable by using Overfill Prevention Devices (OPDs),
which supplement the automatic stop-fill system on the fueling station side, and
pressure release devices (PRDs). Established best practices and safety codes (e.g.,
NFPA) have proven that propane fueled MD/HDVs can be as operationally safe as
the conventionally-fueled counterparts.
As the market penetration of hybrid and electric drivetrain accelerates, and as the
capacity and reliability of lithium ion batteries used in Rechargeable Energy Storage
Systems (RESS) improve, associated potential safety hazards (e.g., electrocution from
stranded energy, thermal runaway leading to battery fire) have become well
understood, preventable, and manageable. Existing and emerging industry technical
and safety voluntary standards, applicable NHTS A regulations and guidance, and the
growing experience with the operation of hybrid and electric MD/HDVs will enable
the safe operation and large-scale adoption of safer and more efficient power-train
electrification technologies.

-------
The safety findings from literature review pertaining to the specific FE technologies
implemented to date in the MD/HDV fleet include:
Telematics—integrating on-board sensors, video, and audio alerts for MD/HDV
drivers—offer potential improvements in both driver safety performance and fuel
efficiency. Both camera and non-camera based telematics setups are currently
integrated with available crash avoidance systems (such as ESC, RSC, LDWS, etc.)
and appear to be well accepted by MD/HDV fleet drivers.
Both experience abroad and the cited US studies of trucks equipped with active speed
limiters indicated a safety benefit, as measured by up to 50 percent reduced crash
rates, in addition to fuel savings and other benefits, with good CMV driver
acceptance. Any negative aspects were small and avoidable if all the speed limitation
devices were set to the same speed, so there would be less need for overtaking at
highway speeds.
• No literature reports of adverse safety impacts were found regarding implementation
of on-board idle-reduction technologies in MD/HDVs (such as automatic start-stop,
direct-fired heaters, and APUs).
There was no clear consensus from the literature regarding the relative crash rates and
highway safety impacts of LCVs, due to lack of sufficient data and controls and
inconsistent study methodologies. Recent safety evaluations of LCVs and ongoing
MAP-21 mandated studies will clarify and quantify this issue.
Tire technologies for FE (including ATIS, TPMS, LRR and single-wide tires)
literature raised potential safety concerns regarding lower stability or loss of control,
e.g., when tire pressure is uneven or a single wide tire blows out on the highway.
However, systems such as automated tire monitoring systems and stability enhancing
electronic systems (ABS, ESC, and RSC) may compensate and mitigate any adverse
safety impacts.
Aerodynamic technologies that offer significant fuel savings have raised potential
concerns about vehicle damage or injury in case of detached fairings or skirts,
although there were no documented incidents of this type in the literature.
Some light weighting materials may pose some fire safety and crashworthiness
hazards, depending on their performance in structural or other vehicle subsystem
applications (chassis, power-train, and crash box or safety cage). Some composites
(fiberglass, plastics, CFRC, foams) may become brittle on impact or due to
weathering from UV exposure or extreme cold. Industry has developed advanced,
high performance lightweight material options tailored to their automotive
applications, e.g., thermoplastics resistant to UV and weathering. No examples of
such lightweight material failures on MD/HDVs were identified in the literature.
Chapter 3 provides complementary inputs on the potential safety issues associated with
FE technologies and alternative fuels obtained from Subject Matter Experts (SMEs). The broad

-------
cross-section of SMEs consulted had experience with the operation of "green" truck and bus
fleets, were Federal program managers, or were industry developers of FE systems for
MD/HDVs. Safety concerns raised by the SMEs can be prevented or mitigated by complying
with applicable regulations and safety standards and best practices, and are being addressed by
evolving technologies, such as electronic collision prevention devices. Although SMEs raised
some safety concerns, their experience indicates that system- or fuel-specific hazards can be
prevented or mitigated by observing applicable industry standards, and by training managers,
operators and maintenance staff in safety best practices. Specific safety concerns raised by
SMEs based on their experience included:
Alternative fuels did not raise major safety concerns, but generally required better
education and training of staff and operators. There was a concern expressed
regarding high pressure (4000 psi) CNG cylinders that could potentially explode in a
crash scenario or if otherwise ruptured. However, aging CNG fuel tank safety can be
assured by enforcing regulations such as FMVSS No. 304, and by periodic inspection
and end-of-life disposal and replacement. A propane truck fleet manager stated that
the fuel was as safe as or safer than gasoline, and reported no safety issues with the
company's propane, nor with hybrid gasoline-electric trucks. OEMs of drivetrain
hybridization and electrification systems, including advanced Lithium Ion batteries
for RESS, indicated that they undergo multiple safety tests and are designed with fail-
safes for various misuse and abuse scenarios. Integration of hybrid components
downstream by bodybuilders in retrofits, as opposed to new vehicles, was deemed a
potential safety risk. Another potential safety concern raised was the uncertain
battery lifetime due to variability of climate, duty-cycles, and aging. Without state-
of-charge indicators, this could conceivably leave vehicles underpowered or stranded
if the battery degrades and is not serviced or replaced in a timely manner.
ITS and telematics raised no safety concerns; on the contrary, fleet managers stated
that "efficient drivers are safer drivers." Monitoring and recording of driver behavior,
combined with coaching, appeared to reduce distracted and aggressive driving and
provided significant FE and safety benefits.
A wide-base single tire safety concern was the decrease in tire redundancy in case of
a tire blowout at highway speeds. For LRRs, a concern was that they could
negatively affect truck stopping distance and stability control.
A speed-limiter safety concern was related to scenarios when such trucks pass other
vehicles on the highway instead of staying in the right-hand lane behind other
vehicles. By combining speed limiters with driver training programs, overall truck
safety could actually improve, as shown by international practice.
Aerodynamic systems' safety performance to date was satisfactory, with no instances
of on-road detaching. However, covering underside or other components with
aerodynamic fairings can make them harder to inspect, such as worn lugs, CNG relief
valve shrouds, wheel covers, and certain fairings. Drivers and inspectors need to be
able to see through wheel covers and to be able to access lug nuts through them.
These covers must also be durable to withstand frequent road abuse.

-------
For lightweighting materials, the safety concern raised was lower crashworthiness
(de-bonding or brittle fracture on impact) and the potential for decreased survivability
in vehicle fires depending on the specific material choice and its application.
The key finding from the literature review and SME interviews is that there appear to be
no major safety hazards preventing the adoption of FE technologies, or the increased use of
alternative fuels and vehicle electrification. In view of the scarcity of hard data currently
available on actual highway crashes that can be directly or causally attributed to adoption of FE
technologies and/or alternative fuels by MD/HDVs, and the limited experience with commercial
truck and transit bus fleets operations equipped with these technologies, it was not possible to
perform a quantitative, probabilistic risk assessment, or even a semi-quantitative preliminary
hazard analysis (PHA).
Chapter 4 employs a deterministic scenario-based hazard analysis of potential crash or
other safety concerns identified from the literature review or raised by subject matter experts
(SMEs) interviewed (e.g., interfaces with charging or refueling infrastructure). For each specific
hazard scenario discussed, the recommended prevention or mitigation options, including
compliance with applicable NHTSA or FMCSA regulations, and voluntary industry standards
and best practices are identified, along with FE technology or fuel-specific operator training.
SMEs safety concerns identified in Sec 3.3 were complemented with actual incidents, and
developed into the hazard scenarios analyzed in Chapter 4.
The scenario-based deterministic hazard analysis reflected not only the literature findings
and SMEs' safety concerns, but also real truck or bus mishaps that have occurred in the past.
Key hazard analysis scenarios included: CNG-fueled truck and bus vehicle fires or explosions
due to tank rupture, when pressurized fuel tanks were degraded due to aging or when PRDs
failed; LNG truck crashes leading to fires, or LNG refueling-related mishaps; the flammability or
brittle fracture issues related to lightweighting materials in crashes; reduced safety performance
for either LRR or wide-base tires; highway pile-ups when LCVs attempt to pass at highway
speeds; aerodynamic components detaching while the vehicle traveled on a busy highway or
urban roadway; and fires resulting in overheated lithium ion batteries in electric or hybrid buses.
These hypothetical worst case scenarios appear to be preventable or able to be mitigated by
observing safety regulations and voluntary standards, or with engineering and operational best
practices.
Chapter 5 reviews and discusses the existing federal and state regulatory framework for
safely operating MD/HDVs equipped with FE technologies or powered by alternative fuels. The
review identifies potential regulatory barriers to their large-scale deployment in the national fleet
that could delay achievement of desired fuel consumption and environmental benefits, while
ensuring equal or better safety performance.
Chapter 6 summarizes the major findings and recommendations of this preliminary safety
analysis of fuel efficiency technologies and alternative fuels adopted by MD/HDVs. The
scenario-based hazard analysis, based on the literature review and experts' inputs, indicates that
MD/HDVs equipped with advanced FE technologies and/or using alternative fuels have
manageable potentially adverse safety impacts. The findings suggest that the potential safety

-------
hazards identified during operation, maintenance, and crash scenarios can be prevented or
mitigated by complying with safety regulations and voluntary standards and industry best
practices. The study also did not identify any major regulatory barriers to rapid adoption of FE
technologies and alternative fuels by the MD/HDV fleet.
9.1.4	Oak Ridge National Laboratory (ORNL) Research on Low Rolling
Resistance Truck Tires
DOT's Federal Motor Carrier Safety Administration and NHTSA sponsored a test
program conducted by Oak Ridge National Laboratory to explore the effects of tire rolling
resistance levels on Class 8 tractor-trailer stopping distance performance over a range of loading
and surface conditions.2 The objective was to determine whether there a relationship exists
between tire rolling resistance and stopping distance for vehicles of this type. The overall results
of this research suggest that tire rolling resistance is not a reliable indicator of Class 8 tractor-
trailer stopping distance. The correlation coefficients (R2 values) for linear regressions of wet
and dry stopping distance versus overall vehicle rolling resistance values did not meet the
minimum threshold for statistical significance for any of the test conditions. Correlation between
CRR and stopping distance was found to be negligible for the dry tests for both loading
conditions. While correlation was higher for the wet testing (showing a slight trend in which
lower CRRs correspond to longer stopping distances), it still did not meet the minimum
threshold for statistical significance. In terms of compliance with Federal safety standards, it
was found that the stopping distance performance of the vehicle with the four tire sets studied in
this research (with estimated tractor CRRs which varied by 33 percent), were well under the
FMVSSNo. 121 stopping distance requirements.
9.1.5	Additional Safety Considerations
The agencies' considered the-Organic Rankine Cycle waste heat recovery (WHR) as a
fuel saving technology in the rulemaking timeframe. The basic approach of these systems is to
use engine waste heat from multiple sources to evaporate a working fluid through a heat
exchanger, which is then passed through a turbine or equivalent expander to create mechanical or
electrical power. The working fluid is then condensed as it passes through a heat exchanger and
returns to back to the fluid tank, and pulled back to the flow circuit through a pump to continue
the cycle.
Despite the promising performance of pre-prototype WHR systems, manufacturers have
not yet arrived at a consensus on which working fluid(s) to be used in WHR systems to balance
concerns regarding performance, global warming potential (GWP), and safety. Working fluids
have a high GWP (conventional refrigerant), are expensive (low GWP refrigerant), are hazardous
(such as ammonia, etc.), are flammable (ethanol/methanol), or can freeze (water). One challenge
is determining how to seal the working fluid properly under the vacuum condition and high
temperatures to avoid safety issues for flammable/hazardous working fluids. Because of these
challenges, choosing a working fluid will be an important factor for system safety, efficiency,
and overall production viability.
The agencies believe manufacturers will require additional time and development
effort to assure that a working fluid that is both appropriate, given the noted challenges, and has

-------
a low GWP for use in waste heat recovery systems. Based on this and other factors, the analysis
used for both the proposed Preferred Alternative and for this final rule assumes that WHR will
not achieve a significant market penetration for diesel tractor engines (i.e., greater than 5
percent) until 2027, which will provide time for these considerations to be addressed. The
agencies assume no use of this technology in the HD pickups and vans and vocational vehicle
segments.
9.2 Safety Related Comments to the NPRM
The agencies received safety related to the NPRM focused on the vehicle and operator
safety benefits of central tire inflation systems, potential safety and traction impacts of low
rolling resistance tires, and recommendations that NHTSA continue evaluations of potential
safety impacts of fuel saving technologies.
AIR CTI, Inc., a supplier of central tire inflation systems, highlighted the safety benefits
to both vehicle operation and the operators themselves through proper tire pressure management.
More specifically, the proper tire inflation levels for the load being carried contributes to both
proper handing for road conditions and reducing irregular road surface vibration from being
transmission to vehicle component and, ultimately, the vehicle operator, where there may be
potential health implications over prolonged exposure.
The agencies appreciate the additional points provided by AIR CTI in terms of not only
the potential fuel efficiency benefits of central tire inflation systems but the potential equipment
longevity benefits, vehicle dynamic impacts, and the potential to reduce driver fatigue and injury
through proper tire inflation for the load being carried.
The American Trucking Associations (ATA) commented on the potential impact of Low
Rolling Resistance Tires by indicating that, "The safety effects of LRRTs are not totally
understood. While the ".. .agencies analysis indicate that this proposal should have no adverse
impact on vehicle or engine safety," ATA remains leery of potential unintended consequences
resulting from new generation tires that have yet to be developed. This especially holds true in
terms of overall truck braking distances." The Owner-Operator Independent Drivers Association
(OOIDA) similarly commented on LRRTs and their ability to meet the tractions needs in
mountainous regions.
The agencies continue to stand behind the low rolling resistance tire research conducted
to date, which includes the study mentioned in the previous section, along with any research
supporting the development, and maintenance, of FMVSS No. 121. The agencies agree, though,
that continuing research will be important as new tire technologies enter the marketplace, and
like the extensive rolling resistance testing conducting to support the Phase 1 regulation and, in
part, this final rule, the agencies will continue to monitor developments in the tire supply
marketplace through the EPA SmartWay program and other, potential, research. NHTSA notes
that FMVSS No. 121 will continue to play a role in ensuring the safety of both current and future
tire technologies.
The ATA also expressed support for the NHTSA study mentioned in the previous
section, Review and Analysis of Potential Safety Impacts of and Regulatory Barriers to Fuel

-------
Efficiency Technologies and Alternative Fuels in Medium- and Heavy-Duty Vehicles. More
specifically, ATA requested that DOT/NHTSA and the DOT Volpe Center continue "to assess
and evaluate potential safety impacts that may be attributed to the use of fuel efficiency devices."
The agencies appreciate ATA's support and acknowledge of this comprehensive, peer-reviewed
assessment and we look forward to continuing this work as the need arises.
9.3 The Agencies' Assessment of Potential Safety Impacts
NHTSA and EPA considered the potential safety impact of technologies that improve
MDHD vehicle fuel efficiency and GHG emissions as part of the assessment of regulatory
alternatives and selection of the final regulatory approach. The safety assessment of the
technologies in this final rule was informed by two NAS reports, an analysis of safety effects of
HD pickups and vans using estimates from the DOT report on the effect of mass reduction and
vehicle size on safety, and agency-sponsored safety testing and research. The agencies
considered safety from the perspective of both direct effects and indirect effects.
In terms of direct effects on vehicle safety, research from NAS and Volpe, and direct
testing of technologies like the ORNL tire work, indicate that there are no major safety hazards
associated with the adoption of technologies that improve MDHD vehicle fuel efficiency and
GHG emissions or the increased use of alternative fuels and vehicle electrification. The findings
suggest that the potential safety hazards identified during operation, maintenance, and crash
scenarios can be prevented or mitigated by complying with safety regulations, voluntary
standards, and industry best practices. Tire testing showed tire rolling resistance did not impact
of Class 8 tractor-trailer stopping distance for the tires tested. For HD pickup and vans, mass
reduction is anticipated to reduce the net incidence of highway fatalities, more specifically
related to the majority of HD pickup and vans weigh more than 4,594 lbs. Taken together, these
studies suggest that the fuel efficiency improving technologies assessed in the studies can be
implemented with no degradation in overall safety.
However, analysis anticipates that the indirect effect of these standards, by reducing the
operating costs, will lead to increased travel by tractor-trailers and HD pickups and vans and,
therefore, more crashes involving these vehicles.

-------
References
1	Brecher, A., Epstein, A. K„ & Breck, A. (2015, June). "Review and analysis of potential safety impacts of and
regulatory barriers to fuel efficiency technologies and alternative fuels in medium- and heavy-duty vehicles."
(Report No. DOT HS 812 159). Washington DC: National Highway Traffic Safety Administration.
2	Lascurain, M.B. (2015, June). "Effects of tire rolling resistance levels on Class 8 tractor trailer stopping distance
performance." Washington, DC: National Highway Traffic Safety Administration.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Chapter 10: CAFE Model for HD Pickups and Vans
In the NPRM, the agencies conducted coordinated and complementary analyses using
two analytical methods for the heavy-duty pickup and van segment, both of which used the same
version of NHTSA's CAFE model to analyze technology. The agencies have also used two
analytical methods for the joint final rule. However, unlike the NPRM, for the joint final rule,
the agencies are using different versions of NHTSA's CAFE model to analyze technology. The
Method B approach continues to use the same version of the model and inputs that was used for
both methods in the NPRM. Method A uses an updated version of the CAFE model and some
updated inputs.
In this chapter, both versions of the CAFE modeling system are described and used to
analyze technology use and per-vehicle costs under each regulatory alternative, including the no
action alternative (which reflects continuation of previously-promulgated standards). The
Method A analysis uses the CAFE model which includes changes made subsequent to the
NPRM, and the Method B analysis uses the CAFE model which includes only those changes
made for the NPRM. However, this model is more comprehensive and also projects other
impacts. NHTSA addresses these other impacts in the EIS and these are also presented here.A
NHTSA developed the CAFE model in 2002 to support the 2003 issuance of CAFE
standards for MYs 2005-2007 light trucks. NHTSA has since significantly expanded and refined
the model, and has applied the model to support every ensuing CAFE rulemaking for both light-
duty and heavy-duty. For this analysis, the model was reconfigured to use the work based
attribute metric of "work factor" established in the Phase 1 rule instead of the light duty
"footprint" attribute metric.
Past analyses conducted using the CAFE model have been subjected to extensive and
detailed review and comment, much of which has informed the model's expansion and
refinement. NHTSA's use of the model was considered and supported in Center for Biological
Diversity v. National Highway Traffic Safety Admin., 538 F.3d 1172, 1194 (9th Cir. 2008). For
further discussion see 76 FR 57198, and the model has been subjected to formal peer review and
review by the General Accounting Office (GAO) and National Research Council (NRC).
NHTSA makes public the model, source code, and—except insofar as doing so will compromise
confidential business information (CBI) manufacturers have provided to NHTSA—all model
inputs and outputs underlying published rulemaking analyses.
Although the CAFE model can also be used for more aggregated analysis (e.g., involving
"representative vehicles," single-year snapshots, etc.), NHTSA designed the model with a view
toward (a) detailed simulation of manufacturers' potential actions given a defined set of
standards, followed by (b) calculation of resultant impacts and economic costs and benefits. The
model is intended to describe actions manufacturers could take in light of defined standards and
other input assumptions and estimates, not to predict actions manufacturers will take in light of
A EPA uses its MOVES model to project these other impacts as discussed in Chapters 5 through 8 of this RIA.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
competing product and market interests (e.g. engine power, customer features, technology
acceptance, etc.).
For the proposal, the agencies conducted coordinated and complementary analyses using
two analytical methods for the heavy-duty pickup and van segment by employing both NHTSA's
CAFE model and EPA's MOVES model. The agencies used EPA's MOVES model to estimate
fuel consumption and emissions impacts for tractor-trailers (including the engine that powers the
tractor), and vocational vehicles (including the engine that powers the vehicle). Additional
calculations were performed to determine corresponding monetized program costs and benefits.
For heavy-duty pickups and vans, the agencies performed complementary analyses, which we
refer to as "Method A" and "Method B."
For the final rule, NHTSA's Method A uses a modified version of the CAFE model
developed since the NPRM, as well as accompanying updates to CAFE model inputs, to project
a pathway the industry could use to comply with each regulatory alternative and the estimated
effects on fuel consumption, emissions, benefits and costs were industry to do so. Method A is
presented below in Section 10.2 and differs from the Method A analysis provided in the NPRM.
NHTSA considered the results of the Method A analysis for decision making for the final rule.
EPA's Method B analysis continues to use the CAFE model and inputs developed for the
NPRM to identify technology pathways the industry could potentially use to comply with each
regulatory alternative, along with resultant impacts on per vehicle costs should that compliance
path be utilized, and the MOVES model was used to calculate corresponding changes in total
fuel consumption and annual emissions. The results are presented in Section 10.3. Additional
calculations were performed to determine corresponding monetized program costs and benefits.
NHTSA's consideration of the Method A analysis and EPA's consideration of the Method B
analysis led the agencies to the same conclusions regarding the selection of the Phase 2
standards. See Sections 10.2 and 10.3 for additional discussion of these two methods and the
feasibility of the standards.
10.1 Overview of the CAFE Model
As a starting point, the model makes use of an input file defining the analysis fleet—that
is, a set of specific vehicle models (e.g., Ford F250) and model configurations (e.g., Ford F250
with 6.2-liter V8 engine, 4WD, and 6-speed manual transmission) estimated or assumed to be
produced by each manufacturer in each model year to be included in the analysis. The analysis
fleet includes key engineering attributes (e.g., curb weight, payload and towing capacities,
dimensions, presence of various fuel-saving technologies) of each vehicle model, engine, and
transmissions, along with estimates or assumptions of future production volumes. It also
specifies the extent to which specific vehicle models share engines, transmissions, and vehicle
platforms, and describes each manufacturer's estimated or assumed product cadence (i.e., timing
for freshening and redesigning different vehicles and platforms). This input file also specifies a
payback period used to estimate the potential that each manufacturer might apply technology to
improve fuel economy beyond levels required by standards.
A second input file to the model contains a variety of contextual estimates and
assumptions. Some of these inputs, such as future fuel prices and vehicle survival and mileage

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
accumulation (versus vehicle age), are relevant to estimating manufacturers' potential
application of fuel-saving technologies. Some others, such as fuel density and carbon content,
vehicular and upstream emission factors, the social cost of carbon dioxide emissions, and the
discount rate, are relevant to calculating physical and economic impacts of manufacturers'
application of fuel-saving technologies.
A third input file contains estimates and assumptions regarding the future applicability,
availability, efficacy, and cost of various fuel-saving technologies. Efficacy is expressed in
terms of the percentage reduction in fuel consumption, cost is expressed in dollars, and both
efficacy and cost are expressed on an incremental basis (i.e., estimates for more advanced
technologies are specified as increments beyond less advanced technologies). The input file also
includes "synergy factors" used to make adjustments accounting for the potential that some
combinations of technologies may result fuel savings or costs different from those indicated by
incremental values.
Finally, a fourth model input file specifies standards to be evaluated. Standards are
defined on a year-by-year basis separately for each regulatory class (passenger cars, light trucks,
and heavy-duty pickups and vans). Regulatory alternatives are specified as discrete scenarios,
with one scenario defining the no-action alternative or "baseline," all other scenarios defining
regulatory alternatives to be evaluated relative to that no-action alternative.
Given these inputs, the model estimates each manufacturer's potential year-by-year
application of fuel-saving technologies to each engine, transmission, and vehicle. Subject to a
range of engineering and planning-related constraints (e.g., secondary axle disconnect can't be
applied to 2-wheel drive vehicles, many major technologies can only be applied practicably as
part of a vehicle redesign, and applied technologies carry forward between model years), the
model attempts to apply technology to each manufacturer's fleet in a manner that minimizes
"effective costs" (accounting, in particular, for technology costs and avoided fuel outlays),
continuing to add improvements as long as doing so will help toward compliance with specified
standards or will produce fuel savings that "pay back" at least as quickly as specified in the input
file mentioned above.
After estimating the extent to which each manufacturer might add fuel-saving
technologies under each specified regulatory alternative, the model calculates a range of physical
impacts, such as changes in highway travel (i.e., VMT), changes in fleetwide fuel consumption,
changes in highway fatalities, and changes in vehicular and upstream greenhouse gas and criteria
pollutant emissions. The model also applies a variety of input estimates and assumptions to
calculate economic costs and benefits to vehicle owners and society, based on these physical
impacts. These are considered Method A results.
Since the manufacturers of HD pickups and vans generally only have one basic pickup
truck and van with different versions {i.e., different wheelbases, cab sizes, two-wheel drive, four-
wheel drive, etc.) there exists less flexibility than in the light-duty fleet to coordinate model
improvements over several years. As such, the CAFE model allows changes to the HD pickups
and vans to meet new standards according to estimated redesign cycles included as a model
input. As noted above, the opportunities for large-scale changes {e.g., new engines,
transmission, vehicle body and mass) thus occur less frequently than in the light-duty fleet,

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
typically at spans of eight or more years for this analysis. However, opportunities for gradual
improvements not necessarily linked to large scale changes can occur between the redesign
cycles (i.e., model refresh). Examples of such improvements are upgrades to an existing vehicle
model's engine, transmission and aftertreatment systems.
10.1.1 How Did the Agencies Develop the Analysis Fleets
As discussed above, both agencies used a version of NHTSA's CAFE modeling system
to estimate technology costs and application rates under each regulatory alternative considered in
the NPRM. The modeling system relies on many inputs, including an analysis fleet. NHTSA
uses the MY 2015 existing fleet as its analysis fleet in "Method A" and EPA continues to use the
MY 2014 fleet as its analysis fleet. In order to estimate the impacts of potential standards, it is
necessary to estimate the composition of the future vehicle fleet. Doing so enables estimation of
the extent to which each manufacturer may need to add technology in response to a given series
of attribute-based standards, accounting for the mix and fuel consumption of vehicles in each
manufacturer's regulated fleet. The agencies created analysis fleets in order to track the volumes
and types of fuel economy-improving and CO2 -reducing technologies that are already present in
the existing vehicle fleet. This aspect of the analysis fleets helps to keep the CAFE model from
adding technologies to vehicles that already have these technologies, which will result in "double
counting" of technologies' costs and benefits. An additional step involved projecting the fleets'
sales into MYs 2019-2030. This represents the fleet volumes that the agencies believe will exist
in MYs 2019-2030. The following presents an overview of the information and methods applied
to develop the analyses fleets, and some basic characteristics of that fleet.
Most of the information about the vehicles that make up the 2014 analysis fleet (used in
the NPRM and Method B of the FRM) and the 2015 analysis fleet (used in Method A of the
FRM) was gathered from the 2014 and 2015 Pre-Model Year Reports submitted to EPA by the
manufacturers under Phase 1 of Fuel Efficiency and GHG Emission Program for Medium- and
Heavy-Duty Trucks, MYs 2014-2018. The major manufacturers of class 2b and class 3 trucks
(Chrysler, Ford and GM) were asked to voluntarily submit updates to their Pre-Model Year
Reports. The agencies used these updated data in constructing the analysis fleet for these
manufacturers. The agencies agreed to treat this information as Confidential Business
Information (CBI) until the publication of the proposed rule. This information can be made
public at this time because by now all MY2014 and MY2015 vehicle models have been
produced, which makes data about them essentially public information.
In addition to information about each vehicle, the agencies need additional information
about the fuel economy-improving/CCh-reducing technologies already on those vehicles in order
to assess how much and which technologies to apply to determine a path toward future
compliance. To correctly account for the cost and effectiveness of adding technologies, it is
necessary to know the technology penetration in the existing vehicle fleet. Otherwise, "double-
counting" of technology could occur. Thus, the agencies augmented this information with data
from public and commercial sources6 that include more complete technology descriptions, e.g.
for specific engines and transmissions.
B e.g., manufacturers' web sites, Wards Automotive.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The resultant analysis fleets are provided in detail atNHTSA's web site, along with all
other inputs to and outputs from both the NPRM and the current analysis. The agencies invited
but did not receive comment on this analysis.
10.1.1.1	Vehicle Redesign Schedules and Platforms
10.1.1.2	Pickup Trucks
Product cadence in the Class 2b and 3 pickup market has historically ranged from 7-9
years between major redesigns. However, due to increasing competitive pressures and consumer
demands the agency anticipates that manufacturers will generally shift to shorter design cycles
resembling those of the light duty market. Pickup truck manufacturers in the Class 2b and 3
segments are shown to adopt redesign cycles of six years, allowing two redesigns prior to the end
of the regulatory period in 2027.
10.1.1.1.1.1	Ford
In the 2b/3 pickup truck market, Ford produces the F250, F350 and F450, currently based
on the P3 platform. These models adopted the Super Duty moniker in 1999, and began using
architecture and product cadence distinct from the F150 light-duty pickup models. The first full
redesign of these models occurred in 2008, with smaller redesigns in 2005 and 2011.
NHTSA estimates that the next major redesign of Ford's 2b/3 products will occur in or
about 2017, trailing Ford's announced update of a redesigned F150 in its light-duty pickup
portfolio, with a more rapid product cadence leading to a subsequent redesign in 2023 and
refreshes in 2020 and 2029.
10.1.1.1.1.2	General Motors
General Motors HD pickup trucks, the Silverado and Sierra HD series, are based on the
GMT910 platform and were introduced as a 2007 model. GM has announced a redesigned HD
pickup for the 2015 model year. NHTSA estimates that, like Ford, GM will adopt an
approximate six-year product cadence in the HD truck market, with redesigns in 2015 and 2021.
10.1.1.1.1.3	Fiat (Ram)
The current Ram HD models, on the D2/DJ platform, are anticipated for a major redesign
in the 2018 model year, and the agencies estimate that the product will adopt a similar, shorter
life cycle of six years, with a subsequent redesign in the 2024 model year.
10.1.1.3	Vans
The 2b/3 van market has changed markedly from five years ago. Ford, Nissan, Ram and
Daimler have adopted vans of "Euro Van" appearance, and in many cases now use smaller
turbocharged gasoline or diesel engines in the place of larger, naturally-aspirated V8s. The 2014
Model Year used in this analysis represents a period where most manufacturers, with the
exception of General Motors, have recently introduced a completely redesigned product after

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
many years. The van segment has historically been one of the slowest to be redesigned of any
product segment, with some products going two decades or more between redesigns.
Due to new entrants in the field and increased competition, the agencies anticipate that
most manufacturers will increase the pace of product redesigns in the van segment, but that they
will continue to trail other segments. The cycle time used in this analysis is approximately ten
years between major redesigns, allowing manufacturers' only one major redesign during the
regulatory period.
10.1.1.1.1.4 General Motors
The GM Savana/Chevrolet Express, built on the GMT600 platform, has been produced
since 1996 with a facelift in 2003. The van is currently due for a redesign, and while it is
unknown when this will occur, the agencies anticipate a major redesign due to strong
competitive pressure from other manufacturers will occur in or about 2017, with no further
redesigns occurring until after 2025.
10.1.1.1.1.5Ford
2014 marks the first year in more than three decades that Ford has used a completely new
platform for its vans. The Transit replaces the Econoline except in Chassis Cab or cutaway
configurations. The agencies anticipate that Ford will gradually shift production volume to the
Transit, and will not redesign the Transit until 2025, with one intermediate product freshening.
10.1.1.1.1.6Fiat (Ram)
The product cycle of the van from the Ram brand has less of a historical precedent. Fiat
currently offers the Promaster (a variant of the Ducato van sold in other markets). Previously
Chrysler sold Sprinter vans in an agreement with Daimler from 2003, and had previously
manufactured its own full-sized van.
The Promaster has just been introduced to the US market, and the agencies anticipate that
Fiat will offer a refreshed version in 2020 prior to a full redesign in 2025.
10.1.1.1.1.7Nissan
The Nissan NV launched for the 2012 model using the F-alpha platform shared with the
light-duty Nissan Titan pickup truck. Trade publications and internet sources suggest the next-
generation Nissan Titan could debut in model year 2016, and the agencies anticipate that the NV
van may adopt some of the features and components of the Titan for a mid-cycle freshening of
the NV, with a full redesign in 2021.
10.1.1.1.1.8 Daimler
Daimler introduced its current Sprinter van for the 2007 model year on the NCV3
platform. U.S. models received an update across 2014 and 2015, with rear wheel drive models

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
arriving one year ahead of AWD models. The agencies anticipate that Daimler will redesign the
Sprinter for 2017 with a subsequent freshening in model year 2021.
10.1.1.4 Sales Volume Forecast
Since each manufacturer's required average fuel consumption and GHG levels are sales-
weighted averages of the fuel economy/GHG targets across all model offerings, sales volumes
play a critical role in estimating that burden. The CAFE model requires a forecast of sales
volumes, at the vehicle model-variant level, in order to simulate the technology application
necessary for a manufacturer to achieve compliance in each model year for which outcomes are
simulated.
For this analysis, the agencies relied on the pre-model-year compliance submissions from
manufacturers to provide sales volumes at the model level based on the level of disaggregation in
which the models appear in the compliance data. However, the agencies only use these reported
volumes without adjustment for the reference fleet model year (MY 2014 or MY 2015). For all
future model years, we combine the manufacturer submissions with sales projections from the
2014 (for the NPRM and Method B of the FRM) or 2015 (for Method A of the FRM) Annual
Energy Outlook Reference Case and IHS Automotive to determine model variant level sales
volumes in future years.c The projected sales volumes by class that appear in the Annual Energy
Outlook as a result of a collection of assumptions about economic conditions, demand for
commercial miles traveled, and technology migration from light-duty pickup trucks in response
to the concurrent light-duty CAFE/GHG standards. These are shown in Chapter 2 of the RIA.
The projection of total sales volumes for the Class 2b and 3 market segment was based on
the total volumes in the 2014 AEO Reference Case in the NPRM and for Method B of this FRM.
For the purposes of this analysis, the AEO2014 calendar year volumes have been used to
represent the corresponding model-year volumes. While AEO2014 provides enough resolution
in its projections to separate the volumes for the Class 2b and 3 segments, the agencies deferred
to the vehicle manufacturers and chose to rely on the relative shares present in the pre-model-
year compliance data. This methodology remains the same for the Method A FRM analysis, but
we have replaced the 2014 AEO reference case with the 2015 AEO reference case. A
description of key characteristics of the 2014 and 2015 analysis fleets follows.
10.1.1.4.1 Summary of the 2014 Analysis Fleet
The projection of total sales volumes for the Class 2b and 3 market segment was based on
the total volumes in the 2014 AEO Reference Case. For the purposes of the Method B analysis,
the AEO2014 calendar year volumes have been used to represent the corresponding model-year
volumes. While AEO2014 provides enough resolution in its projections to separate the volumes
for the Class 2b and 3 segments (see Figure 10-1), the agencies deferred to the vehicle
manufacturers and chose to rely on the relative shares present in the pre-model-year compliance
data.
c Tables from AEO's forecast are available at http://www3.eia.gov/oiaf/aeo/tablebrowser/. The agencies also made
use of the IHS Automotive Light Vehicle Production Forecast (August 2014).

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
900
^ ^ ^ ^ ^ ^ ^ ^ ^
	2b 	3 — Total
Figure 10-1 AEO2014 Sales Projections for 2b/3 Vehicles
The relative sales share by vehicle type (van or pickup truck, in this case) was derived
from a sales forecast that the agencies purchased from IHS Automotive, and applied to the total
volumes in the AEO2014 projection. Table 10-1 shows the implied shares of the total new 2b/3
vehicle market broken down by manufacturer and vehicle type.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-1 2014 IHS Automotive Market Share Forecast for 2b/3 Vehicles


MODEL YEAR MARKET SHARE
Manufacturer
Style
2015
2016
2017
2018
2019
2020
2021
Daimler
Van
3%
3%
3%
3%
3%
3%
3%
Fiat
Van
2%
2%
2%
2%
2%
2%
3%
Ford
Van
16%
17%
17%
17%
18%
18%
18%
General Motors
Van
12%
12%
11%
12%
13%
13%
13%
Nissan
Van
2%
2%
2%
2%
2%
2%
2%









Daimler
Pickup
0%
0%
0%
0%
0%
0%
0%
Fiat
Pickup
14%
14%
14%
14%
11%
12%
12%
Ford
Pickup
28%
27%
30%
30%
30%
27%
26%
General Motors
Pickup
23%
23%
21%
21%
21%
22%
23%
Nissan
Pickup
0%
0%
0%
0%
0%
0%
0%
Within those broadly defined market shares, volumes at the manufacturer/model-variant
level were constructed by applying the model-variant's share of manufacturer sales in the pre-
model-year compliance data for the relevant vehicle style, and multiplied by the total volume
estimated for that manufacturer and that style.
After building out a set of initial future sales volumes based on the sources described
above, the agencies attempted to incorporate new information about changes in sales mix that
will not be captured by either the existing sales forecasts or the simulated technology changes in
vehicle platforms. In particular, Ford has announced intentions to phase out their existing
Econoline vans, gradually shifting volumes to the new Transit platform for some model variants
(notably chassis cabs and cutaways variants) and eliminating offerings outright for complete
Econoline vans as early as model year 2015. In the case of complete Econoline vans, the
volumes for those vehicles were allocated to MY2015 Transit vehicles based on assumptions
about likely production splits for the powertrains of the new Transit platform. The volumes for
complete Econoline vans were shifted at ratios of 50 percent, 35 percent, and 15 percent for 3.7
L, 3.5 L Eco-boost, and 3.2 L diesel, respectively. Within each powertrain, sales were allocated
based on the percentage shares present in the pre-model-year compliance data. The chassis cab
and cutaway variants of the Econoline were phased out linearly between MY2015 and MY2020,
at which time the Econolines cease to exist in any form and all corresponding volume resides
with the Transits.
The tables below summarize some of the characteristics of the MY2014 based analysis
fleet for Class 2b and Class 3 trucks. Table 10-2 shows production by manufacturer and
indicates that Ford is dominant with 52 percent of this market.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-2 Estimated MY2014 Production by Manufacturer
MANUFACTURER
PRODUCTION
PERCENT
Daimler
25,327
4.0%
Fiat
138,902
21.8%
Ford
330,919
51.9%
General Motors
129,435
20.3%
Nissan
13,526
2.1%
Total
638,109
100.0%
Table 10-3 shows production by class with 80 percent of production in class 2b, those
trucks with a GVW between 8,501 and 10,000 lbs.
Table 10-3 Estimated MY2014 Production by Class
GVW CLASS
PRODUCTION
PERCENT
2b (8,501-10,000 lbs.)
506,989
79.5%
3 (10,001-14,000 lbs.)
131,120
20.5%
Total
638,109
100.0%
Table 10-4 shows production by style or body type. Pickup trucks make up 52 percent of
production and vans 42 percent of production.
Table 10-4 Estimated MY2014 Production by Vehicle Style
STYLE
PRODUCTION
PERCENT
Chassis Cab
19,724
3.1%
Cutaway
20,539
3.2%
Pickup
333,100
52.2%
Van
264,746
41.5%
Total
638,109
100.0%
Table 10-5 shows production by engine type. Diesel powered trucks make up a
significant share (40 percent) of this market in comparison to light duty vehicles.
Table 10-5 Estimated MY2014 Production by Engine Type
ENGINE TYPE
PRODUCTION
PERCENT
Diesel
252,744
39.6%
Gasoline
105,604
16.5%
FFV
279,761
43.8%
Total
638,109
100.0%
Table 10-6 shows production by drive type with an almost equal division between two
wheel drive (55 percent) and four wheel drive (45 percent).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-6 Estimated MY2014 Production by Drive
DRIVE
PRODUCTION
PERCENT
4WD
286,122
44.8%
FWD
23,309
3.7%
RWD
328,678
51.5%
Total
638,109
100.0%
The following tables show some of the characteristics of the baseline analysis fleet at the
manufacturer level. Table 10-7 and Table 10-8show production by manufacturer for class 2b and
class 3 trucks respectively. As noted above Ford is the dominant manufacturer with 52 percent
of the market in both class 2b and class 3 trucks. While Fiat and General Motors have
comparable shares of the class 2b market (20 percent and 22 percent respectively), Fiat (at 31
percent) has a significantly larger share of the class 3 market than General Motors (at 13
percent).
Table 10-7 Estimated MY2014 Production Class 2b by Manufacturer
MANUFACTURER
PRODUCTION
PERCENT
Daimler
19,556
3.9%
Fiat
98,722
19.5%
Ford
262,687
51.8%
General Motors
112,498
22.2%
Nissan
13,526
2.7%
Total
506,989
100.0%
Table 10-8 Estimated MY2014 Production Class 3 by Manufacturer
MANUFACTURER
PRODUCTION
PERCENT
Daimler
5,771
4.4%
Fiat
40,180
30.6%
Ford
68,232
52.0%
General Motors
16,937
12.9%
Nissan
-
0.0%
Total
131,120
100.0%
As noted above pickup trucks were the dominant body style in Class 2b and 3 trucks.
Table 10-9 shows pickup truck production by manufacturer. Only three manufactures share this
market with Ford the leader at 43 percent, followed by Fiat at 35 percent and General Motors at
22 percent.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-9 Estimated MY2014 Production Pickups by Manufacturer
MANUFACTURER
PRODUCTION
PERCENT
Daimler
-
0.0%
Fiat
115,593
34.7%
Ford
142,580
42.8%
General Motors
74,927
22.5%
Nissan
-
0.0%
Total
333,100
100.0%
All five manufactures share the Class 2b and 3 van market. Table 10-10 shows van
production by manufacturer. Ford is again dominant with 57 percent of the market followed by
General Motors at 21 percent with the remainder divided among Fiat, Daimler and Nissan.
Table 10-10 Estimated MY2014 Production Vans by Manufacturer
MANUFACTURER
PRODUCTION
PERCENT
Daimler
21,900
8.3%
Fiat
23,309
8.8%
Ford
151,503
57.2%
General Motors
54,508
20.6%
Nissan
13,526
5.1%
Total
264,746
100.0%
Table 10-11 and Table 10-12 give an indication of the significance of diesel powered
trucks in the class 2b and 3 market. Table 10-11 shows the distribution of diesel trucks by
manufacturer. Ford is the leader at 40 percent followed by Fiat at 34 percent. Table 10-12 shows
diesel production as a percent of total production for each manufacturer. At either end of the
spectrum are Nissan at 0 percent and Daimler at 100 percent. Of the producers with significant
market share Fiat leads with 62 percent of its production in diesels, followed by General Motors
at 32 percent and Ford at 30 percent.
Table 10-11 Estimated MY2014 Production Diesel Powered Trucks by Manufacturer
MANUFACTURER
PRODUCTION
PERCENT
Daimler
25,327
10.0%
Fiat
86,124
34.1%
Ford
100,208
39.6%
General Motors
41,085
16.3%
Nissan
-
0.0%
Total
252,744
100.0%

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-12 Estimated MY2014 Diesel Penetration by Manufacturer
MANUFACTURER
DIESEL
PRODUCTION
TOTAL
PRODUCTION
PERCENT
DIESEL
Daimler
25,327
25,327
100.0%
Fiat
86,124
138,902
62.0%
Ford
100,208
330,919
30.3%
General Motors
41,085
129,435
31.7%
Nissan
-
13,526
0.0%
Total
252,744
638,109
39.6%
The resultant analysis fleet for Method A (and both Method A and B in the NPRM) is
provided in detail at NHTSA's web site, along with all other inputs to and outputs from Method
A (and NPRM) analysis.
10.1.1.4.2 Summary of the 2015 Analysis Fleet
For Method A, the projection of total sales volumes for the Class 2b and 3 market
segment was based on the total volumes in the 2015 AEO Reference Case. For the purposes of
the Method A analysis, the AEO2015 calendar year volumes have been used to represent the
corresponding model-year volumes. While AEO2015 provides enough resolution in its
projections to separate the volumes for the Class 2b and 3 segments (see Figure 10-2), NHTSA
deferred to the vehicle manufacturers and chose to rely on the relative shares present in the pre-
model-year compliance data.

900

800
„	„

V)
"D
700
£

CO

V)
3
600
O

-C
+•»
500
0)

E
400
3

O

>
300
V)

_0>

(0
200
to


100

0
^ ^ ^ ^ ^
2b ^—3	Total
Figure 10-2 AEO2015 Sales Projections for 2b/3 Vehicles
As with the 2014 analysis fleet, the relative sales share by vehicle type (van or pickup
truck, in this case) was derived from a sales forecast that the agencies purchased from IHS

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Automotive, and applied to the total volumes in the AEO2015 projection. Table 10-13 shows
the implied shares of the total new 2b/3 vehicle market broken down by manufacturer and
vehicle type.
Table 10-13 2015 IHS Automotive Market Share Forecast for 2b/3 Vehicles

MODEL YEAR MARKET SHARE
Manufacturer
Style
2016
2017
2018
2019
2020
2021
Daimler
Van
2%
2%
2%
3%
3%
3%
Fiat
Van
3%
3%
3%
3%
3%
3%
Ford
Van
16%
16%
16%
17%
18%
19%
General Motors
Van
7%
7%
7%
7%
8%
8%
Nissan
Van
1%
1%
1%
1%
2%
2%








Daimler
Pickup
0%
0%
0%
0%
0%
0%
Fiat
Pickup
14%
14%
14%
14%
15%
14%
Ford
Pickup
29%
30%
31%
31%
28%
28%
General Motors
Pickup
28%
27%
26%
25%
24%
24%
Nissan
Pickup
0%
0%
0%
0%
0%
0%
Within those broadly defined market shares, volumes at the manufacturer/model-variant
level were constructed by applying the model-variant's share of manufacturer sales in the pre-
model-year compliance data for the relevant vehicle style, and multiplied by the total volume
estimated for that manufacturer and that style.
The tables below summarize some of the characteristics of the MY2015 based analysis
fleet for Class 2b and Class 3 trucks.
Table 10-14 shows production by manufacturer and indicates that Ford is dominant with
45 percent of this market.
Table 10-14 Estimated MY2015 Production by Manufacturer
MANUFACTURER
PRODUCTION
PERCENT
Daimler
20,828
2.8%
Fiat
126,916
16.9%
Ford
334,859
44.6%
General Motors
254,852
34.0%
Nissan
12,728
1.7%
Total
750,183
100.0%
Table 10-15 shows production by class with 74 percent of production in class 2b, those
trucks with a GVW between 8,501 and 10,000 lbs.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-15 Estimated MY2015 Production by Class
GVW CLASS
PRODUCTION
PERCENT
2b (8,501-10,000 lbs.)
555,415
74.0%
3 (10,001-14,000 lbs.)
194,768
26.0%
Total
750,183
100.0%
Table 10-16 shows production by engine type. Diesel powered trucks make up a
significant share (46 percent) of this market in comparison to light duty vehicles.
Table 10-16 Estimated MY2015 Production by Engine Type
ENGINE TYPE
PRODUCTION
PERCENT
Diesel
342,376
45.6%
Gasoline
160,018
21.3%
FFV
242,510
32.3%
CNG
5,279
0.8%
Total
750,183
100.0%
Table 10-17 shows production by drive type with more four-wheel drive vehicles (62
percent) than two-wheel drive vehicles (38 percent) in the MY 2015 medium/heavy-duty fleet.
Table 10-17 Estimated MY2015 Production by Drive
DRIVE
PRODUCTION
PERCENT
4WD
467,761
62.4%
FWD
19,863
2.6%
RWD
262,559
35.0%
Total
750,183
100.0%
The following tables show some of the characteristics of the baseline analysis fleet at the
manufacturer level. Table 10-18 Table 10-19show production by manufacturer for class 2b and
class 3 trucks respectively. As noted above Ford is the dominant manufacturer with 41 percent
of the market in the class 2b and 56 percent of the market in the class 3 trucks. While General
Motors trails Ford in the class 3 market with 22 percent of the market, they have almost as much
of the class 2b market (38 percent). Fiat has a similar share as General Motors in the class 3
market (19 percent), but makes up about half as much of the class 2b market (16 percent). Both
Nissan and Daimler play a small part in the class 2b market, and of the two, only Daimler has a
small share in the class 3 market.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-18 Estimated MY2015 Production Class 2b by Manufacturer
MANUFACTURER
PRODUCTION
PERCENT
Daimler
14,110
2.5%
Fiat
89,707
16.2%
Ford
226,725
40.8%
General Motors
212,145
38.2%
Nissan
12,728
2.3%
Total
555,415
100.0%
Table 10-19 Estimated MY2015 Production Class 3 by Manufacturer
MANUFACTURER
PRODUCTION
PERCENT
Daimler
6,718
3.5%
Fiat
37,209
19.1%
Ford
108,134
55.5%
General Motors
42,707
21.9%
Nissan
-
0.0%
Total
194,768
100.0%
As noted above pickup trucks were the dominant body style in Class 2b and 3 trucks.
Table 10-20 shows pickup truck production by manufacturer. Only three manufactures share this
market with Ford the leader at 43 percent, followed by General Motors at 37 percent and Fiat at
19 percent.
Table 10-20 Estimated MY2015 Production Pickups by Manufacturer
MANUFACTURER
PRODUCTION
PERCENT
Daimler
-
0.0%
Fiat
107,053
19.3%
Ford
239,835
43.3%
General Motors
206,772
37.4%
Nissan
-
0.0%
Total
553,660
100.0%
All five manufactures share the Class 2b and 3 van market. Table 10-21 shows van
production by manufacturer. Ford is again dominant with 48 percent of the market followed by
General Motors at 25 percent with the remainder divided among Fiat (10 percent), Daimler (11
percent) and Nissan (7 percent).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-21 Estimated MY2014 Production Vans by Manufacturer
MANUFACTURER
PRODUCTION
PERCENT
Daimler
20,828
10.6%
Fiat
19,863
10.1%
Ford
95,024
48.4%
General Motors
48,080
24.5%
Nissan
12,728
6.5%
Total
196,523
100.0%
Table 10-22 and Table 10-23 give an indication of the significance of diesel powered
trucks in the class 2b and 3 market. Table 10-22 shows the distribution of diesel trucks by
manufacturer. Ford is the leader at 43 percent followed by General Motors at 28 percent and Fiat
at 23 percent. Daimler plays a minor role in the diesel market with 6 percent of the market.
Table 10-22 Estimated MY2014 Production Diesel Powered Heavy-Duty Vehicles by Manufacturer
MANUFACTURER
PRODUCTION
PERCENT
Daimler
20,828
6.1%
Fiat
79,478
23.2%
Ford
147,075
43.0%
General Motors
94,995
27.7%
Nissan
-
0.0%
Total
342,376
100.0%
Table 10-23 shows diesel production as a percent of total production for each
manufacturer. At either end of the spectrum are Nissan at 0 percent and Daimler at 100 percent.
Of the producers with significant market share Fiat leads with 63 percent of its production in
diesels, followed by Ford at 44 percent and General Motors at 37 percent.
Table 10-23 Estimated MY2014 Diesel Penetration by Manufacturer
MANUFACTURER
DIESEL
PRODUCTION
TOTAL
PRODUCTION
PERCENT
DIESEL
Daimler
20,828
20,828
100.0%
Fiat
79,478
126,916
62.6%
Ford
147,075
334,859
43.9%
General Motors
94,995
254,852
37.3%
Nissan
-
12,728
0.0%
Total
342,376
750,183
45.6%
The resultant 2015 analysis fleet used in Method B is provided in detail at NHTSA's web
site, along with all other inputs to and outputs from the Method B analysis.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
10.1.2 Other Analysis Inputs
In addition to the inputs summarized above, the analysis of potential standards for HD
pickups and vans makes use of a range of other estimates and assumptions specified as inputs to
the CAFE modeling system. Some significant inputs (e.g., estimates of future fuel prices) also
applicable to other MDHD segments are discussed below in Section IX. Others more specific to
the analysis of HD pickups and vans are as follows:
10.1.2.1	Vehicle Survival and Mileage Accumulation
The analysis estimates the travel, fuel consumption, and emissions over the useful lives
of vehicles produced during model years 2014-2030. Doing so requires initial estimates of these
vehicles' survival rates (i.e., shares expected to remain in service) and mileage accumulation
rates (i.e., anticipated annual travel by vehicles remaining in service), both as a function of
vehicle vintage (i.e., age). These estimates are based on an empirical analysis of changes in the
fleet of registered vehicles over time from HIS/Polk data, in the case of survival rates. The
NPRM and Method A of the FRM use data collected as part of the last Vehicle In Use Survey
(the 2002 VIUS) for the mileage accumulation schedule. Method A of the FRM uses mileage
accumulation schedules from 2014 Polk/IHS odometer reading data. The changes to the VMT
schedules for Method A of the current analysis are further described below in the Method A
FRM specific changes.
10.1.2.2	Rebound Effect
Expressed as an elasticity of mileage accumulation with respect to the fuel cost per mile
of operation, the agencies have applied a rebound effect of 10 percent for today's analysis. Other
rebound effects are considered in sensitivity analyses in Sections D and E.
10.1.2.3	On-Road "Gap"
The model was run with a 20 percent adjustment to reflect differences between on-road
and laboratory performance.
10.1.2.4	Fleet Population Profile
Though not reported here, cumulative fuel consumption and CO2 emissions are presented
in the accompanying EIS, and these calculations utilize estimates of the numbers of vehicles
produced in each model year remaining in service in calendar year 2014. The initial age
distribution of the registered vehicle population in 2014 is based on vehicle registration data
acquired by NHTSA from R.L. Polk Company. For Method A, these values were updated to
reflect newer data acquired by NHTSA from Polk.
10.1.2.5	Past Fuel Consumption Levels

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Though not reported here, cumulative fuel consumption and CO2 emissions are presented
in the accompanying EIS, and these calculations require estimates of the performance of vehicles
produced prior to model year 2014. Consistent with AEO 2014, the model was run with the
assumption that gasoline and diesel HD pickups and vans averaged 14.9 mpg and 18.6 mpg,
respectively, with gasoline versions averaging about 48 percent of production. For Method A,
these values were updated to reflect AEO2015, such that gasoline and diesel versions were
projected to average 16.0 mpg and 20.0 mpg, respectively.
10.1.2.6	Long-Term Fuel Consumption Levels
Though not reported here, longer-term estimates of fuel consumption and emissions are
presented in the accompanying EIS. These estimates include calculations involving vehicle
produced after MY 2030 and, consistent with AEO 2014, the model was run with the assumption
that fuel consumption and CO2 emission levels will continue to decline at 0.05 percent annually
(compounded) after MY 2030.
10.1.2.7	Payback Period
To estimate in what sequence and to what degree manufacturers might add fuel-saving
technologies to their respective fleets, the CAFE model iteratively ranks remaining opportunities
(i.e., applications of specific technologies to specific vehicles) in terms of effective cost, primary
components of which are the technology cost and the avoided fuel outlays, attempting to
minimize effective costs incurred.0 Depending on inputs, the model also assumes manufacturers
may improve fuel consumption beyond requirements insofar as doing so will involve
applications of technology at negative effective cost—i.e., technology application for which
buyers' up-front costs are quickly paid back through avoided fuel outlays. This calculation
includes only fuel outlays occurring within a specified payback period. For both Method A and
Method B, a payback period of 6 months was applied for the dynamic baseline case, or
Alternative lb. Thus, for example, a manufacturer already in compliance with standards is
projected to apply a fuel consumption improvement projected to cost $250 (i.e., as a cost that
could be charged to the buyer at normal profit to the manufacturer) and reduce fuel costs by $500
in the first year of vehicle operation. The agencies have conducted the same analysis applying a
payback period of 0 months for the flat baseline case, or Alternative la. For Method A,
Alternative lb is the primary analysis, and Alternative la is one of a range of cases included in
the sensitivity analysis.
10.1.2.8	Civil Penalties
EPCA and EISA require that a manufacturer pay civil penalties if it does not have enough
credits to cover a shortfall with one or both of the light-duty CAFE standards in a model year.
While these provisions do not apply to HD pickups and vans, at this time, the CAFE model will
show civil penalties owed in cases where available technologies and credits are estimated to be
insufficient for a manufacturer to achieve compliance with a standard. These model-reported
estimates have been excluded from this analysis. For Method A, this aspect of the model has
D Volpe CAFE Model, available at http://www.nhtsa.gov/fuel-economy

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
been modified to also exclude from the calculation of "effective cost" used to select among
available options to add specific technologies to specific vehicles.
10.1.2.9 Coefficients for Fatality Calculations
Both the NPRM and the current analysis consider the potential effects on crash safety of
the technologies manufacturers may apply to their vehicles to meet each of the regulatory
alternatives. NHTSA research has shown that vehicle mass reduction affects overall societal
fatalities associated with crashesE and, most relevant to this rule, mass reduction in heavier light-
and medium-duty vehicles has an overall beneficial effect on societal fatalities. Reducing the
mass of a heavier vehicle involved in a crash with another vehicle(s) makes it less likely there
will be fatalities among the occupants of the other vehicles. In addition to the effects of mass
reduction, the analysis anticipates that these standards, by reducing the cost of driving HD
pickups and vans, will lead to increased travel by these vehicles and, therefore, more crashes
involving these vehicles. The Method B analysis considers overall impacts considering both of
these factors, using a methodology similar to NHTSA's analyses for the MYs 2017 - 2025
CAFE and GHG emission standards.
The Method B analysis includes estimates of the extent to which HD pickups and vans
produced during MYs 2014-2030 may be involved in fatal crashes, considering the mass,
survival, and mileage accumulation of these vehicles, taking into account changes in mass and
mileage accumulation under each regulatory alternative. These calculations make use of the
same coefficients applied to light trucks in the MYs 2017-2025 CAFE rulemaking analysis.
Baseline rates of involvement in fatal crashes are 13.03 and 13.24 fatalities per billion miles for
vehicles with initial curb weights above and below 4,594 lbs, respectively. Considering that the
data underlying the corresponding statistical analysis included observations through calendar
year 2010, these rates are reduced by 9.6 percent to account for subsequent impacts of recent
Federal Motor Vehicle Safety Standards (FMVSS) and anticipated behavioral changes (e.g.,
continued increases in seat belt use). For vehicles above 4,594 lbs—i.e., the majority of the HD
pickup and van fleet—mass reduction is estimated to reduce the net incidence of highway
fatalities by 0.34 percent per 100 lbs. of removed curb weight. For the few HD pickups and vans
below 4,594 lbs, mass reduction is estimated to increase the net incidence of highway fatalities
by 0.52 percent per 100 lbs. Consistent with DOT guidance, the social cost of highway fatalities
is estimated using a value of statistical life (VSL) of $9.36m in 2014, increasing thereafter at
1.18 percent annually.
The Method A analysis uses the same methodology as described above, but applies
coefficients that have been updated to reflect more current data, updated statistical analysis by
NHTSA staff, and updated DOT guidance regarding the VSL. Baseline rates of involvement in
fatal crashes are 16.06 and 14.35 fatalities per billion miles for pickups and vans with initial curb
weights above and below 4,947 lbs, respectively. Considering that the data underlying the
corresponding statistical analysis included observations through calendar year 2012, these rates
are reduced by 9.6 percent to account for subsequent impacts of recent Federal Motor Vehicle
Safety Standards (FMVSS) and anticipated behavioral changes (e.g., continued increases in seat
E U.S. DOT/NHTSA, Relationships Between Fatality Risk Mass and Footprint in MY 2000-2007 PC and LTVs, ID:
NHTSA-2010-0131-0336, Posted August 21, 2012.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
belt use). For vehicles above 4,947 lbs—i.e., the majority of the HD pickup and van fleet—mass
reduction is estimated to reduce the net incidence of highway fatalities by 0.72 percent per 100
lbs. of removed curb weight. For HD pickups and vans below 4,947 lbs (accounting for any
applied mass reduction), mass reduction is estimated to reduce the net incidence of highway
fatalities by 0.10 percent per 100 lbs. Consistent with DOT guidance, the social cost of highway
fatalities is estimated using a value of statistical life (VSL) of $9.4m from 2015 forward.
10.1.2.10	Compliance Credit Provisions
Today's analysis accounts for the potential to over comply with standards and thereby
earn compliance credits, applying these credits to ensuring compliance requirements. In doing
so, the agencies treat any unused carried-forward credits as expiring after five model years,
consistent with current and standards. For today's analysis, the agencies are not estimating the
potential to "borrow"—i.e., to carry credits back to past model years.
10.1.2.11	Emission Factors
While CAFE model calculates vehicular CO2 emissions directly on a per-gallon basis
using fuel consumption and fuel properties (density and carbon content), the model calculates
emissions of other pollutants (methane, nitrogen oxides, ozone precursors, carbon monoxide,
sulfur dioxide, particulate matter, and air toxics) on a per-mile basis. In doing so, the Method A
analysis used corresponding emission factors estimated using EPA's MOVES model.F To
estimate emissions (including CO2) from upstream processes involved in producing, distributing,
and delivering fuel, NHTSA has applied emission factors—all specified on a gram per gallon
basis—derived from Argonne National Laboratory's GREET model.0
10.1.2.12	Refueling Time Benefits
To estimate the value of time savings associated with vehicle refueling, the Method A
analysis used estimates that an average refueling event involves refilling 60 percent of the tank's
capacity over the course of 3.5 minutes, at an hourly cost of $27.22.
10.1.2.13	External Costs of Travel
Changes in vehicle travel will entail economic externalities. To estimate these costs, the
Method A analysis used estimates that congestion-, accident-, and noise-related externalities will
total 5.1 0/mi., 2.8 0/mi., and 0.1 0/mi., respectively.
10.1.2.14	Ownership and Operating Costs
Method A results predict that the total cost of vehicle ownership and operation will change not
just due to changes in vehicle price and fuel outlays, but also due to some other costs likely to
vary with vehicle price. To estimate these costs, NHTSA has applied factors of 5.5 percent (of
F EPA MOVES model available at http://www3.epa. gov/otaa/models/moves/index.htm (last accessed Feb 23,
2015).
G GREET (Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation) Model, Argonne National
Laboratory, https://greet.es.anl.gov/.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
price) for taxes and fees, 15.3 percent for financing, 19.2 percent for insurance, 1.9 percent for
relative value loss. The Method A analysis also estimates that average vehicle resale value will
increase by 25 percent of any increase in new vehicle price.
10.1.3 What Technologies Did the Agencies Consider
The agencies considered over 35 vehicle technologies that manufacturers could use to
improve the fuel consumption and reduce CO2 emissions of their vehicles during MYs 2021-
2027. The majority of the technologies described in this section are readily available, well
known and proven in other vehicle sectors, and could be incorporated into vehicles once
production decisions are made. Other technologies considered may not currently be in
production, but are beyond the research phase and under development, and are expected to be in
production in highway vehicles over the next few years. These are technologies that are capable
of achieving significant improvements in fuel economy and reductions in CO2 emissions, at
reasonable costs. The agencies did not consider technologies in the research stage because there
is insufficient time for such technologies to move from research to production during the model
years covered by this final action.
The technologies considered in the agencies' analysis are briefly described below. They
fall into five broad categories: engine technologies, transmission technologies, vehicle
technologies, electrification/accessory technologies, and hybrid technologies.
In this class of trucks and vans, diesel engines are installed in about half of all vehicles.
The buyer's decision to purchase a diesel versus gasoline engine depends on several factors
including initial purchase price, fuel operating costs, durability, towing capability and payload
capacity amongst other reasons. As discussed in above, the agencies generally prefer to set
standards that do not distinguish between fuel types where technological or market-based reasons
do not strongly argue otherwise. However, as with Phase 1, we continue to believe that
fundamental differences between spark ignition and compression ignition engines warrant unique
fuel standards, which is also important in ensuring that our program maintains product choices
available to vehicle buyers. Therefore, we are maintaining separate standards for gasoline and
diesel vehicles. In the context of our technology discussion for heavy-duty pickups and vans, we
are treating gasoline and diesel engines separately so each has a set of baseline technologies. We
discuss performance improvements in terms of changes to those baseline engines. Our cost and
inventory estimates contained elsewhere reflect the current fleet baseline with an appropriate mix
of gasoline and diesel engines. Note that we are not basing these standards on a targeted switch
in the mix of diesel and gasoline vehicles. We believe our standards require similar levels of
technology development and cost for both diesel and gasoline vehicles. Hence the program is
not intended to force, nor discourage, changes in a manufacturer's fleet mix between gasoline
and diesel vehicles.
The following contains a description of technologies the agencies considered as
potentially available in the rule timeframe, and hence, having potential to be part of a compliance
pathway for these vehicles. Additionally, the agencies did not receive any comments indicating
that the technology effectiveness estimates used in the determination of potential reductions in

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
GHGs and fuel consumption are not representative of the expected ranges for expected duty
cycles.
10.1.3.1 Engine Technologies
The agencies reviewed the engine technology estimates used in the 2017-2025 light-duty
rule, the 2014-2018 heavy-duty rule, and the 2015 NHTSA Technology Study. In doing so the
agencies reconsidered all available sources and updated the estimates as appropriate. The
section below describes both diesel and gasoline engine technologies considered for this
program.
10.1.3.1.1	Low Friction Lubricants
One of the most basic methods of reducing fuel consumption in both gasoline and diesel
engines is the use of lower viscosity engine lubricants. More advanced multi-viscosity engine
oils are available today with improved performance in a wider temperature band and with better
lubricating properties. This can be accomplished by changes to the oil base stock (e.g.,
switching engine lubricants from a Group I base oils to lower-friction, lower viscosity Group III
synthetic) and through changes to lubricant additive packages (e.g., friction modifiers and
viscosity improvers). The use of 5W-30 motor oil is now widespread and auto manufacturers are
introducing the use of even lower viscosity oils, such as 5W-20 and 0W-20, to improve cold-
flow properties and reduce cold start friction. However, in some cases, changes to the
crankshaft, rod and main bearings and changes to the mechanical tolerances of engine
components may be required. In all cases, durability testing will be required to ensure that
durability is not compromised. The shift to lower viscosity and lower friction lubricants will also
improve the effectiveness of valvetrain technologies such as cylinder deactivation, which rely on
a minimum oil temperature (viscosity) for operation.
10.1.3.1.2	Engine Friction Reduction
In addition to low friction lubricants, manufacturers can also reduce friction and improve
fuel consumption by improving the design of both diesel and gasoline engine components and
subsystems. Approximately 10 percent of the energy consumed by a vehicle is lost to friction,
and just over half is due to frictional losses within the engine.H Examples include improvements
in low-tension piston rings, piston skirt design, roller cam followers, improved crankshaft design
and bearings, material coatings, material substitution, more optimal thermal management, and
piston and cylinder surface treatments. Additionally, as computer-aided modeling software
continues to improve, more opportunities for evolutionary friction reductions may become
available. All reciprocating and rotating components in the engine are potential candidates for
friction reduction, and minute improvements in several components can add up to a measurable
fuel efficiency improvement.
H "Impact of Friction Reduction Technologies on Fuel Economy," Fenske, G. Presented at the March 2009 Chicago
Chapter Meeting of the ' Society of Tribologists and Lubricated Engineers' Meeting, March 18th, 2009. Available
at: http://www.chicagostle.org/program/2008-
2009/Impact%20of%20Friction%20Reduction%20Technologies%20on%20Fuel%20Economy%20-
%20with%20VGs%20removed.pdf (last accessed July 9, 2009).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
10.1.3.1.3	Engine Parasitic Demand Reduction
In addition to physical engine friction reduction, manufacturers can reduce the
mechanical load on the engine from parasitics, such as oil, fuel, and coolant pumps. The high-
pressure fuel pumps of direct-injection gasoline and diesel engines have particularly high
demand. Example improvements include variable speed or variable displacement water pumps,
variable displacement oil pumps, more efficient high pressure fuel pumps, valvetrain upgrades
and shutting off piston cooling when not needed.
10.1.3.1.4	Coupled Cam Phasing
Valvetrains with coupled (or coordinated) cam phasing can modify the timing of both the
inlet valves and the exhaust valves an equal amount by phasing the camshaft of an overhead
valve engine.1 For overhead valve engines, which have only one camshaft to actuate both inlet
and exhaust valves, couple cam phasing is the only variable valve timing (VVT) implementation
option available and requires only one cam phaser.J We also considered variable valve lift
(VVL), which alters the intake valve lift in order to reduce pumping losses and more efficiently
ingest air.
10.1.3.1.5	Cylinder Deactivation
In conventional spark-ignited engines throttling the airflow controls engine torque output.
At partial loads, efficiency can be improved by using cylinder deactivation instead of throttling.
Cylinder deactivation can improve engine efficiency by disabling or deactivating (usually) half
of the cylinders when the load is less than half of the engine's total torque capability - the valves
are kept closed, and no fuel is injected - as a result, the trapped air within the deactivated
cylinders is simply compressed and expanded as an air spring, with reduced friction and heat
losses. The active cylinders combust at almost double the load required if all of the cylinders
were operating. Pumping losses are significantly reduced as long as the engine is operated in
this "part-cylinder" mode.
Cylinder deactivation control strategy relies on setting maximum manifold absolute
pressures or predicted torque within a range in which it can deactivate the cylinders. Noise and
vibration issues reduce the operating range to which cylinder deactivation is allowed, although
manufacturers are exploring vehicle changes that enable increasing the amount of time that
cylinder deactivation might be suitable. Some manufacturers may choose to adopt active engine
mounts and/or active noise cancellations systems to address Noise Vibration and Harshness
(NVH) concerns and to allow a greater operating range of activation.
1 Although couple cam phasing appears only in the single overhead cam and overhead valve branches of the decision
tree, it is noted that a single phaser with a secondary chain drive would allow couple cam phasing to be applied to
direct overhead cam engines. Since this would potentially be adopted on a limited number of direct overhead cam
engines NHTSA did not include it in that branch of the decision tree.
1 It is also noted that coaxial camshaft developments would allow other variable valve timing options to be applied
to overhead valve engines. However, since they would potentially be adopted on a limited number of overhead valve
engines, NHTSA did not include them in the decision tree.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Cylinder deactivation has seen a recent resurgence thanks to better valvetrain designs and
engine controls. General Motors and Chrysler Group have incorporated cylinder deactivation
across a substantial portion of their V8-powered lineups, including some heavy duty
applications.
10.1.3.1.6	Stoichiometric Gasoline Direct Injection
SGDI engines inject fuel at high pressure directly into the combustion chamber (rather
than the intake port in port fuel injection). SGDI requires changes to the injector design, an
additional high pressure fuel pump, new fuel rails to handle the higher fuel pressures and
changes to the cylinder head and piston crown design. Direct injection of the fuel into the
cylinder improves cooling of the air/fuel charge within the cylinder, which allows for higher
compression ratios and increased thermodynamic efficiency without the onset of combustion
knock. Recent injector design advances, improved electronic engine management systems and
the introduction of multiple injection events per cylinder firing cycle promote better mixing of
the air and fuel, enhance combustion rates, increase residual exhaust gas tolerance and improve
cold start emissions. SGDI engines achieve higher power density and match well with other
technologies, such as boosting and variable valvetrain designs.
Most manufacturers have introduced vehicles with SGDI engines in light duty sectors,
including GM and Ford and have announced their plans to increase dramatically the number of
SGDI engines in their portfolios. SGDI has not been introduction on heavy duty applications at
this time however as these largely dedicated heavy duty engines approach their redesign window,
they are expected to become SGDI engines.
10.1.3.1.7	Turbocharging and Downsizing
The specific power of a naturally aspirated engine is primarily limited by the rate at
which the engine is able to draw air into the combustion chambers. Turbocharging and
supercharging (grouped together here as boosting) are two methods to increase the intake
manifold pressure and cylinder charge-air mass above naturally aspirated levels. Boosting
increases the airflow into the engine, thus increasing the specific power level, and with it the
ability to reduce engine displacement while maintaining performance. This effectively reduces
the pumping losses at lighter loads in comparison to a larger, naturally aspirated engine.
Almost every major manufacturer currently markets a vehicle with some form of
boosting. While boosting has been a common practice for increasing performance for several
decades, turbocharging has considerable potential to improve fuel economy and reduce CO2
emissions when the engine displacement is also reduced. Specific power levels for a boosted
engine often exceed 100 hp/L, compared to average naturally aspirated engine power densities of
roughly 70 hp/L. As a result, engines can be downsized roughly 30 percent or higher while
maintaining similar peak output levels. In the last decade, improvements to turbocharger turbine
and compressor design have improved their reliability and performance across the entire engine
operating range. New variable geometry turbines and ball-bearing center cartridges allow faster
turbocharger spool-up (virtually eliminating the once-common "turbo lag") while maintaining
high flow rates for increased boost at high engine speeds. Low speed torque output has been
dramatically improved for modern turbocharged engines. However, even with turbocharger

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
improvements, maximum engine torque at very low engine speed conditions, for example launch
from standstill, is increased less than at mid and high engine speed conditions. The potential to
downsize engines may be less on vehicles with low displacement to vehicle mass ratios for
example a very small displacement engine in a vehicle with significant curb weight, in order to
provide adequate acceleration from standstill, particularly up grades or at high altitudes.
The use of GDI in combination with turbocharging and charge air cooling reduces the
fuel octane requirements for knock limited combustion enabling the use of higher compression
ratios and boosting pressures. Recently published data with advanced spray-guided injection
systems and more aggressive engine downsizing targeted towards reduced fuel consumption and
CO2 emissions reductions indicate that the potential for reducing CO2 emissions for
turbocharged, downsized GDI engines may be as much as 15 to 30 percent relative to port-fuel-
injected engines.14'15'16'17'18 Confidential manufacturer data suggests an incremental range of fuel
consumption and CO2 emission reduction of 4.8 to 7.5 percent for turbocharging and
downsizing. Other publicly-available sources suggest a fuel consumption and CO2 emission
reduction of 8 to 13 percent compared to current-production naturally-aspirated engines without
friction reduction or other fuel economy technologies: a joint technical paper by Bosch and
Ricardo suggesting fuel economy gain of 8 to 10 percent for downsizing from a 5.7 liter port
injection V8 to a 3.6 liter V6 with direct injection using a wall-guided direct injection system; a
Renault report suggesting a 11.9 percent NEDC fuel consumption gain for downsizing from a 1.4
liter port injection in-line 4-cylinder engine to a 1.0 liter in-line 4-cylinder engine, also with
wall-guided direct injection; and a Robert Bosch paper suggesting a 13 percent NEDC gain for
downsizing to a turbocharged DI engine, again with wall-guided injection. These reported fuel
economy benefits show a wide range depending on the SGDI technology employed.
Note that for this analysis we determined that this technology path is only applicable to
heavy duty applications that have operating conditions more closely associated with light duty
vehicles. This includes vans designed mainly for cargo volume or modest payloads and having
similar GCWR to light duty applications. These vans cannot tow trailers heavier than similar
light duty vehicles and are largely already sharing engines of significantly smaller displacement
and cylinder count compared to heavy duty vehicles designed mainly for trailer towing.
ACEEE commented that 10 percent of pick-ups in the heavy duty sector are candidates
for turbocharging and downsizing if they do not require higher payloads or towing capacity.
Other commenters suggested that downsizing that has occurred in light duty could also occur in
heavy duty. As discussed above, the agencies evaluated turbocharging and downsizing in
vehicles like vans which are not typically designed for extensive trailer towing. When we looked
at pick-ups, we determined that consumers needing a pick-up without higher payload and trailer
towing requirements would migrate to the lower cost light-duty versions which are typically
identical in cabin size and seating as the heavy-duty versions but have less work capability.
Because of this, in the agencies assessment, the heavy-duty pickups retained the high trailer
towing and payload requirements and the corresponding larger engines. AAPC comments
supported this approach as the correct combination of engine to intended use and even provided
in their comments data indicating that turbocharged and downsized engines are more fuel
efficient at lighter loads however under working conditions expected of a heavy-duty pick-up
they are actually less fuel efficient than the larger engines.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
10.1.3.1.8	Cooled Exhaust-Gas Recirculation
Cooled exhaust gas recirculation or Boosted EGR is a combustion concept that involves
utilizing EGR as a charge diluent for controlling combustion temperatures and cooling the EGR
prior to its introduction to the combustion system. Higher exhaust gas residual levels at part load
conditions reduce pumping losses for increased fuel economy. The additional charge dilution
enabled by cooled EGR reduces the incidence of knocking combustion and obviates the need for
fuel enrichment at high engine power. This allows for higher boost pressure and/or compression
ratio and further reduction in engine displacement and both pumping and friction losses while
maintaining performance. Engines of this type use GDI and both dual cam phasing and discrete
variable valve lift. The EGR systems considered in this final rule, consistent with the rule, will
use a dual-loop system with both high and low pressure EGR loops and dual EGR coolers. The
engines will also use single-stage, variable geometry turbocharging with higher intake boost
pressure available across a broader range of engine operation than conventional turbocharged SI
engines. Such a system is estimated to be capable of an additional 3 to 5 percent effectiveness
relative to a turbocharged, downsized GDI engine without cooled-EGR. The agencies have also
considered a more advanced version of such a cooled EGR system that employs very high
combustion pressures by using dual stage turbocharging.
10.1.3.1.9	Lean-burn Combustion
The agencies considered the concept that gasoline engines that are normally
stoichiometric mainly for emission reasons can run lean over a range of operating conditions and
utilize diesel like aftertreatment systems to control NOx. For this analysis, we determined that
the modal operation nature of this technology is currently only beneficial at light loads and will
not be appropriate for a heavy duty application purchase specifically for its high work and load
capacity.
10.1.3.2 Diesel Engine Technologies
Diesel engines have several characteristics that give them superior fuel efficiency
compared to conventional gasoline, spark-ignited engines. Pumping losses are much lower due
to lack of (or greatly reduced) throttling. The diesel combustion cycle operates at a higher
compression ratio, with a very lean air/fuel mixture, and turbocharged light-duty diesels typically
achieve much higher torque levels at lower engine speeds than equivalent-displacement
naturally-aspirated gasoline engines. Additionally, diesel fuel has a higher energy content per
gallon.K However, diesel fuel also has a higher carbon to hydrogen ratio, which increases the
amount of CO2 emitted per gallon of fuel used by approximately 15 percent over a gallon of
gasoline.
Based on confidential business information and the 2010 NAS Report, two major areas of
diesel engine design could be improved during the timeframe of this final rule. These areas
include aftertreatment improvements and a broad range of engine improvements.
K Burning one gallon of diesel fuel produces about 15 percent more carbon dioxide than gasoline due to the higher
density and carbon to hydrogen ratio.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
10.1.3.2.1	Aftertreatment Improvements
The HD diesel pickup and van segment has largely adopted the SCR type of
aftertreatment system to comply with criteria pollutant emission standards. As the experience
base for SCR expands over the next few years, many improvements in this aftertreatment system
such as construction of the catalyst, thermal management, and reductant optimization may result
in a reduction in the amount of fuel used in the process. However, due to uncertainties with
these improvements regarding the extent of current optimization and future criteria emissions
obligations, the agencies are not considering aftertreatment improvements as a fuel-saving
technology in the rulemaking analysis.
10.1.3.2.2	Engine Improvements
Diesel engines in the HD pickup and van segment are expected to have several
improvements in their base design in the 2021-2027 timeframe. These improvements include
items such as improved combustion management, optimal turbocharger design, and improved
thermal management.
10.1.3.3 Transmission Technologies
The agencies have also reviewed the transmission technology estimates used in the 2017-
2015 light-duty and 2014-2018 heavy-duty final rules. In doing so, NHTSA and EPA considered
or reconsidered all available sources including the 2015 NHTSA Technology Study and updated
the estimates as appropriate. The section below describes each of the transmission technologies
considered for this rule.
10.1.3.3.1	Automatic 8-Speed Transmissions
Manufacturers can also choose to replace 6-speed automatic transmissions with 8-speed
automatic transmissions. Additional ratios allow for further optimization of engine operation
over a wider range of conditions, but this is subject to diminishing returns as the number of
speeds increases. As additional gear sets are added, additional weight and friction are introduced
requiring additional countermeasures to offset these losses. Some manufacturers are replacing 6-
speed automatics already, and 7 to 10-speed automatics have entered production.
10.1.3.3.2	High Efficiency Transmission
For this rule, a high efficiency transmission refers to some or all of a suite of incremental
transmission improvement technologies that should be available within the 2019 to 2027
timeframe. The majority of these improvements address mechanical friction within the
transmission. These improvements include but are not limited to: shifting clutch technology
improvements, improved kinematic design, dry sump lubrication systems, more efficient seals,
bearings and clutches (reducing drag), component superfinishing and improved transmission
lubricants.
10.1.3.3.3	Secondary Axle Disconnect

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The ability to disconnect some of the rotating components in the front axle on 4wd
vehicles when the secondary axle is not needed for traction. This will reduce friction and
increase fuel economy.
10.1.3.4 Electrification/Accessory Technologies
10.1.3.4.1	Electrical Power Steering or Electrohydraulic Power Steering
Electric power steering (EPS) or Electrohydraulic power steering (EHPS) provides a
potential reduction in CO2 emissions and fuel consumption over hydraulic power steering
because of reduced overall accessory loads. This eliminates the parasitic losses associated with
belt-driven power steering pumps which consistently draw load from the engine to pump
hydraulic fluid through the steering actuation systems even when the wheels are not being
turned. EPS is an enabler for all vehicle hybridization technologies since it provides power
steering when the engine is off. EPS may be implemented on most vehicles with a standard 12V
system. Some heavier vehicles may require a higher voltage system which may add cost and
complexity.
10.1.3.4.2	Improved Accessories
The accessories on an engine, including the alternator, coolant and oil pumps are
traditionally mechanically-driven. A reduction in CO2 emissions and fuel consumption can be
realized by driving them electrically, and only when needed ("on-demand").
Electric water pumps and electric fans can provide better control of engine cooling. For
example, coolant flow from an electric water pump can be reduced and the radiator fan can be
shut off during engine warm-up or cold ambient temperature conditions which will reduce warm-
up time, reduce warm-up fuel enrichment, and reduce parasitic losses.
Indirect benefit may be obtained by reducing the flow from the water pump electrically
during the engine warm-up period, allowing the engine to heat more rapidly and thereby
reducing the fuel enrichment needed during cold operation and warm-up of the engine. Faster oil
warm-up may also result from better management of the coolant warm-up period. Further benefit
may be obtained when electrification is combined with an improved, higher efficiency engine
alternator used to supply power to the electrified accessories.
Intelligent cooling can more easily be applied to vehicles that do not typically carry
heavy payloads, so larger vehicles with towing capacity present a challenge, as these vehicles
have high cooling fan loads.L However, towing vehicles tend to have large cooling system
capacity and flow scaled to required heat rejection levels when under full load situations such as
towing at GCWR in extreme ambient conditions. During almost all other situations, this design
characteristic may result in unnecessary energy usage for coolant pumping and heat rejection to
the radiator.
L In the CAFE model, improved accessories refers solely to improved engine cooling.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The agencies considered whether to include electric oil pump technology for the
rulemaking. Because it is necessary to operate the oil pump any time the engine is running,
electric oil pump technology has insignificant effect on efficiency. Therefore, the agencies
decided to not include electric oil pump technology.
10.1.3.4.3	Mild Hybrid
Mild hybrid systems offer idle-stop functionality and a limited level of regenerative
braking and power assist. These systems replace the conventional alternator with a belt or crank
driven starter/alternator and may add high voltage electrical accessories (which may include
electric power steering and an auxiliary automatic transmission pump). The limited electrical
requirements of these systems allow the use of lead-acid batteries or supercapacitors for energy
storage, or the use of a small lithium-ion battery pack.
10.1.3.4.4	Strong Hybrid
A hybrid vehicle is a vehicle that combines two significant sources of propulsion energy,
where one uses a consumable fuel (like gasoline), and one is rechargeable (during operation, or
by another energy source). Hybrid technology is well established in the U.S. light-duty market
and more manufacturers are adding hybrid models to their lineups. Hybrids reduce fuel
consumption through three major mechanisms:
•	The internal combustion engine can be optimized (through downsizing, modifying the
operating cycle, or other control techniques) to operate at or near its most efficient
point more of the time. Power loss from engine downsizing can be mitigated by
employing power assist from the secondary power source.
•	A significant amount of the energy normally lost as heat while braking can be
captured and stored in the energy storage system for later use.
•	The engine is turned off when it is not needed, such as when the vehicle is coasting or
when stopped.
Hybrid vehicles utilize some combination of the three above mechanisms to reduce fuel
consumption and CO2 emissions. The effectiveness of fuel consumption and CO2 reduction
depends on the utilization of the above mechanisms and how aggressively they are pursued. One
area where this variation is particularly prevalent is in the choice of engine size and its effect on
balancing fuel economy and performance. Some manufacturers choose not to downsize the
engine when applying hybrid technologies. In these cases, overall performance (acceleration) is
typically improved beyond the conventional engine. However, fuel efficiency improves less than
if the engine was downsized to maintain the same performance as the conventional version. The
non-downsizing approach is used for vehicles like trucks where towing and/or hauling are an
integral part of their performance requirements. In these cases, if the engine is downsized, the
battery can be quickly drained during a long hill climb with a heavy load, leaving only a
downsized engine to carry the entire load. Because towing capability is currently a heavily-
marketed truck attribute, manufacturers are hesitant to offer a truck with downsized engine

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
which can lead to a significantly diminished towing performance when the battery state of charge
level is low, and therefore engines are traditionally not downsized for these vehicles.
Strong Hybrid technology utilizes an axial electric motor connected to the transmission
input shaft and connected to the engine crankshaft through a clutch. The axial motor is a
motor/generator that can provide sufficient torque for launch assist, all electric operation, and the
ability to recover significant levels of braking energy.
10.1.3.4.5 Air Conditioning Systems
These technologies include improved hoses, connectors and seats for leakage control.
They also include improved compressors, expansion valves, heat exchangers and the control of
these components for the purposes of improving tailpipe CO2 emissions as a result of AJC useM.
10.1.3.5 Vehicle Technologies
10.1.3.5.1 Mass Reduction
Mass reduction is a technology that can be used in a manufacturer's strategy to meet the
Heavy Duty Greenhouse Gas Phase 2 standards. Vehicle mass reduction (also referred to as
"down-weighting" or 'light-weighting"), decreases fuel consumption and GHG emissions by
reducing the energy demand needed to overcome inertia forces, and rolling
resistance. Automotive companies have worked with mass reduction technologies for many
years and a lot of these technologies have been used in production vehicles. The weight savings
achieved by adopting mass reduction technologies offset weight gains due to increased vehicle
size, larger powertrains, and increased feature content (sound insulation, entertainment systems,
improved climate control, panoramic roof, etc.). Sometimes mass reduction has been used to
increase vehicle towing and payload capabilities.
Manufacturers employ a systematic approach to mass reduction, where the net mass
reduction is the addition of a direct component or system mass reduction, also referred to as
primary mass reduction, plus the additional mass reduction taken from indirect ancillary systems
and components, also referred to as secondary mass reduction or mass compounding. There are
more secondary mass reductions achievable for light-duty vehicles compared to heavy-duty
vehicles, which are limited due to the higher towing and payload requirements for these vehicles.
Mass reduction can be achieved through a number of approaches, even while maintaining
other vehicle functionalities. As summarized by NAS in its 2011 light duty vehicle report,N
there are two key strategies for primary mass reduction: 1) changing the design to use less
material; 2) substituting lighter materials for heavier materials.
M See RIA Chapter 2.3 for more detailed technology descriptions.
N Committee on the Assessment of Technologies for Improving Light-Duty Vehicle Fuel Economy; National
Research Council, "Assessment of Fuel Economy Technologies for Light-Duty Vehicles," 2011. Available at
http://www.nap.edu/catalog.php?record_id=12924 (last accessed Jun 27, 2012).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The first key strategy of using less material compared to the baseline component can be
achieved by optimizing the design and structure of vehicle components, systems and vehicle
structure. Vehicle manufacturers have long used these continually-improving CAE tools to
optimize vehicle designs. For example, the Future Steel Vehicle (FSV) project0 sponsored by
World Auto Steel used three levels of optimization: topology optimization, low fidelity 3G
(Geometry Grade and Gauge) optimization, and subsystem optimization, to achieve 30 percent
mass reduction in the body structure of a vehicle with a mild steel unibody structure. Using less
material can also be achieved through improving the manufacturing process, such as by using
improved joining technologies and parts consolidation. This method is often used in
combination with applying new materials.
The second key strategy to reduce mass of an assembly or component involves the
substitution of lower density and/or higher strength materials. Material substitution includes
replacing materials, such as mild steel, with higher-strength and advanced steels, aluminum,
magnesium, and composite materials. In practice, material substitution tends to be quite specific
to the manufacturer and situation. Some materials work better than others for particular vehicle
components, and a manufacturer may invest more heavily in adjusting to a particular type of
advanced material, thus complicating its ability to consider others. The agencies recognize that
like any type of mass reduction, material substitution has to be conducted not only with
consideration to maintaining equivalent component strength, but also to maintaining all the other
attributes of that component, system or vehicle, such as crashworthiness, durability, and noise,
vibration and harshness (NVH).
If vehicle mass is reduced sufficiently through application of the two primary strategies
of using less material and material substitution described above, secondary mass reduction
options may become available. Secondary mass reduction is enabled when the load requirements
of a component are reduced as a result of primary mass reduction. If the primary mass reduction
reaches a sufficient level, a manufacturer may use a smaller, lighter, and potentially more
efficient powertrain while maintaining vehicle acceleration performance. If a powertrain is
downsized, a portion of the mass reduction may be attributed to the reduced torque requirement
which results from the lower vehicle mass. The lower torque requirement enables a reduction in
engine displacement, changes to transmission torque converter and gear ratios, and changes to
final drive gear ratio. The reduced powertrain torque enables the downsizing and/or mass
reduction of powertrain components and accompanying reduced rotating mass (e.g., for
transmission, driveshafts/halfshafts, wheels, and tires) without sacrificing powertrain
durability. Likewise, the combined mass reductions of the engine, drivetrain, and body in turn
reduce stresses on the suspension components, steering components, wheels, tires, and brakes,
which can allow further reductions in the mass of these subsystems. Reducing the unsprung
masses such as the brakes, control arms, wheels, and tires further reduce stresses in the
suspension mounting points, which will allow for further optimization and potential mass
reduction. However, pickup trucks have towing and hauling requirements which must be taken
0 SAE World Congress, "Focus B-pillar 'tailor rolled' to 8 different thicknesses," Feb. 24, 2010. Available at
http://www.sae.org/mags/AEI/7695 (last accessed Jun. 10, 2012).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
into account when determining the amount of secondary mass reduction that is possible and so it
is less than that of passenger cars.
In 2015, EPA completed a multi-year study with FEV North America, Inc. on the
lightweighting of a light-duty pickup truck, a 2011 GMC Silverado, titled "Mass Reduction and
Cost Analysis -Light-Duty Pickup Trucks Model Years 2020-2025."p Results contain a cost
curve for various mass reduction percentages with the main solution being evaluated for a 20.8
percent (510 kg/1122 lb.) mass reduction resulting in an increased direct incremental
manufacturing cost of $2228. In addition, the report outlines the compounding effect that occurs
in a vehicle with performance requirements including hauling and towing. Secondary mass
evaluation was performed on a component level based on an overall 20 percent vehicle mass
reduction. Results revealed 84 kg of the 510 kg, or 20 percent of the overall mass reduction,
were from secondary mass reduction. Information on this study is summarized in SAE paper
2015-01-0559. NHTSA has also sponsored an on-going pickup truck lightweighting project.
This project uses a more recent baseline vehicle, a MY 2014 GMC Silverado, and the project
will be finished in 2016. Both projects will be utilized for the light-duty GHG and CAFE
Midterm Evaluation mass reduction baseline characterization and may be used to update
assumptions of mass reduction for HD pickups and vans for the final Phase 2 rulemaking.
In order to determine if technologies identified on light duty trucks are applicable to
heavy-duty pickups, EPA contracted with FEV North America, Inc. to perform a scaling study in
order to evaluate whether the technologies identified for the light-duty truck would be applicable
for a heavy-duty pickup truck. In this study a 2013MY Silverado 2500, a 2007 Mercedes
Sprinter and a 2010 Renault MasterQ were analyzed. A 2013MY Silverado 2500 was purchased
and torn down. The mass reduction results were 18.9 percent mass reduction at a cost of $2372
and focused on aluminum intensive with AHSS frame. The Mercedes Sprinter and Renault
Master analyses were performed based on information from the A2Macl database. The results
were 18.15 percent mass reduction at a cost add of $2,293 for the Mercedes Sprinter and 18.55
percent mass reduction at a cost add of $2293 for the Master.
In September 2015, Ford announced that its MY 2017 F-Series Super duty pickup (F250)
would be manufactured with an aluminum body and overall the truck will be 350 lbs. lighter (5
percent-6 percent) than the current gen truck with steel.^ This is less overall mass reduction
than the resultant lightweighting effort on the MY 2015 F-150, which achieved up to 750 lb
decrease in curb weight (12 percent-13 percent) per vehicle.1 Strategies were employed by Ford
in the F250 to "improve the productivity of the Super Duty." In addition, Ford added several
safety systems (and consequent mass) including cameras, lane departure warning, brake assist,
p "Mass Reduction and Cost Analysis - Light-Duty Pickup Trucks Model Years 2020-2025," FEV, North America,
Inc., April 2015, Document no. EPA-420-R-15-006.
Q "Mass Reduction and Cost Analysis Heavy Duty Pickup Truck and Light Commercial Vans," 2016, EPA-420-D-
16-003.
R http://www.techtimes.com/articles/87961/20150925/ford-s-2017-f-250-super-duty-with-an-aluminum-body-is-the-
toughest-smartest-and-most-capable-super-duty-ever.htm, September 25, 2015.
s https://www.ford.com/trucks/superduty/2017/.
T "2008/9 Blueprint for Sustainability," Ford Motor Company. Available at: http:// www.ford.com/go/sustainability
(last accessed February 8, 2010).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
etc. More details on the F250 will be known once it is released; however, a review of the F150
vehicle aluminum intensive design shows that it has an aluminum cab structure, body panels, and
suspension components, as well as a high strength steel frame and a smaller, lighter and more
efficient engine. The Executive Summary to Ducker Worldwide's 2014 reportu states that the
MY 2015 F-150 contains 1080 lbs. of aluminum with at least half being aluminum sheet and
extrusions for body and closures. Ford's engine range for its light duty truck fleet includes a
2.7L EcoBoost V-6. The integrated loop, between Ford and the aluminum sheet suppliers, of
aluminum manufacturing scrap and new aluminum sheet is integral to making aluminum a
feasible lightweighting technology option for Ford. It is also possible that the strategy of
aluminum body panels will be applied to the heavy duty F-350 version when it is redesigned.v
The RIA for this rulemaking shows that 10 percent or less mass reduction is part of the
projected strategy for compliance for HD pickups and vans. The cost and effectiveness
assumptions for mass reduction technology are described in the RIA.
10.1.3.5.2	Low Rolling Resistance Tires
Tire rolling resistance is the frictional loss associated mainly with the energy dissipated
in the deformation of the tires under load and thus influences fuel efficiency and CO2
emissions. Other tire design characteristics (e.g., materials, construction, and tread design)
influence durability, traction (both wet and dry grip), vehicle handling, and ride comfort in
addition to rolling resistance. A typical LRR tire's attributes will include: increased tire
inflation pressure, material changes, and tire construction with less hysteresis, geometry changes
(e.g., reduced aspect ratios), and reduction in sidewall and tread deflection. These changes will
generally be accompanied with additional changes to suspension tuning and/or suspension
design.
10.1.3.5.3	Aerodynamic Drag Reduction
Many factors affect a vehicle's aerodynamic drag and the resulting power required to
move it through the air. While these factors change with air density and the square and cube of
vehicle speed, respectively, the overall drag effect is determined by the product of its frontal area
and drag coefficient, Cd. Reductions in these quantities can therefore reduce fuel consumption
and CO2 emissions. Although frontal areas tend to be relatively similar within a vehicle class
(mostly due to market-competitive size requirements), significant variations in drag coefficient
can be observed. Significant changes to a vehicle's aerodynamic performance may need to be
implemented during a redesign (e.g., changes in vehicle shape). However, shorter-term
aerodynamic reductions, with a somewhat lower effectiveness, may be achieved through the use
of revised exterior components (typically at a model refresh in mid-cycle) and add-on devices
that currently being applied. The latter list will include revised front and rear fascias, modified
u "2015 North American Light Vehicle Aluminum Content Study - Executive Summary," June 2014,
http://www.drivealuminum.org/research-resources/PDF/Research/2014/2014-ducker-report (last accessed February
26, 2015).
v http://www.foxnews.eom/leisure/2014/09/30/ford-confirms-increased-aluminum-use-on-next-gen-super-duty-
pickups/.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
front air dams and rear valances, addition of rear deck lips and underbody panels, and lower
aerodynamic drag exterior mirrors.
10.1.4 How Did the Agencies Determine the Costs and Effectiveness of Each of
These Technologies
Building on the technical analysis underlying the 2017-2025 MY light-duty vehicle rule,
the 2014-2018 MY heavy-duty vehicle rule, and the 2015 NHTSA Technology Study, the
agencies took a fresh look at technology cost and effectiveness values for purposes of this rule.
For costs, the agencies reconsidered both the direct (or "piece") costs and indirect costs of
individual components of technologies. For the direct costs, the agencies followed a bill of
materials (BOM) approach employed by the agencies in the light-duty rule as well as referencing
costs from the 2014-2018 MY heavy-duty vehicle rule and a new cost survey performed by Tetra
Tech in 2014.
For two technologies, stoichiometric gasoline direct injection (SGDI) and turbocharging
with engine downsizing, the agencies relied to the extent possible on the available tear-down
data and scaling methodologies used in EPA's ongoing study with FEV, Incorporated. This
study consists of complete system tear-down to evaluate technologies down to the nuts and bolts
to arrive at very detailed estimates of the costs associated with manufacturing them.w
For the other technologies, considering all sources of information and using the BOM
approach, the agencies worked together intensively to determine component costs for each of the
technologies and build up the costs accordingly. Where estimates differ between sources, we
have used engineering judgment to arrive at what we believe to be the best cost estimate
available today, and explained the basis for that exercise of judgment.
Once costs were determined, they were adjusted to ensure that they were all expressed in
2012 dollars (see Section IX.B.l.e of this Preamble), and indirect costs were accounted for using
a methodology consistent with the new ICM approach developed by EPA and used in the Phase
1 rule, and the 2012-2016 and 2017-2025 light-duty rules. NHTSA and EPA also reconsidered
how costs should be adjusted by modifying or scaling content assumptions to account for
differences across the range of vehicle sizes and functional requirements, and adjusted the
associated material cost impacts to account for the revised content. We present the individual
technology costs used in this analysis in Chapter 2.12 of the Draft RIA.
Regarding estimates for technology effectiveness, the agencies used the estimates from
the 2014 Southwest Research Institute study as a baseline, which was designed specifically to
inform this rulemaking. In addition, the agencies used 2017-2025 light-duty rule as a reference,
and adjusted these estimates as appropriate, taking into account the unique requirement of the
heavy-duty test cycles to test at curb weight plus half payload versus the light-duty requirement
of curb plus 300 lb. The adjustments were made on an individual technology basis by assessing
the specific impact of the added load on each technology when compared to the use of the
w U.S. Environmental Protection Agency, "Draft Report - Light-Duty Technology Cost Analysis Pilot Study,"
Contract No. EP-C-07-069, Work Assignment 1-3, September 3, 2009.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
technology on a light-duty vehicle. The agencies also considered other sources such as the 2010
NAS Report, recent compliance data, and confidential manufacturer estimates of technology
effectiveness. The agencies reviewed effectiveness information from the multiple sources for
each technology and ensured that such effectiveness estimates were based on technology
hardware consistent with the BOM components used to estimate costs. Together, the agencies
compared the multiple estimates and assessed their validity, taking care to ensure that common
BOM definitions and other vehicle attributes such as performance and drivability were taken into
account.
The agencies note that the effectiveness values estimated for the technologies may
represent average values applied to the baseline fleet described earlier, and do not reflect the
potentially limitless spectrum of possible values that could result from adding the technology to
different vehicles. For example, while the agencies have estimated an effectiveness of 0.5
percent for low friction lubricants, each vehicle could have a unique effectiveness estimate
depending on the baseline vehicle's oil viscosity rating. Similarly, the reduction in rolling
resistance (and thus the improvement in fuel efficiency and the reduction in CO2 emissions) due
to the application of LRR tires depends not only on the unique characteristics of the tires
originally on the vehicle, but on the unique characteristics of the tires being applied,
characteristics which must be balanced between fuel efficiency, safety, and performance.
Aerodynamic drag reduction is much the same—it can improve fuel efficiency and reduce CO2
emissions, but it is also highly dependent on vehicle-specific functional objectives. For purposes
of this final rule, the agencies believe that employing average values for technology effectiveness
estimates is an appropriate way of recognizing the potential variation in the specific benefits that
individual manufacturers (and individual vehicles) might obtain from adding a fuel-saving
technology.
The assessment of the technology effectiveness and costs was determined from a
combination of sources. First an assessment was performed by SwRI under contract with the
agencies to determine the effectiveness and costs on several technologies that were generally not
considered in the Phase 1 GHG rule time frame. Some of the technologies were common with
the light-duty assessment but the effectiveness and costs of individual technologies were
appropriately adjusted to match the expected effectiveness and costs when implemented in a
heavy-duty application. Finally, the agencies performed extensive outreach to suppliers of
engine, transmission and vehicle technologies applicable to heavy-duty applications to get
industry input on cost and effectiveness of potential GHG and fuel consumption reducing
technologies. The agencies did not receive comments disputing the expected technology
effectiveness values or costs developed with input from industry.
To achieve the levels of the Phase 2 standards for gasoline and diesel powered heavy-
duty vehicles, a combination of the technologies previously discussed will be required respective
to unique gasoline and diesel technologies and their challenges. Although some of the
technologies may already be implemented in a portion of heavy-duty vehicles, none of the
technologies discussed are considered ubiquitous in the heavy-duty fleet. Also, as will be
expected, the available test data show that some vehicle models will not need the full
complement of available technologies to achieve these standards. Furthermore, many
technologies can be further improved (e.g., aerodynamic improvements) from today's best

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
levels, and so allow for compliance without needing to apply a technology that a manufacturer
might deem less desirable.
Technology costs for HD pickups and vans are shown in Table 10-24. These costs reflect
direct and indirect costs to the vehicle manufacturer for the 2021 model year. See Chapter 2.11
of the RIA for a more complete description of the basis of these costs.
Table 10-24 Technology Costs for HD Pickups/Vans Inclusive of Indirect Cost Markups for MY2021 (2012$)
TECHNOLOGY
GASOLINE
DIESEL
Engine changes to accommodate low friction lubes
$6
$6
Engine friction reduction - level 1
$116
$116
Engine friction reduction - level 2
$254
$254
Dual cam phasing
$183
$183
Cylinder deactivation
$196
N/A
Stoichiometric gasoline direct injection
$451
N/A
Turbo improvements
N/A
$16
Cooled EGR
$373
$373
Turbocharging & downsizing3
$671
N/A
"Right-sized" diesel from larger diesel
N/A
$0
8s automatic transmission (increment to 6s automatic transmission)
t
>457 $457
Improved accessories - level 1
$82
$82
Improved accessories - level 2
$132
$132
Low rolling resistance tires - level 1
$10
$10
Passive aerodynamic improvements (aero 1)
$51
$51
Passive plus Active aerodynamic improvements (aero2)
$230
$230
Electric (or electro/hydraulic) power steering
$151
$151
Mass reduction (10% on a 6500 lb vehicle)
$318
$318
Driveline friction reduction
$139
$139
Stop-start (no regenerative braking)
$539
$539
Mild HEV
$2730
$2730
Strong HEV, without inclusion of any engine changes
$6779
$6779
Note:
a Cost to downsize from a V8 OHC to a V6 OHC engine with twin turbos.
As explained above, the CAFE model works by adding technologies in an incremental
fashion to each particular vehicle in a manufacturer's fleet until that fleet complies with the
imposed standards. It does this by following a predefined set of decision trees whereby the
particular vehicle is placed on the appropriate decision tree and it follows the predefined
progression of technology available on that tree. At each step along the tree, a decision is made
regarding the cost of a given technology relative to what already exists on the vehicle along with
the fuel consumption improvement it provides relative to the fuel consumption at the current
location on the tree, prior to deciding whether to take that next step on the tree or remain in the
current location. Because the model works in this way, the input files must be structured to
provide costs and effectiveness values for each technology relative to whatever technologies
have been added in earlier steps along the tree. Table 10-25presents the cost and effectiveness
values used in the CAFE model input files.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-25 CAFE Model Input Values for Cost & Effectiveness for Given Technologies"
TECHNOLOGY
FC
SAVINGS
INCREMENTAL COST (2012$)A B C
2021
2025
2027
Improved Lubricants and Engine
Friction Reduction
1.60%
24
24
23
Coupled Cam Phasing (SOHC)
3.82%
48
43
39
Dual Variable Valve Lift (SOHC)
2.47%
42
37
34
Cylinder Deactivation (SOHC)
3.70%
34
30
27
Intake Cam Phasing (DOHC)
0.00%
48
43
39
Dual Cam Phasing (DOHC)
3.82%
46
40
37
Dual Variable Valve Lift (DOHC)
2.47%
42
37
34
Cylinder Deactivation (DOHC)
3.70%
34
30
27
Stoichiometric Gasoline Direct
Injection (OHC)
0.50%
71
61
56
Cylinder Deactivation (OHV)
3.90%
216
188
172
Variable Valve Actuation (OHV)
6.10%
54
47
43
Stoichiometric Gasoline Direct
Injection (OHV)
0.50%
71
61
56
Engine Turbocharging and Downsizing




Small Gasoline Engines
8.00%
518
441
407
Medium Gasoline Engines
8.00%
-12
-62
-44
Large Gasoline Engines
8.00%
623
522
456
Cooled Exhaust Gas Recirculation
3.04%
382
332
303
Cylinder Deactivation on
Turbo/downsized Eng.
1.70%
33
29
26
Lean-Burn Gasoline Direct Injection
4.30%
1,758
1,485
1,282
Improved Diesel Engine Turbocharging
2.51%
22
19
18
Engine Friction & Parasitic Reduction




Small Diesel Engines
3.50%
269
253
213
Medium Diesel Engines
3.50%
345
325
273
Large Diesel Engines
3.50%
421
397
334
Downsizing of Diesel Engines (V6 to I-
4)
11.10%
0
0
0
8-Speed Automatic Transmission
5.00%
482
419
382
Electric Power Steering
1.00%
160
144
130

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Improved Accessories (Level 1)
0.93%
93
83
75
Improved Accessories (Level 2)
0.93%
57
54
46
Stop-Start System
1.10%
612
517
446
Integrated Starter-Generator
3.20%
1,040
969
760
Strong Hybrid Electric Vehicle
17.20%
3,038
2,393
2,133
Mass Reduction (5%)
1.50%
0.28
0.24
0.21
Mass Reduction (additional 5%)
1.50%
0.87
0.75
0.66
Reduced Rolling Resistance Tires
1.10%
10
9
9
Low-Drag Brakes
0.40%
106
102
102
Driveline Friction Reduction
0.50%
153
137
124
Aerodynamic Improvements (10%)
0.70%
58
52
47
Aerodynamic Improvements (add'l
10%)
0.70%
193
182
153
Notes:
a Values for other model years available in CAFE model input files available at NHTSA web site.
b For mass reduction, cost reported on mass basis (per pound of curb weight reduction).
0 The model output has been adjusted to 2013$
d 8 speed automatic transmission costs include costs for high efficiency gearbox and aggressive shift logic whereas
those costs were kept separate in prior analyses.
In addition to the base technology cost and effectiveness inputs described above, the
CAFE model accommodates inputs to adjust accumulated effectiveness under circumstances
when combining multiple technologies could result in underestimation or overestimation of total
incremental effectiveness relative to an "unevolved" baseline vehicle. These so-called synergy
factors may be positive, where the combination of the technologies results in greater
improvement than the additive improvement of each technology, or negative, where the
combination of the technologies is lower than the additive improvement of each technology. The
synergy factors used in the NPRM and Method B of the FRM are described in Table 10-26.
Method A of the FRM uses synergies derived from a simulation project NHTSA undertook with
Autnomie Argonne National Lab. A description of these changes is given Section D(8).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-26 Technology Pair Effectiveness Synergy Factors for HD Pickups and Vans
TECHNOLOGY
PAIR
ADJUSTMENT

TECHNOLOGY
PAIR
ADJUSTMENT
8SPD/CCPS
-4.60%

IATC/CCPS
-1.30%
8SPD/DEACO
-4.60%

IATC/DEACO
-1.30%
8SPD/ICP
-4.60%

IATC/ICP
-1.30%
8SPD/TRBDS1
4.60%

IATC/TRBDS1
1.30%
AER02/SHEV1
1.40%

MR1/CCPS
0.40%
CCPS/IACC1
-0.40%

MR1/DCP
0.40%
CCPS/IACC2
-0.60%

MR1/VVA
0.40%
DCP/IACC1
-0.40%

MR2/ROLL1
-0.10%
DCP/IACC2
-0.60%

MR2/SHEV1
-0.40%
DEACD/IATC
-0.10%

NAUTO/CCPS
-1.70%
DEACO/IACC2
-0.80%

NAUTO/DEACO
-1.70%
DEACO/MHEV
-0.70%

NAUTO/ICP
-1.70%
DEACS/IATC
-0.10%

NAUTO/SAX
-0.40%
DTURB/IATC
1.00%

NAUTO/TRBDS1
1.70%
DTURB/MHEV
-0.60%

ROLL 1/AERO 1
0.10%
DTURB/SHEV1
-1.00%

ROLL1/SHEV1
1.10%
DVVLD/8SPD
-0.60%

R0LL2/AER02
0.20%
DVVLD/IACC2
-0.80%

SHFTOPT/MHEV
-0.30%
DVVLD/IATC
-0.60%

TRBDS1/MHEV
0.80%
DVVLD/MHEV
-0.70%

TRBDS1/SHEV1
-3.30%
DVVLS/8SPD
-0.60%

TRBDS1/VVA
-8.00%
DVVLS/IACC2
-0.80%

TRBDS2/EPS
-0.30%
DVVLS/IATC
-0.50%

TRBDS2/IACC2
-0.30%
DVVLS/MHEV
-0.70%

TRBDS2/NAUTO
-0.50%



VVA/IACC1
-0.40%



VVA/IACC2
-0.60%



VVA/IATC
-0.60%
The CAFE model also accommodates inputs to adjust accumulated incremental costs
under circumstances when the application sequence could result in underestimation or
overestimation of total incremental costs relative to an "unevolved" baseline vehicle. For
today's analysis, the agencies have applied one such adjustment, increasing the cost of medium-
sized gasoline engines by $513 in cases where turbocharging and engine downsizing is applied
with variable valve actuation.
The analysis performed using Method A also applied cost inputs to address some costs
encompassed neither by the agencies' estimates of the direct cost to apply these technologies, nor
by the agencies' methods for "marking up" these costs to arrive at increases in the new vehicle

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
purchase costs. To account for the additional costs that could be incurred if a technology is
applied and then quickly replaced, the CAFE model accommodates inputs specifying a "stranded
capital cost" specific to each technology. For this analysis, the model was run with inputs to
apply about $78 of additional cost (per engine) if gasoline engine turbocharging and downsizing
(separately for each "level" considered) is applied and then immediately replaced, declining
steadily to zero by the tenth model year following initial application of the technology. The
model also accommodates inputs specifying any additional changes owners might incur in
maintenance and post-warranty repair costs. For this analysis, the model was run with inputs
indicating that vehicles equipped with less rolling-resistant tires could incur additional tire
replacement costs equivalent to $21-$23 (depending on model year) in additional costs to
purchase the new vehicle. The agencies did not, however, include inputs specifying any
potential changes repair costs that might accompany application of any of the above
technologies. A sensitivity analysis using Method A, discussed below, includes a case in which
repair costs are estimated using factors consistent with those underlying the indirect cost
multipliers used to markup direct costs for the agencies' central analysis.
10.1.5 Regulatory Alternatives Considered by the Agencies
As discussed above, the model considers regulatory alternatives. The results of regulatory
alternatives are considered relative to a "no action" alternative where existing standards persist,
but no further regulatory action is taken (in this case the MY2018 standards from Phase I are the
last regulatory action taken). The agencies also considered four regulatory alternatives. The
preferred alternative with a standard that increases 2.5 percent in stringency annually for MY's
2021-2027, and three others with annual increases in stringency of: 2.0 percent, 3.5 percent, and
4.0 percent for MY's 2021-2025. For each of the "action alternatives" (i.e., those involving
stringency increases beyond the no-action alternative), the annual stringency increases are
applied as follows: An annual stringency increase of r is applied by multiplying the model year
2020 target functions (identical to those applicable to model year 2018) by 1 - r to define the
model year 2021 target functions, multiplying the model year 2021 target functions by 1 - r to
define the model year 2022 target functions, continuing through 2025 for all alternatives except
for the preferred Alternative 3 which extends through 2027. In summary, the agencies have
considered the following five regulatory alternatives in the CAFE model.
Table 10-27 Considered Regulatory Alternatives
REGULATORY
ALTERNATIVE
ANNUAL STRINGENCY INCR
EASE
2019-2020
2021-2025
2026-2027
1: No Action
None
None
None
2: 2.0%/y
None
2.0%
None
3: 2.5%/y
None
2.5%
2.5%
4: 3.5%/y
None
3.5%
None
5: 4.0%/y
None
4.0%
None

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
10.1.6 NPRM Modifications of the Model
The NPRM analysis (and the current analysis) reflect several changes made to the model
since 2012, when NHTSA used the model to estimate the effects, costs, and benefits of final
CAFE standards for light-duty vehicles produced during MYs 2017-2021, and augural standards
for MYs 2022-2025. Some of these changes specifically enable analysis of potential fuel
consumption standards (and, hence, CO2 emissions standards harmonized with fuel consumption
standards) for heavy-duty pickups and vans; other changes implement more general
improvements to the model. Key changes include the following:
•	Changes to accommodate standards for heavy-duty pickups and vans, including attribute-
based standards involving targets that vary with "work factor."
•	Explicit calculation of test weight, taking into account test weight "bins" and differences
in the definition of test weight for light-duty vehicles (curb weight plus 300 pound) and
heavy-duty pickups and vans (average of GVWR and curb weight).
•	Procedures to estimate increases in payload when curb weight is reduced, increases in
towing capacity if GVWR is reduced, and calculation procedures to correspondingly
update calculated work factors.
•	Expansion of model inputs, procedures, and outputs to accommodate technologies not
included in prior analyses.
•	Changes to the algorithm used to apply technologies, enabling more explicit accounting
for shared vehicle platforms and adoption and "inheritance" of major engine changes.
These changes are reflected in updated model documentation available at NHTSA's web
site, the documentation also providing more information about the model's purpose, scope,
structure, design, inputs, operation, and outputs. The agencies invited but did not receive
comments on the CAFE model used for the NPRM analysis and used in this final rule for the
Method B analysis.
10.1.6.1 Product Cadence
Past comments on the CAFE model have stressed the importance of product cadence—
i.e., the development and periodic redesign and freshening of vehicles—in terms of involving
technical, financial, and other practical constraints on applying new technologies, and NHTSA
has steadily made changes to the model with a view toward accounting for these considerations.
For example, early versions of the model added explicit "carrying forward" of applied
technologies between model years, subsequent versions applied assumptions that most
technologies would be applied when vehicles are freshened or redesigned, and more recent
versions applied assumptions that manufacturers would sometimes apply technology earlier than
"necessary" in order to facilitate compliance with standards in ensuing model years. Thus, for
example, if a manufacturer is expected to redesign many of its products in model years 2018 and
2023, and the standard's stringency increases significantly in model year 2021, the CAFE model
will estimate the potential that the manufacturer will add more technology than necessary for

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
compliance in MY 2018, in order to carry those product changes forward through the next
redesign and contribute to compliance with the MY 2021 standard.
The model also accommodates estimates of overall limits (expressed as "phase-in caps"
in model inputs) on the rates at which manufacturers' may practicably add technology to their
respective fleets. So, for example, even if a manufacturer is expected to redesign half of its
production in MY 2016, if the manufacturer is not already producing any strong hybrid electric
vehicles (SHEVs), a phase-in cap can be specified in order to assume that manufacturer will stop
applying SHEVs in MY 2016 once it has done so to at least 3 percent of its production in that
model year.
After the light-duty rulemaking analysis accompanying the 2012 final rule regarding
post-2016 CAFE standards and related GHG emissions standards, NHTSA staff began work on
CAFE model changes expected to better reflect additional considerations involved with product
planning and cadence. These changes, summarized below, interact with preexisting model
characteristics discussed above.
10.1.6.2 Platforms and Technology
The term "platform" is used loosely in industry, but generally refers to a common
structure shared by a group of vehicle variants. The degree of commonality varies, with some
platform variants exhibiting traditional "badge engineering" where two products are
differentiated by little more than insignias, while other platforms be used to produce a broad
suite of vehicles that bear little outer resemblance to one another.
Given the degree of commonality between variants of a single platform, manufacturers
do not have complete freedom to apply technology to a vehicle: while some technologies (e.g.
low rolling resistance tires) are very nearly "bolt-on" technologies, others involve substantial
changes to the structure and design of the vehicle, and therefore necessarily are constant between
vehicles that share a common platform. NHTSA staff has, therefore, modified the CAFE model
such that all mass reduction and aero technologies are forced to be constant between variants of a
platform. The agencies requested but did not receive comment on the suitability of this
viewpoint, and which technologies can deviate from one platform variant to another.
Within the analysis fleet, each vehicle is associated with a specific platform. As the
CAFE model applies technology, it first defines a platform "leader" as the vehicle variant of a
platform with the highest technology utilization vehicle of mass reduction and aerodynamic
technologies. As the vehicle applies technologies, it effectively harmonizes to the highest
common denominator of the platform. If there is a tie, the CAFE model begins applying
aerodynamic and mass reduction technology to the vehicle with the lowest average sales across
all available model years. If there remains a tie, the model begins by choosing the vehicle with
the highest average MSRP across all available model years. The model follows this formulation
due to previous market trends suggesting that many technologies begin deployment at the high-
end, low-volume end of the market as manufacturers build their confidence and capability in a
technology, and later expand the technology across more mainstream product lines.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
In the HD pickup and van market, there is a relatively small amount of diversity in
platforms produced by manufacturers: typically 1-2 truck platforms and 1-2 van platforms.
However, accounting for platforms will take on greater significance in future analyses involving
the light-duty fleet. The agency requested but did not receive comments on the general use of
platforms within CAFE rulemaking.
10.1.6.3	Engine and Transmission Inheritance
In practice, manufacturers are limited in the number of engines and transmissions that
they produce. Typically a manufacturer produces a number of engines—perhaps six or eight
engines for a large manufacturer—and tunes them for slight variants in output for a variety of car
and truck applications. Manufacturers limit complexity in their engine portfolio for much the
same reason as they limit complexity in vehicle variants: they face engineering manpower
limitations, and supplier, production and service costs that scale with the number of parts
produced.
In previous usage of the CAFE model, engines and transmissions in individual models
were allowed relative freedom in technology application, potentially leading to solutions that
would, if followed, involve unaccounted-for costs associated with increased complexity in the
product portfolio. The lack of a constraint in this area allowed the model to apply different
levels of technology to the engine in each vehicle at the time of redesign or refresh, independent
of what was done to other vehicles using a previously identical engine.
In the current version of the CAFE model, engines and transmissions that are shared
between vehicles must apply the same levels of technology in all technologies dictated by engine
or transmission inheritance. This forced adoption is referred to as "engine inheritance" in the
model documentation.
As with platform-shared technologies, the model first chooses an "engine leader" among
vehicles sharing the same engine. The leader is selected first by the vehicle with the lowest
average sales across all available model years. If there is a tie, the vehicle with the highest
average MSRP across model years is chosen. The model applies the same logic with respect to
the application of transmission changes. As with platforms, this is driven by the concept that
vehicle manufacturers typically deploy new technologies in small numbers prior to deploying
widely across their product lines.
10.1.6.4	Interactions between Regulatory Classes
Like earlier versions, the current CAFE model provides for integrated analysis spanning
different regulatory classes, accounting both for standards that apply separately to different
classes and for interactions between regulatory classes. Light vehicle CAFE standards are
specified separately for passenger cars and light trucks. However, there is considerable sharing
between these two regulatory classes. Some specific engines and transmissions are used in both
passenger cars and light trucks, and some vehicle platforms span these regulatory classes. For
example, some sport-utility vehicles are offered in 2WD versions classified as passenger cars and
4WD versions classified as light trucks. Integrated analysis of manufacturers' passenger car and
light truck fleets provides the ability to account for such sharing and reduce the likelihood of

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
finding solutions that could involve impractical levels of complexity in manufacturers' product
lines. In addition, integrated analysis provides the ability to simulate the potential that
manufactures could earn CAFE credits by over complying with one standard and use those
credits toward compliance with the other standard (i.e., to simulate credit transfers between
regulatory classes).
HD pickups and vans are regulated separately from light-duty vehicles. While
manufacturers cannot transfer credits between light-duty and MDHD classes, there is some
sharing of engineering and technology between light-duty vehicles and HD pickups and vans.
For example, some passenger vans with GVWR over 8,500 lbs. are classified as medium-duty
passenger vehicles (MDPVs) and thus included in manufacturers' light-duty truck fleets, while
cargo vans sharing the same nameplate are classified as HD vans.
10.1.6.5 Phase-In Caps
The CAFE model retains the ability to use phase-in caps (specified in model inputs) as
proxies for a variety of practical restrictions on technology application. Unlike vehicle-specific
restrictions related to redesign, refreshes or platforms/engines, phase-in caps constrain
technology application at the vehicle manufacturer level. They are intended to reflect a
manufacturer's overall resource capacity available for implementing new technologies (such as
engineering and development personnel and financial resources), thereby ensuring that resource
capacity is accounted for in the modeling process.
In previous CAFE rulemakings, redesign/refresh schedules and phase-in caps were the
primary mechanisms to reflect an OEM's limited pool of available resources during the
rulemaking time frame and the years leading up to the rulemaking time frame, especially in years
where many models may be scheduled for refresh or redesign. The newly-introduced
representation platform-, engine-, and transmission-related considerations discussed above
augment the model's preexisting representation of redesign cycles and accommodation of phase-
in caps. Considering these new constraints, inputs for today's analysis de-emphasize reliance on
phase-in caps.
In the NPRM and Method B of the FRM application of the CAFE model, phase-in caps
are used only for the most advanced technologies included in the analysis, i.e., SHEVs and lean-
burn GDI engines, considering that these technologies are most likely to involve implementation
costs and risks not otherwise accounted for in corresponding input estimates of technology cost.
For these two technologies, the agencies have applied caps that begin at 3 percent (i.e., 3 percent
of the manufacturer's production) in MY 2017, increase at 3 percent annually during the ensuing
nine years (reaching 30 percent in the MY 2026), and subsequently increasing at 5 percent
annually for four years (reaching 50 percent in MY 2030). Note that the agencies did not feel
that lean-burn engines were feasible in the timeframe of this rulemaking, so decided to reject any
model runs where they were selected. Due to the cost ineffectiveness of this technology, it was
never chosen. The agencies did not receive comments specifically on this approach for phase-in
caps. The agencies received comments regarding the general feasibility of SHEVs in this market
segment, with some commenters commenting that SHEVs are not feasible for HD pickups and
vans. These comments are discussed in below. While the agencies have retained the above
approach for SHEV phase-in caps, the agencies have conducted a sensitivity analysis setting the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
SHEV caps at zero, showing that the Phase 2 standards are feasible and appropriate without the
use of SHEVs. This sensitivity analysis is described in Section 10.3.1 below.
For Method A of the NPRM the phase-in caps have been set to 100 percent, so that the
model no longer relies on phase-in caps to limit the early-year application of advanced
technologies. This changes is further described in the Method B of the FRM specific section
below.
10.1.6.6	Impact of Vehicle Technology Application Requirements
Compared to prior analyses of light-duty standards, these model changes, along with
characteristics of the HD pickup and van fleet result in some changes in the broad characteristics
of the model's application of technology to manufacturers' fleets. First, since the number of HD
pickup and van platforms in a portfolio is typically small, compliance with standards may appear
especially "lumpy" (compared to previous applications of the CAFE model to the more highly
segmented light-duty fleet), with significant over compliance when widespread redesigns
precede stringency increases, and/or significant application of carried-forward (aka "banked")
credits.
Second, since the use of phase-in caps has been de-emphasized and manufacturer
technology deployment remains tied strongly to estimated product redesign and freshening
schedules, technology penetration rates may jump more quickly as manufacturers apply
technology to high-volume products in their portfolio.
By design, restrictions that enforce commonality of mass reduction and aerodynamic
technologies on variants of a platform, and those that enforce engine inheritance, will result in
fewer vehicle-technology combinations in a manufacturer's future modeled fleet. These
restrictions are expected to more accurately capture the true costs associated with producing and
maintaining a product portfolio.
10.1.6.7	Accounting for Test Weight, Payload, and Towing Capacity
As mentioned above, NHTSA has also revised the CAFE model to explicitly account for
the regulatory "binning" of test weights used to certify light-duty fuel economy and HD pickup
and van fuel consumption for purposes of evaluating fleet-level compliance with fuel economy
and fuel consumption standards. For HD pickups and vans, test weight (TW) is based on
adjusted loaded vehicle weight (ALVW), which is defined as the average of gross vehicle weight
rating (GVWR) and curb weight (CW). TW values are then rounded, resulting in TW "bins":
ALVW < 4,000 lb.: TW rounded to nearest 125 lb.
4000 lb. < ALVW < 5,500 lb.: TW rounded to nearest 250 lb.
ALVW > 5,500 lb.: TW rounded to nearest 500 lb.
This "binning" of TW is relevant to calculation of fuel consumption reductions
accompanying mass reduction. Model inputs for mass reduction (as an applied technology) are
expressed in terms of a percentage reduction of curb weight and an accompanying estimate of

-------
*** E.O. 12866 Review - Revised - Do Not Cite, Quote, or Release During Review **'
the percentage reduction in fuel consumption, setting aside rounding of test weight. Therefore,
to account for rounding of test weight, NHTSA has modified these calculations as follows:
\ j-?f-i	_	w ^^unrounded TW
LroundedTW — 1/1/ X
Where:
ACW = % change in curb weight (from model input),
AF C unrounded tw = % change in fuel consumption (from model input), without TW
rounding,
ATW = % change in test weight (calculated), and
AFC rounded tw = % change in fuel consumption (calculated), with TW rounding.
As a result, some applications of vehicle mass reduction will produce no compliance
benefit at all, in cases where the changes in ALVW are too small to change test weight when
rounding is taken into account. On the other hand, some other applications of vehicle mass
reduction will produce significantly more compliance benefit than when rounding is not taken
into account, in cases where even small changes in ALVW are sufficient to cause vehicles' test
weights to increase by, e.g., 500 lbs. when rounding is accounted for. Model outputs now
include initial and final TW, GVWR, and GCWR (and, as before, CW) for each vehicle model in
each model year. The agencies invited but did not receive comment on how TW is modeled.
In addition, considering that the regulatory alternatives in the agencies' analysis all
involve attribute-based standards in which underlying fuel consumption targets vary with "work
factor" (defined by the agencies as the sum of three quarters of payload, one quarter of towing
capacity, and 500 lb. for vehicles with 4WD), NHTSA has modified the CAFE model to apply
inputs defining shares of curb weight reduction to be "returned" to payload and shares of GVWR
reduction to be returned to towing capacity. The standards' dependence on work factor provides
some incentive to increase payload and towing capacity, both of which are buyer-facing
measures of vehicle utility. In the agencies' judgment, this provides reason to assume that if
vehicle mass is reduced, manufacturers are likely to "return" some of the change to payload
and/or towing capacity. For this analysis, the agencies have applied the following assumptions:
•	GVWR will be reduced by half the amount by which curb weight is reduced. In other
words, 50 percent of the curb weight reduction will be returned to payload
•	GCWR will not be reduced. In other words, 100 percent of any GVWR reduction
will be returned to towing capacity
•	GVWR/CW and GCWR/GVWR will not increase beyond levels observed among the
majority of similar vehicles (or, for outlier vehicles, initial values)

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-28 Ratios for Modifying GVW and GCW as a Function of Mass Reduction

MAXIMUM RATIOS ASSUMED ENABLED
BY MASS REDUCTION
Group
GVWR/CW
GCWR/GVWR
Unibody
1.75
1.50
Gasoline pickups > 13k GVWR
2.00
1.50
Other gasoline pickups
1.75
2.25
Diesel SRW pickups
1.75
2.50
All other
1.75
2.25
The first of two of these inputs are specified along with standards for each regulatory
alternative, and the GVWR/CW and GCWR/GVWR "caps" are specified separately for each
vehicle model in the analysis fleet.
In addition, NHTSA has changed the model to prevent HD pickup and van GVWR from
falling below 8,500 lbs. when mass reduction is applied (because doing so will cause vehicles to
be reclassified as light-duty vehicles), and to treat any additional mass for hybrid electric
vehicles as reducing payload by the same amount (e.g., if adding a strong HEV package to a
vehicle involves a 350 pound penalty, GVWR is assumed to remain unchanged, such that
payload is also reduced by 350 lbs).
The agencies invited but did not receive comment on estimating how changes in vehicle
mass may impact fuel consumption, GVWR, and GCWR.
10.1.7 Subsequent Changes to the CAFE Model (for Method A)
Since issuing the NPRM, NHTSA has made further changes to the CAFE model, in order
to estimate the potential impacts of simultaneous standards for both light-duty vehicles and HD
pickups and vans. Among the updates most relevant to analysis supporting the final standards
for HD pickups and vans, the current model: includes refinements to enable accounting for
platforms, engines, and transmissions sharing between light-duty and HD pickups and vans;
reflects refinements to how models for the first application of new technology are identified
among shared platforms, engines, and transmissions; allows payback period, discount rate,
survival rates, and mileage accumulation schedules to be specified separately for each vehicle
class; makes use of large scale simulation modeling to more accurately account for synergies
among technologies to estimate the fuel consumption impact of different combinations of
technologies; provides the ability to selectively exclude fine payment from the "effective cost"
calculation used to simulation manufacturers' decisions regarding the application of fuel-saving
technologies; and expands the use of forward planning to estimate decisions to use credits that
would otherwise expire. Changes to the CAFE model are discussed at greater length below and
in the CAFE model documentation.
Also since issuing the NPRM, NHTSA has revised many model inputs to reflect
information that has become available since the proposal. Among the updates most relevant to
analysis supporting the final rule, these inputs reflect: an updated vehicle-level market forecast
based on data regarding the 2015 model year fleet and a new commercially-available

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
manufacturer- and segment-level market forecast, and spanning light-duty vehicles and HD
pickups and vans; newer fuel prices and total vehicle production volumes from the Energy
Information Administration's Annual Energy Outlook 2015; a database, based on a large-scale
full vehicle simulation study, of estimates of the effect of thousands of different combinations of
technologies on fuel consumption; and updated mileage accumulation schedules based on a
database of more than 70 million odometer readings.
NHTSA implemented these changes to the CAFE model and accompanying inputs to
support both today's final rule promulgating new fuel consumption standards for HD pickups
and vans and the Draft Technical Assessment Report regarding agency's consideration of CAFE
standards for light duty vehicles for model years 2022-2025. This provided a basis to analyze
the fleets simultaneously, accounting for interactions between the fleets; the draft RIA (p. 10-18)
accompanying the NPRM identified this as a planned improvement for the final rule, and some
stakeholders' comments (e.g., CARB,X UCS,Y and CBDZ) indicated that such interactions
should be accounted for. Implementing the changes at the same time for both actions also
provided means to release mutually consistent analyses intended for publication nearly
concurrently, and for review by many of the same stakeholders (e.g., by manufacturers
producing both light-duty vehicle and HD pickups and vans).
The remainder of this section summarizes changes to the CAFE model and inputs made
subsequent to the NPRM analysis, summarizes results of the updated analysis, and discusses.
10.1.7.1 Interactions between Regulatory Classes
Like earlier versions, the current CAFE model provides for integrated analysis spanning
different regulatory classes, accounting both for standards that apply separately to different
classes and for interactions between regulatory classes. Light vehicle CAFE standards are
specified separately for passenger cars and light trucks. However, there is considerable sharing
between these two regulatory classes. Some specific engines and transmissions are used in both
passenger cars and light trucks, and some vehicle platforms span these regulatory classes. For
example, some sport-utility vehicles are offered in 2WD versions classified as passenger cars and
4WD versions classified as light trucks. Integrated analysis of manufacturers' passenger car and
light truck fleets provides the ability to account for such sharing and reduce the likelihood of
finding solutions that could involve impractical levels of complexity in manufacturers' product
lines. In addition, integrated analysis provides the ability to simulate the potential that
manufactures could earn CAFE credits by over complying with one standard and use those
credits toward compliance with the other standard (i.e., to simulate credit transfers between
regulatory classes).
HD pickups and vans are regulated separately from light-duty vehicles. While
manufacturers cannot transfer credits between light-duty and MDHD classes, there is some
sharing of engineering and technology between light-duty vehicles and HD pickups and vans.
For example, some passenger vans with GVWR over 8,500 pounds are classified as medium-
x CARB, Docket No. NHTSA-2014-0132-0125, at 17-18; 52-53.
Y UCS, Docket No. EPA-HQ-OAR-2014-0827-1329, at pages 23-24
z CBD, Docket No. NHTSA-2014-0132-0101 at pages 8-9.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
duty passenger vehicles (MDPVs) and thus included in manufacturers' light-duty truck fleets,
while cargo vans sharing the same nameplate are classified as HD vans.
Today's analysis uses an overall analysis fleet spanning both the light-duty and HD
pickup and van fleets. As discussed below, doing so shows some technology "spilling over" to
HD pickups and vans due, for example, to the application of technology in response to current
light-duty standards. For most manufacturers, these interactions appear relatively small. For
Nissan, however, they appear considerable, because Nissan's heavy-duty vans use engines also
used in Nissan's light-duty SUVs.
In the NPRM proposing new standards for heavy-duty pickups and vans, NHTSA and
EPA comment on the expansion of the analysis fleet such that the impacts of new HD pickup and
van standards can be estimated within the context of an integrated analysis of light-duty vehicles
and HD pickups and vans, accounting for interactions between the fleets. As mentioned above,
some environmental organizations specifically cited commonalities and overlap between light-
and heavy-duty products.
10.1.7.2 Phase-In Caps
The model also accommodates estimates of overall limits (expressed as "phase-in caps"
in model inputs) on the rates at which manufacturers' may practicably add technology to their
respective fleets. So, for example, even if a manufacturer is expected to redesign half of its
production in MY 2016, if the manufacturer is not already producing any strong hybrid electric
vehicles (SHEVs), a phase-in cap can be specified in order to assume that manufacturer will stop
applying SHEVs in MY 2016 once it has done so to at least 3 percent of its production in that
model year. Today's analysis sets all of these caps at 100 percent, relying on other model
constraints (in particular, the assumption that many technologies are most practicably applied as
part of a vehicle freshening or redesign) to estimate practicable technology application pathways.
The CAFE model retains the ability to use phase-in caps (specified in model inputs) as
proxies for a variety of practical restrictions on technology application. Unlike vehicle-specific
restrictions related to redesign, refreshes or platforms/engines, phase-in caps constrain
technology application at the vehicle manufacturer level. Introduced in the 2006 version of the
CAFE model, they were intended to reflect a manufacturer's overall resource capacity available
for implementing new technologies (such as engineering and development personnel and
financial resources), thereby ensuring that resource capacity is accounted for in the modeling
process.
In previous fuel efficiency rulemakings, redesign/refresh schedules and phase-in caps
were the primary mechanisms to reflect an OEM's limited pool of available resources during the
rulemaking time frame and the years leading up to the rulemaking time frame, especially in years
where many models may be scheduled for refresh or redesign. The newly-introduced
representation platform-, engine-, and transmission-related considerations discussed above
augment the model's preexisting representation of redesign cycles, and as discussed above,
inputs for today's analysis de-emphasize reliance on phase-in caps.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
10.1.7.3	Impact of Vehicle Technology Application Requirements
Compared to prior analyses of light-duty standards, these model changes result in some
changes in the broad characteristics of the model's application of technology to manufacturers'
fleets. Since the use of phase-in caps has been de-emphasized and manufacturer technology
deployment remains tied strongly to estimated product redesign and freshening schedules,
technology penetration rates may jump more quickly as manufacturers apply technology to high-
volume products in their portfolio.
By design, restrictions that enforce commonality of mass reduction and aerodynamic
technologies on variants of a platform, and those that enforce engine inheritance, will result in
fewer vehicle-technology combinations in a manufacturer's future modeled fleet. As explained
in the NPRM proposing new standards for HD pickups and vans, these restrictions are expected
to more accurately capture the true costs associated with producing and maintaining a product
portfolio.
10.1.7.4	Accounting for Credits
The changes discussed above relate specifically to the model's approach to simulating
manufacturers' potential addition of fuel-saving technology in response to fuel efficiency
standards and fuel prices within an explicit product planning context. The model's approach to
simulating compliance decisions also accounts for the potential to earn and use fuel consumption
credits, as provided by EPCA/EISA. Like past versions, the current CAFE model can be used to
simulate credit carry-forward (a.k.a. banking) between model years and transfers between the
passenger car and light truck fleets, but not credit carry-back (a.k.a. borrowing) between model
years or trading between manufacturers. Unlike past versions, the current CAFE model provides
a basis to specify (in model inputs) fuel consumption credits available from model years earlier
than those being simulated explicitly. For example, with today's analysis representing model
years 2015-2032 explicitly, credits specified as being available from model year 2014 are made
available for use through model year 2019 (given the current 5-year limit on carry-forward of
credits).
As discussed in the CAFE model documentation, the model's default logic attempts to
maximize credit carry-forward—that is to "hold on" to credits for as long as possible. Although
the model uses credits before expiry if needed to cover shortfalls when insufficient opportunity
to add technology is available to achieve compliance with a standard, the model will otherwise
carry forward credits until they are about to expire, at which point it will use them before adding
technology. As further discussed in the CAFE model documentation, model inputs can be used
to adjust this logic to shift the use of credits ahead by one or more model years.
The example presented below illustrates how some of aspects of the current model logic
around credits impacts estimation of technology application by a manufacturer within the context
of a specified set of standards, focusing here on the model's estimate of Ford's potential
technology application under the preferred alternative. Overall results for Ford and other
manufacturers are summarized in Section E.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
¦x
X
/	—x—X—>
LO t£>
oo cn
1,400
1,200
1,000
800
600
400
200
0
o^HrMro^TLntDr-vcocno
c\ic\ic\ic\ic\ic\ic\ic\ic\ic\im
oooooooooooooooo
(NC\lc\l(N(NrMrMrMfNfNfNfNfNfNfNfN
V>
m
o
r\J
CD
(D
s_
(J
C
o
u
>
3
tLO
(D
ai
Model Year
•Achieved Fuel Consumption
•Avg Reg-Cost X Required Fuel Consumption
Figure 10-3 Example of a Possible Compliance Strategy for Ford
Several aspects of the estimated achieved and required fuel consumption levels shown
above are notable. First, the characteristics of Ford's fleet as represented in today's analysis fleet
are such that the heavy duty pickup and van fleet falls short of average fuel efficiency standard in
MY's 2023 through 2027. However, they exceed their standard for MY's 2016 through 2022.
The current analysis uses logic that reflect the potential that Ford could use the 5-year carry
forward provision to use fuel efficiency credits earned in MY's 2018 through MY 2022, to cover
the shortfalls for MY's 2023 to 2027. The model assumes Ford will use as many of the MY 2018
expiring credits as necessary to cover the shortfall in MY 2023. For MY 2024 they will use all
available MY 2019 credits before applying any additional MY 2020 credits necessary to cover
the shortfall (in this particular case there are enough MY 2019 credits to cover the shortfall in
MY 2024). This pattern continues for all model years where there is a shortfall—the model
applies the oldest remaining credits first. Even so, today's analysis indicates Ford could be
required to pay civil penalties for noncompliance without the addition of modest fuel savings in
MY 2027. The change to the model which accounts for credits earned prior to MY 2015 is not
illustrated in this example. However, Ford comes in with fuel consumption credits from MY's
prior to MY 2015; if they had come in with an initial shortfall, they could have used these
banked credits to cover, at least a portion, of that shortfall.
As discussed above, these results provide an estimate, based on analysis inputs, of one
way General Motors could add fuel-saving technologies to its products under the preferred
alternative considered here, and are not a prediction of what General Motors would do under this
alternative. In addition, it should be recognized that specific results vary among manufacturers
and among regulatory alternatives (and under different analytical inputs). Still, the example
should serve to illustrate how the ability to model credit banking can impact results.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
10.1.7.5 Integrating Vehicle Simulation Results into the Synergy Values
The CAFE model does not itself evaluate which technologies will be available, nor does
it evaluate how effective or reliable they will be. The technological availability and
effectiveness rather, are predefined inputs to the model based on the agencies' judgements and
not outputs from the model, which is simply a tool for calculating the effects of combining input
assumptions.
In previous versions of the CAFE Model, technology effectiveness values entered into
the model as a single number for each technology (for each of several classes), intended to
represent the incremental improvement in fuel consumption achieved by applying that
technology to a vehicle in a particular class. At a basic level, this implied that successive
application of new vehicle technologies resulted in an improvement in fuel consumption (as a
percentage) that was the product of the individual incremental effectiveness of each technology
applied. Since this construction fails to capture interactive effects - cases where a given
technology either improves or degrades the impact of subsequently applied technologies - the
CAFE Model applied "synergy factors." The synergy factors were defined for a relatively small
number of technology pairs, and were intended to represent the result of physical interactions
among pairs of technologies - attempting to account for situations where 2x2^4.
For a more specific example, for a vehicle with an initial fuel consumption of FCo, if two
technologies are applied, one with an incremental effectiveness of 5 percent, and a second with
an incremental effectiveness of 10 percent, the effectiveness after the application of both
technologies without consideration of synergies could be expressed as follows:
FCo*(l-.05)*(l-.l)
Which is equivalent to:
FCo*(l-.145)
This suggests that the combined effectiveness of the two technologies is 14.5 percent. The
synergy factors aim to correct for cases where fuel consumption improvements are not perfectly
multiplicative, and the combined fuel consumption in the example above is either greater than or
less than 14.5 percent.
For this analysis, the CAFE Model has been modified to accommodate the results of the
large-scale vehicle simulation study conducted by Argonne National Laboratory (described in
more detail in the light-duty TAR). While Autonomie, Argonne's vehicle simulation model,
produces absolute fuel consumption values for each simulation record, the results have been
modified in a way that preserves much of the existing structure of the CAFE Model's
compliance logic, but still faithfully reproduces the totality of the simulation outcomes present in
the database. Fundamentally, the implementation represents a translation of the absolute values
in the simulation database into incremental improvements and a substantially expanded set of
synergy factors.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Since the simulation efforts only included light-duty vehicles, the effectiveness values for
heavy duty were not integrated into the heavy-duty fleet; for future rule-makings NHTSA hopes
to extend the vehicle simulation efforts to include simulations that would be relevant for heavy-
duty pickups and vans. While the effectiveness values for individual technologies remain the
same, the synergies between two or more technologies incorporate information from Autonomie
Argonne's light-duty pickup simulations. While these synergy values are not a perfect
approximation of the interaction of technology applications particular to heavy-duty vehicles, it
is consistent with what we did in the NPRM (where we also used synergy values from light-duty
pickups).
Updating the synergy values to use Argonne's simulation efforts does two things: 1) it
allows that these synergies may occur between more than two technologies, and 2) because the
synergies are multiplicative, rather than additive, it allows for the consideration that the order of
other technology applications matter in determining the incremental percentage improvement
correction of the synergy value. Instead of having one additive incremental percentage synergy
value for a pair of technologies, regardless of the order of technology application between these
pair of technologies, the synergy values are dependent on the initial state and ending point of a
vehicle within the database.
As stated, in the past, synergy values in the Volpe model were represented as pairs.
However, the new values are 7-tuples and there is one for every point in the database. The
synergy factors are based (entirely) on values in the Argonne database, producing one for each
unique technology combination for each technology class, and are calculated as
FCk
Sk~ FC0-U(.l~Xi)
where Sk is the synergy factor for technology combination k, FCo is the fuel consumption
of the reference vehicle (in the database), x; is the fuel consumption improvement of each
technology i represented in technology combination k (where some technologies are present in
combination k, and some are precedent technologies that were applied, incrementally, before
reaching the current state on one of the paths).
In order to incorporate the results of the Argonne database, while still preserving the
basic structure of the CAFE model's technology module, it was necessary to translate the points
in the database into locations on the technology tree. ^ By recognizing that most of the paths on
the technology tree are unrelated, or separable, it is possible to decompose the technology tree
into a small number of paths and branches by technology type. To achieve this level of linearity,
we define technology groups - only one of which is new. They are: engine cam configuration
(CONFIG), engine technologies (ENG), transmission technologies (TRANS), electrification
(ELEC), mass reduction levels (MR), aerodynamic improvements (AERO), and rolling
resistance (ROLL). The combination of technology levels along each of these paths define a
unique technology combination that corresponds to a single point in the database for each
technology class. These technology state definitions are more important for defining synergies
AA Complete details in the technology tree used to develop the synergies for the heavy-duty rule are available in the
light-duty Draft TAR.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
than for determining incremental effectiveness, but the paths are incorporated into both. Again,
because we did not simulate results applicable to the heavy-duty fleet, we did not use the
database to define the incremental technology effectiveness, but only to adjust for the unique
interaction of different combinations of technology.
As an example, a technology state vector describing a vehicle with a SOHC engine,
variable valve timing (only), a 6-speed automatic transmission, a belt-integrated starter
generator, mass reduction (level 1), aerodynamic improvements (level 2), and rolling resistance
(level 1) would be specified as S0HC;VVT;AT6;BISG;MR1;AER02;R0LL1. Once a vehicle is
assigned a technology state (one of the tens of thousands of unique 7-tuples, defined as
CONFIG;ENG;TRANS;ELEC;MR;AERO;ROLL), adding a new technology to the vehicle
simply represents progress from one technology state to another. The vehicle's fuel consumption
is:
FCi = FC0-(l~FCIi)-Sk/S0
where FCi is the fuel consumption resulting from the application of technology z, FCo is
the vehicle's fuel consumption before technology z is applied, FCIi is the incremental fuel
consumption (percentage) improvement associated with technology z, Sk is the synergy factor
associated with the combination, k, of technologies the vehicle technology z is applied, and So
the synergy factor associated with the technology state that produced fuel consumption FCo.
The synergy factor is defined in a way that captures the incremental improvement of moving
between points in the database, where each point is defined uniquely as a 7-tuple describing its
cam configuration, highest engine technology, transmission, electrification type, mass reduction
level, and level of aerodynamic or rolling resistance improvement. For the current heavy-duty
adoption, it is only these synergy values that were used in the current analysis. While, like with
the individual fuel consumption improvements, there is likely not a simple mapping from light-
duty pickups to heavy-duty pickups (size and power matter), the previous synergy values were
also an adoption from light-duty pickups. The integration of the simulation data allows for a
more complete set of synergies that account for the order of technology application and the
interaction of more than two individual technologies.
10.1.7.6 Updating Mileage Accumulation Schedules
In order to develop new mileage accumulation schedules for vehicles regulated under
NHTSA's fuel efficiency and CAFE programs (classes 1-3), NHTSA purchased a data set of
vehicle odometer readings from IHS/Polk (Polk). Polk collects odometer readings from
registered vehicles when they encounter maintenance facilities, state inspection programs, or
interactions with dealerships and OEMs. The (average) odometer readings in the data set
NHTSA purchased are based on over 74 million unique odometer readings across 16 model
years (2000-2015) and vehicle classes present in the data purchase (all registered vehicles less
than 14,000 lbs. GVW).
The Polk data provide a measure of the cumulative lifetime vehicle miles traveled (VMT)
for vehicles, at the time of measurement, aggregated by the following parameters: make, model,
model year, fuel type, drive type, door count, and ownership type (commercial or personal).
Within each of these subcategories they provide the average odometer reading, the number of

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
odometer readings in the sample from which Polk calculated the averages, and the total number
of that subcategory of vehicles in operation. From these NHTSA was able to develop new
estimates of vehicle miles traveled by age as inputs for the CAFE Model.
10.1.7.6.1 Updated Schedules
The new medium-duty van/pickup schedule in Figure 10-4predicts higher annual VMT
for vehicles between ages one through five years, and lower annual VMT for all other vehicle
ages, than the old schedule. Over the first 30-year span, the new schedule predicts that medium-
duty vans/pickups drive 24,249 (9 percent) fewer miles than the old schedule. We predict the
maximum average annual VMT for medium-duty vehicles (23,307 miles) at age two. These
changes to the schedule will have important implications on certain benefits of the standards.
More monetary fuel savings will occur during the first five years of a vehicle's life under the
new schedule, but a decrease in fuel savings will occur overall while using these schedules. For
payback periods shorter than 5 years, the new schedule will show shorter payback periods than
the old schedule. Section 10 of the RIA offers similar figures for light-duty vehicles types. It also
offers further explanation about the shape of the new annual VMT schedule.
Medium-Duty Pickup/Van Milage Accumulation
(Data and Model)
New and Old Medium-Duty Pickups/Vans Schedules
Old Schedule
New Schedule
Figure 10-4 A Comparison of the New and Old Heavy-Duty Van/Pickup Schedules
Figure 10-5 shows that while the maximum share of commercially-owned vehicles
occurs at age one, the registration population-weighted average odometer reading for personally
and commercially owned vehicles are almost identical for this age. However, the share of
commercially-owned vehicles is higher for age two vehicles than all older ages, and there is a
larger spread between the average odometer readings of the two ownership types for this age of
vehicle (while the spread between the average odometer readings for age three is even larger, the
share of commercially-owned vehicles is smaller, and likely counteracts this effect in the
registration population-weighted models). This increase in discrepancy between the average
odometer reading of the ownership types can explain the peak annual VMT at age two.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
o
VMT by Ownership Type for Medium-duty Vans/Pickups

Ownership Type Population Share for Medium-duty Vans/Pickups



CO -

O "
o

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
in the Polk data and the classes in the CAFE model. The only difference between the mapping
for the VMT schedules and the rest of the CAFE model is that we merged the SUV and van body
styles into one class (for reasons described in our discussion of the SUV/van schedule above).
This mapping allowed us to predict the lifetime miles traveled, by the age of a vehicle, for the
categories in the CAFE model.
In estimating the VMT models, we weighted each data point (make/model classification)
by the share of each make/model in the total population of the corresponding CAFE class. This
weighting ensures that the predicted odometer readings, by class and model year, represent each
of vehicle classification among observed vehicles (i.e., the vehicles for which Polk has odometer
readings), based on each vehicles' representation in the registered vehicle population of its class.
Implicit in this weighting scheme, is the assumption that the samples used to calculate each
average odometer reading by make, model, and model year are representative of the total
population of vehicles of that type. Several indicators suggest that this is a reasonable
assumption.
First, the majority of each vehicle make/model is well-represented in the sample.
Histograms and empirical cumulative distribution functions (CDF's) of the ratio of the number
of odometer readings to the total population of those makes/models by each class (Figure 10-6
below), show that for more than 85 percent of make/model combinations, the average odometer
readings are collected for 20 percent or more of the total population. Most make/model
observations have sufficient sample sizes, relative to their representation in the vehicle
population, to produce meaningful average odometer totals at that levelBB
Ratio of the Number Odometer Readings to Population
Cumulative Representation of Odometer Readings
.2	.4	.6	.8
Ratio of Odometer Readings to Population [Representation]
Figure 10-6 Distribution of the Ratio of the Sample Size to the Population Size (By Make/Model/MY)
We also considered whether the representativeness of the odometer sample varies by
vehicle age, since VMT schedules in the CAFE model are specific to each age. To investigate,
we calculated the percentage of vehicle types (by make, model, and model year) that did not
BB
We developed similar figures, stratified by each vehicle class, but these were no more revealing than the figures
for all vehicles.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
have odometer readings. Figure 10-7 shows that all model years, apart from 2015, have
odometer readings for 96 percent or more of the total types of vehicles observed in the fleet.
Population with No Odometer Readings by Model Year
2000	2005	2010	2015
Figure 10-7 Percentage of the Total Vehicle Population with No Odometer Readings across Model
Year
While the preceding discussion supports the coverage of the odometer sample across
makes/models by each model year, it is possible that, for some of those models, an insufficient
number of odometer readings is recorded to create an average that is likely to be representative
of all of those models in operation for a given year. Figure 10-8 below shows the percentage of
all vehicle types for which the number of odometer readings is less than 5 percent of the total
population (for that model). Again, for all model years other than 2015, about 95 percent or
more of vehicles types are represented by at least 5 percent of their population. For this reason,
we included observations from all model years, other than 2015, in the estimation of the new
VMT schedules.
Population with Fewer than 5% of Odometer Readings
2000	2005	2010	2015
Model Year
Figure 10-8 Percentage of Vehicles with Fewer Than 5% of Population in Odometer Readings (By Class)

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
It is possible that the odometer sample is biased. If certain vehicles are over-represented
in the sample of odometer readings relative to the registered vehicle population, a simple
average, or even one weighted by the number of odometer observations will be biased.
However, while weighting by the share of each vehicle in the population will account for this
bias, it would not correct for a sample that entirely omits a large number of makes/models within
a model year. We tested for this by computing the proportion of the count of odometer readings
for each individual vehicle type—within a class and model year—to the total count of readings
for that class and model year. We also compared the population of each make/model—within
each class and model year—to the population of the corresponding class and model year. The
difference of these two ratios shows the difference of the representation of a vehicle type—in its
respective class and model year—in the sample versus the population. All vehicle types are
represented in the sample within 10 percent of their representation in the population, and the
variance between the two representations is normally distributed. This suggests that, on average,
the likelihood that a vehicle is in the sample is comparable to its proportion in the relevant
population, and that there is little under or over sampling of certain vehicle makes/models.cc
Difference in Representation in Sample vs. Population
-.05	0	.05
Difference in Representation
Figure 10-9 Difference in the Share of Each Vehicle in the Population versus the Sample (By Class)
10.1.7.6.3 Es timation
Since model years are sold in in the fall of the previous calendar year, throughout the
same calendar year, and even into the following calendar year—not all registered vehicles of a
make/model/model year will have been registered for at least a year (or more) until age 3. The
result is that some MY2014 vehicles may have been driven for longer than one year, and some
less, at the time the odometer was observed. In order to consider this in our definition of age, we
assign the age of a vehicle to be the difference between the average reading date of a
make/model and the average first registration date of that make/model. The result is that the
continuous age variable reflects the amount of time that a car has been registered at the time of
odometer reading, and presumably the time span that the car has accumulated the miles.
CC
We produced similar figures, stratified by class, but these were no more revealing; the only difference being that
cars are represented in the sample within 5 percent of their representation in the population (with a distribution range
of .05 on either side).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
After creating the "Age" variable, we fit the make/model lifetime VMT data points to a
weighted quartic polynomial regression of the age of the vehicle. The predicted values of the
quartic regressions are used to calculate the marginal annual VMT by age for each class by
calculating differences in estimated lifetime mileage accumulation by age. However, the Polk
data acquired by NHTSA only contains observations for vehicles newer than 16 years of age. In
order to estimate the schedule for vehicles older than the age 15 vehicles in the Polk data, we
combined information about that portion of the schedule from the VMT schedules used in both
the 2017-2021 Final Light Duty Rule and 2019-2025 Medium-Duty NPRM. The light-duty
schedules were derived from the survey data contained in the 2009 National Household Travel
Survey (NHTS) and the 2001 Vehicle in Use Survey (VIUS), for medium-duty trucks.
Based on the vehicle ages for which we have data (from the Polk purchase), the newly
estimated annual schedules differ from the previous version in important ways. Perhaps most
significantly, the annual mileage associated with ages beyond age 8 begin to, and continue to,
trend much lower. The approach taken here attempts to preserve the results obtained through
estimation on the Polk observations, while leveraging the existing (NHTS-based) schedules to
support estimation of the higher ages (age 16 and beyond). Since the two schedules are so far
apart, simply splicing them together would have created not only a discontinuity, but also
precluded the possibility of a monotonically decreasing scale with age (which is consistent with
previous schedules, the data acquired from Polk, and common sense).
From the old schedules, we expect that the annual VMT is decreasing for all ages.
Towards the end of our sample, the predictions for annual VMT increase. In order to force the
expected monotonicity, we perform a triangular smoothing algorithm until the schedule is
monotonic. This performs a weighted average which weights the observations close to the
observation more than those farther from it. The result is a monotonic function, which predicts
similar lifetime VMT for the sample span as the original function. Since we do not have data
beyond 15 years of age, we are not able to correctly capture that part of the annual VMT curve
using only the new dataset. For this reason, we use trends in the old data to extrapolate the new
schedule for ages beyond the sample range.
In order to use the VMT information from the newer data source for ages outside of the
sample, we use the final in-sample age (15 years) as a seed and then apply the proportional trend
from the old schedules to extrapolate the new schedules out to age 30. To do this, we calculated
the annual percentage difference in VMT of the old schedule for ages 15-30. The same annual
percentage difference in VMT is applied to the new schedule to extend beyond the final in-
sample value. This assumes that the overall proportional trend in the outer years is correctly
modeled in the old VMT schedule, and imposes this same trend for the outer years of the new
schedule. The extrapolated schedules are the final input for the VMT schedules in the CAFE
model.
10.1.7.6.4 Comparison to Previous Schedules
The new VMT data suggests that the VMT schedule used in the last Light-Duty CAFE
Final Rule likely does not represent current annual VMT rates. Across all classes, the previous
VMT schedules overestimate the average annual VMT. The previous schedules are based on data
that is outdated and self-reported, while the observations from Polk are between 5 and 7 years

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
newer than those in the NHTS and represent valid odometer readings (rather than self-reported
information).
Additionally, while the NHTS may be a representative sample of households, it is less
likely to be a representative sample of vehicles. However, by properly accounting for vehicle
population weights in the new averages and models, we corrected for this issue in the derivation
of the new schedules.
Insofar as these changes better represent actual VMT, they lead to better estimates of
actual impacts, such as avoided fuel consumption and GHG emissions, safety impacts, and
monetized benefits.
10.1.7.6.5 Future Direction
In consultation with other agencies closely involved with VMT estimation (e.g., FHWA),
NHTSA will continue to seek means to further refine estimated mileage accumulation schedules.
For example, one option under consideration would be to obtain odometer reading data from
successive calendar years, thus providing a more robust basis to consider, for example, the
influence of changing fuel prices or economic conditions on the accumulation of miles by
vehicles of a given age.
10.1.7.7	Updated Analysis Fleet
For the current analysis we updated the reference fleet from MY 2014, to the latest
available MY 2015. The projection of total sales volumes for the Class 2b and 3 market segment
was based on the total volumes in the 2015 AEO Reference Case. For the purposes of this
analysis, the AEO2015 calendar year volumes have been used to represent the corresponding
model-year volumes. While AEO2015 provides enough resolution in its projections to separate
the volumes for the Class 2b and 3 segments, the agencies deferred to the vehicle manufacturers
and chose to rely on the relative shares present in the pre-model-year compliance data.
10.1.7.8	Changes to Costs
10.1.7.8.1 Use of Retail Price Equivalent (RPE) Multiplier to Calculate
Indirect Costs
To produce a unit of output, vehicle manufacturers incur direct and indirect costs. Direct
costs include cost of materials and labor costs. Indirect costs are all the costs associated with
producing the unit of output that are not direct costs - for example, they may be related to
production (such as research and development [R&D]), corporate operations (such as salaries,
pensions, and health care costs for corporate staff), or selling (such as transportation, dealer
support, and marketing). Indirect costs are generally recovered by allocating a share of the costs
to each unit of good sold. Although it is possible to account for direct costs allocated to each
unit of good sold, it is more challenging to account for indirect costs allocated to a unit of goods
sold. To make a cost analysis process more feasible, markup factors, which relate total indirect
costs to total direct costs, have been developed. These factors are often referred to as retail price
equivalent (RPE) multipliers.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Cost analysts and regulatory agencies (including both NHTSA and EPA) have frequently
used these multipliers to predict the resultant impact on costs associated with manufacturers'
responses to regulatory requirements. The best approach, if it were possible, to determining the
impact of changes in direct manufacturing costs on a manufacturer's indirect costs would be to
actually estimate the cost impact on each indirect cost element. However, doing this within the
constraints of an agency's time or budget is not always feasible, and the technical, financial, and
accounting information to carry out such an analysis may simply be unavailable.
The one empirically derived metric that addresses the markup of direct costs to consumer
costs is the RPE multiplier, which is measured from manufacturer 10-K accounting statements
filed with the Securities and Exchange Commission. Over roughly a three decade period, the
measured RPE has been remarkably stable, averaging 1.5, with minor annual variation. The
National Research Council notes that, "Based on available data, a reasonable RPE multiplier
would be 1.5." The historical trend in the RPE is illustrated in Figure 10-10.
RPE History, 1972-1997, and 2007
~ RPE
2000
Figure 10-10 RPE History, 1972-1997, and 2007
RPE multipliers provide, at an aggregate level, the relationship between revenue and
direct manufacturing costs. They are measured by dividing total revenue by direct costs.
However, because this provides only a single aggregate measure, using RPE multipliers results
in the application of a common incremental markup to all technologies. It assures that the
aggregate cost impact across all technologies is consistent with empirical data, but does not
allow for indirect cost discrimination among different technologies. Thus, a concern in using the
RPE multiplier in cost analysis for new technologies added in response to regulatory
requirements is that the indirect costs of vehicle modifications are not likely to be the same for
all different technologies. For example, less complex technologies could require fewer R&D
efforts or less warranty coverage than more complex technologies. In addition, some simple
technological adjustments may, for example, have no effect on the number of corporate
personnel and the indirect costs attributable to those personnel. The use of RPEs, with their
assumption that all technologies have the same proportion of indirect costs, is likely to

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
overestimate the costs of less complex technologies and underestimate the costs of more
complex technologies. However, for regulations such as the CAFE and GHG emission standards
under consideration, which drive changes to nearly every vehicle system, overall average indirect
costs should align with the RPE value. Applying RPE to the cost for each technology assures
that alignment.
Modified multipliers have been developed by EPA, working with a contractor, for use in
rulemakings.1 These multipliers are referred to as indirect cost multipliers (or ICMs). ICMs
assign unique incremental changes to each indirect cost contributor at several different
technology levels.
ICM = (direct cost + adjusted indirect cost)/(direct cost)
Developing the ICMs from the RPE multipliers requires developing adjustment factors
based on the complexity of the technology and the time frame under consideration: the less
complex a technology, the lower its ICM, and the longer the time frame for applying the
technology, the lower the ICM. This methodology was used in the cost estimation for the recent
light-duty MYs 2012-2016 and MYs 2017-2025 rulemaking and for the heavy-duty MYs 2014-
2018 rulemaking. The ICMs for the light-duty context were developed in a peer-reviewed report
from RTI International and were subsequently discussed in a peer-reviewed journal article.2
Importantly, since publication of that peer-reviewed journal article, the agencies have revised the
methodology to include a return on capital (i.e., profits) based on the assumption implicit in
ICMs (and RPEs) that capital costs are proportional to direct costs, and businesses need to be
able to earn returns on their investments.
Since their original development in February 2009, the agencies have made some
changes to both the ICMs factors and to the method of applying those factors relative to the
factors developed by RTI and presented in their reports. We have described and explained those
changes in several rulemakings over the years, most notably the 2017-2025 FR for light vehicles
and the more recent Heavy-duty GHG Phase 2 NPRM.3 In the 2015 NAS study, the committee
stated a conceptual agreement with the ICM method since ICM takes into account design
challenges and the activities required to implement each technology. However, although
endorsing ICMs as a concept, the NAS Committee stated that".. .the empirical basis for such
multipliers is still lacking, and, since their application depends on expert judgment, it is not
possible to determine whether the Agencies' ICMs are accurate or not." NAS also states that
".. .the specific values for the ICMs are critical since they may affect the overall estimates of
costs and benefits for the overall standards and the cost effectiveness of the individual
technologies." The committee did encourage continued research into ICMs given the lack of
empirical data for them to evaluate the ICMs used by the agencies in past analyses. EPA, for its
part, continues to study the issue surrounding ICMs but has not pursued further efforts given
resource constraints and demands in areas such as technology benchmarking and cost teardowns.
On balance, NHTSA believes that the empirically derived RPE is a more reliable basis
for estimating indirect costs. To ensure overall indirect costs in the analysis align with the RPE
value, NHTSA has developed its primary analysis based on applying the RPE value of 1.5 to
each technology. NHTSA also has conducted a sensitivity analysis examining the impact of
applying the ICM approach in the sensitivity analysis portion later in this Section. This marks a

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
change from the NPRM where we use the ICM multiplier to calculate indirect costs as the central
analysis and the RPE multiplier as a sensitivity case.
10.1.7.8.2 Updates to Mass Reduction Based on 2014 Silverado Study
As proposed in the NPRM we have updated the HD pickup and van mass reduction cost
curves with a MY 2014 GMC Silverado ED AG study. The updated mass reduction study
suggests that mass reduction will be more costly for heavy-duty vans and pickups than was
suggested in the NPRM. This can explain the reduction in mass reduction in the current analysis
compared to the NPRM.
NHTSA awarded a contract to ED AG to conduct a vehicle weight reduction feasibility
and cost study of a 2014MY full size pick-up truck. The light weighted version of the full size
pick-up truck (LWT) used manufacturing processes that will likely be available during the model
years 2025-2030 and be capable of high volume production. The goal was to determine the
maximum feasible weight reduction while maintaining the same vehicle functionalities, such as
towing, hauling, performance, noise, vibration, harshness, safety, and crash rating, as the
baseline vehicle, as well as the functionality and capability of designs to meet the needs of
sharing components across same or cross vehicle platform. Consideration was also given to the
sharing of engines and other components with vehicles built on other platforms to achieve
manufacturing economies of scale, and in recognition of resource constraints which limit the
ability to optimize every component for every vehicle.
A comprehensive teardown/benchmarking of the baseline vehicle was conducted for the
engineering analysis. The analysis included geometric optimization of load bearing vehicle
structures, advanced material utilization along with a manufacturing technology assessment that
would be available in the 2017 to 2025 time frame. The baseline vehicle's overall mass, center of
gravity and all key dimensions were determined. Before the vehicle teardown, laboratory
torsional stiffness tests, bending stiffness tests and normal modes of vibration tests were
performed on baseline vehicles so that these results could be compared with the CAE model of
the light weighted design. After conducting a full tear down and benchmarking of the baseline
vehicle, a detailed CAE model of the baseline vehicle was created and correlated with the
available crash test results. The project team then used computer modeling and optimization
techniques to design the light-weighted pickup truck and optimized the vehicle structure
considering redesign of structural geometry, material grade and material gauge to achieve the
maximum amount of mass reduction while achieving comparable vehicle performance as the
baseline vehicle. Only technologies and materials projected to be available for large scale
production and available within two to three design generations (e.g. model years 2020, 2025
and 2030) were chosen for the LWT design. Three design concepts were evaluated: 1) a multi-
material approach; 2) an aluminum intensive approach; and 3) a Carbon Fiber Reinforced
Plastics approach. The multi-material approach was identified as the most cost effective. The
recommended materials (advanced high strength steels, aluminum, magnesium and plastics),
manufacturing processes, (stamping, hot stamping, die casting, extrusions, and roll forming) and
assembly methods (spot welding, laser welding, riveting and adhesive bonding) are currently
used, although some to a lesser degree than others. These technologies can be fully developed
within the normal product design cycle using the current design and development methods.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The design of the LWT was verified, through CAE modeling, that it meets all relevant
crash tests performance. The LS-DYNA finite element software used by the ED AG team is an
industry standard for crash simulation and modeling. The researchers modeled the
crashworthiness of the LWT design using the NCAP Frontal, Lateral Moving Deformable
Barrier, and Lateral Pole tests, along with the IIHS Roof, Lateral Moving Deformable Barrier,
and Frontal Offset (40 percent and 25 percent) tests. All of the modeled tests were comparable
to the actual crash tests performed on the 2014 Silverado in the NHTSA database. Furthermore,
the FMVSS No. 301 rear impact test was modeled and it showed no damage to the fuel system.
The baseline 2014 MY Chevrolet Silverado's platform shares components across several
platforms. Some of the chassis components and other structural components were designed to
accommodate platform derivatives, similar to the components in the baseline vehicle which are
shared across platforms such as GMT 920 (GM Tahoe, Cadillac Escalade, GMC Yukon), GMT
930 platform (Chevy Suburban, Cadillac Escalade ESV, GMC Yukon XL), and GMT 940
platform (Chevy Avalanche and Cadillac Escalade EXT) and GMT 900 platform (GMC Sierra).
As per the National Academy of Science's guidelines, the study assumes engines would be
downsized or redesigned for mass reduction levels at or greater than 10 percent. As a
consequence of mass reduction, several of the components used designs that were developed for
other vehicles in the weight category of light-weighted designed vehicles were used to maximize
economies of scale and resource limitations. Examples include brake systems, fuel tanks, fuel
lines, exhaust systems, wheels, and other components.
Cost is a key consideration when vehicle manufacturers decide which fuel-saving
technology to apply to a vehicle. Incremental cost analysis for all of the new technologies
applied to reduce mass of the light-duty full-size pickup truck designed were calculated. The
cost estimates include variable costs as well as non-variable costs, such as the manufacturer's
investment cost for tooling. The cost estimates include all the costs directly related to
manufacturing the components. For example, for a stamped sheet metal part, the cost models
estimate the costs for each of the operations involved in the manufacturing process, starting from
blanking the steel from coil through the final stamping operation to fabricate the
component. The final estimated total manufacturing cost and assembly cost are a sum total of all
the respective cost elements including the costs for material, tooling, equipment, direct labor,
energy, building and maintenance.
The information from the LWT design study was used to develop a cost curve
representing cost effective full vehicle solutions for a wide range of mass reduction levels. At
lower levels of mass reduction, non-structural components and aluminum closures provide
weight reduction which can be incorporated independently without the redesign of other
components and are stand-alone solutions for the LWV. The holistic vehicle design using a
combination of AHSS and aluminum provides good levels of mass reduction at reasonably
acceptable cost. The LWV solution achieves 17.6 percent mass reduction from the baseline curb
mass. Further two more analytical mass reduction solutions (all aluminum and all carbon fiber
reinforced plastics) were developed to show additional mass reduction that could be potentially
achieved beyond the LWV mass reduction solution point. The aluminum analytical solution
predominantly uses aluminum including chassis frame and other components. The carbon fiber
reinforced plastics analytical solution predominantly uses CFRP in many of the components. The
CFRP analytical solution shows higher level of mass reduction but at very high costs. Note here

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
that both all-Aluminum and all CFRP mass reduction solutions are analytical solutions only and
no computational models were developed to examine all the performance metrics.
An analysis was also conducted to examine the cost sensitivity of major vehicle systems
to material cost and production volume variations.
Table 10-30 lists the components included in the various levels of mass reduction for the
LWV solution. The components are incorporated in a progression based on cost effectiveness.
Table 10-30 Components Included for Different Levels of Mass Reduction
Vehicle
Cumulative
Cumulative
Cumulative Cost
Cumulative Cost
Component/System
Mass Saving
(kg)
MR%

$/kg
Interior Electrical
1.38
0.06%
($28.07)
-20.34
Wiring




Headliner
1.56
0.06%
($29.00)
-18.59
Trim - Plastic
2.59
0.11%
($34.30)
-13.24
Trim - misc.
4.32
0.18%
($43.19)
-10.00
Floor Covering
4.81
0.20%
($45.69)
-9.50
Headlamps
6.35
0.26%
($45.69)
-7.20
HVAC System
8.06
0.33%
($45.69)
-5.67
Tail Lamps
8.46
0.35%
($45.69)
-5.40
Chassis Frame
54.82
2.25%
$2.57
0.05
Front Bumper
59.93
2.46%
$7.89
0.13
Rear Bumper
62.96
2.59%
$11.04
0.18
Towing Hitch
65.93
2.71%
$14.13
0.21
Rear Doors
77
3.17%
$28.09
0.36
Wheels
102.25
4.20%
$68.89
0.67
Front Doors
116.66
4.80%
$92.53
0.79
Fenders
128.32
5.28%
$134.87
1.05
Front/Rear Seat &
157.56
6.48%
$272.57
1.73
Console




Steering Column Assy
160.78
6.61%
$287.90
1.79
Pickup Box
204.74
8.42%
$498.35
2.43
Tailgate
213.14
8.76%
$538.55
2.53
Instrument Panel
218.66
8.99%
$565.06
2.58
Instrument Panel
221.57
9.11%
$580.49
2.62
Plastic Parts




Cab
304.97
12.54%
$1,047.35
3.43
Radiator Support
310.87
12.78%
$1,095.34
3.52
Powertrain
425.82
17.51%
1246.68
2.93
A fitted curve was developed based on the above listed mass reduction points to derive
cost per kilogram at distinct mass reduction points. The current curve shows costs per kilogram

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
approximately six times as expensive for 5 percent mass reduction (MR1) than in the NPRM,
and approximately twice as expensive per kilogram for 7.5 percent mass reduction (MR2), which
explains the reduction in mass reduction in the current analysis relative to the NPRM.
10.2 What Impacts Did NHTSA's "Method A" Analysis Show for
Regulatory Alternatives?
EPCA and EISA require NHTSA to "implement a commercial medium- and heavy-duty
on-highway vehicle and work truck fuel efficiency improvement program designed to achieve
the maximum feasible improvement" and to establish corresponding fuel consumption standards
"that are appropriate, cost-effective, and technologically feasible."DD For both the NPRM and
the current analysis of potential standards for HD pickups and vans, NHTSA applied NHTSA's
CAFE Compliance and Effects Modeling System (sometimes referred to as "the CAFE model"
or "the Volpe model") to aid in determination of the maximally feasible standards. The
subsequent analysis, referred to as "Method A," includes several updates to the model and to
accompanying inputs, as discussed above in Chapter 10.1. The "Method A" results are used as
the primary basis for NHTSA's final determination of the suitability of the Phase 2 standards.
Further discussion of the determination are provided after the discussion of the "Method A"
modeling results above.
10.2.1 Baseline Costs across Manufacturers
As in the NPRM, the main analysis of Method A considers costs, benefits and other
effects of regulatory alternatives relative to the dynamic baseline—or a baseline which assumes
that manufacturers will apply all technologies with associated cost that pays back from retail-
priced fuel savings within 6 months of purchase. The assumption is that consumers are willing
to pay additional technology costs that return in fuel savings within 6-months of purchase, and
that as a result, manufacturers will adopt these technologies regardless of fuel efficiency
standards. We considered alternative runs with voluntary overcompliance of technologies with a
payback period of 0-months (manufacturers will not voluntarily overcomply if there is a cost
associated with a technology), 12-months, 18-months, and 24-months in the sensitivity analysis.
Before considering the effects of increases in the standards, it is important to discuss the
baseline costs. These costs are assumed to be incurred even if no additional regulatory action is
taken to increase standards beyond the existing MY 2018 standards. Table 10-31 shows the
baseline average and total technology costs for each manufacturer in the heavy duty market, and
for the heavy duty industry as a whole for the MY 2021 fleet (cost increases relative to the MY
2015 fleet). The updated CAFE model suggests that under no further increasses to stringency
beyond MY 2018, manufacturers would spend $136 million—an industry average of $180 per
vehicle—on technologies that improve fuel economy in MY 2021. The additonal baseline costs
are not distributed across all manufacturers proportional to their fleet size. The average
technology costs of an individual manufacturer fleet range from $80 per vehicle for Fiat/Chrysler
to $350 per vehicle for General Motors. In order to explain this heterogeneity it is important to
DD 49 USC 32902(k)(2).

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
consider the sources of increased technology costs: compliance actions, inheritance from heavy
duty vehicles, spillover inheritance from the light-duty vehicles, and voluntary overcompliance.
Table 10-31 MY2021 Costs (2013$) under Alternative lb (Central Baseline) for 2b/3 Market
Manufacturer
Average per
Vehicle
Technology
Cost
(2013$)
Total
Technology
Cost
(million
2013$)
Estimated
MY 2015 Fuel
Consumption
(g/100 mi)
Estimated
MY 2018
Standard
(g/100 mi)
Daimler
$150
$3
4.50
4.84
FCA
$80
$10
6.23
5.95
Ford
$90
$33
6.00
5.76
GM
$350
$86
6.52
5.94
Nissan
$230
$3
6.01
5.63
Industry
$180
$136
6.18
5.83
One reason manufacturers incur technology costs in the baseline for MY 2021 vehicles is
to achieve compliance with Phase 1 standards, which end their stringency increases in MY 2018.
Manufacturers will have different standards and different starting positions relative to these
standards. In order to indicate which manufacturers make compliance actions which increase
their baseline technology costs, Table 10-31 includes the MY 2015 estimated average fuel
consumption and the estimated MY 2018 fuel consumption standard—manufacturers with higher
average fuel consumption in MY 2015 than the estimated MY 2018 fuel consumption standard,
will apply technology costs to comply with the final MY 2018 standards. The fuel consumption
standards are determined by setting work factor based targets and computing the manufacturer's
sales-weighted average of these targets. While the individual vehicle targets based on work
factor are the same for all vehicles of the same work factor for model years 2018 and beyond, the
overall fuel efficiency standard for a manufacturer may change from model year to model year
with changes to the work factors of individual vehicle models, as well as changes in relative
production volumes of each vehicle model. The model does not capture all means by which a
manufacturer's average fuel efficiency standard may change under the MY 2018 attribute-based
standards, but does capture changes to work factor—and therefore individual vehicle targets—
due to application of mass reduction. The model also predicts changes to the fleet mix of each
manufacturer using inputs created from AEO2015 and 2015 IHS/Polk production projections.
The technology cost for a manufacturer to meet MY 2018 standards is primarily driven by the
fuel consumption gap between the MY 2015 (baseline) compliance level and the 2018 standard.
From Table 10-38it can be seen that only Daimler meets its most-stringent fuel consumption
standard in 2015 and does not have to apply technology in the baseline to comply with Phase 1
standards.
A second source of technology costs is from inheritance; vehicles with shared platforms
are assumed to inherit technologies applied to the platform leader at their next redesign or refresh
to avoid creating a new body or engine platformEE, even if these actions are no longer necessary
EE For a more complete discussion of inheritance in the model see Chapter 6, Section C.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
to reach compliance. Manufacturers produce a limited set of engine and body platforms as a
strategy to reduce their costs; there is no reason to indicate they will modify this strategy to
comply with standards, for this reason this is an important constraint in the CAFE model. A
similar source of technology costs are costs associated with spillover from the light-duty MY
2017-2021 standards. Regulatory agencies distinctly define the heavy duty and light duty
classes, but from the manufacturer perspective these classes are not clearly delineated. They
share some engine and body platforms across regulatory classes, and sometimes the most cost-
effective choice to comply with standards will involve making changes to these shared platforms.
Comments in the NPRM recommended that we run the model with the ability to capture this
spillover effect between the light-duty and heavy-duty fleets—in response to these comments, in
the current analysis we run the two fleets together with all existing standards from the light-duty
fleet included for all scenarios. Since the MY 2017-2021 light-duty CAFE standards are final,
these and their effects are included in the baseline of the model—they will be in effect whether
or not additional action is taken with heavy-duty standards. While we have included the ability
for the standards from one fleet to affect the other, our modeling has shown that the spilloever
effect from the light-duty fleet into the heavy-duty fleet, and from the heavy-duty fleet into the
light-duty fleet is small. We hope to further develop the model's ability to capture the spillover
effects in future versions of the model.
The final way that manufacturers might accrue additional technology costs in the MY
2021 dynamic baseline scenario is through voluntary overcompliance. As already discussed: in
the baseline case of the central analysis it is assumed that manufacturers will apply technologies
which payback in fuel savings within 6 months of operation, regardless of whether or not the
standards increase in stringency. Depending on the existing technologies and vehicles in a
manufacturer's fleet, they may voluntarily overcomply by adding different technologies, or none
at all.
The MY 2021 costs of the dynamic baseline scenario are lower in the updated analysis
than they were in the NPRM for all manufacturers other than Nissan and Daimler. The average
technology costs across the industry are less than half the NPRM costs—dropping from
$440/vehicle to $180/vehicle. The largest drop in average costs across the manufacturers is for
GM; their costs dropped from $780/vehicle to $350/vehicle. The modeled costs for Nissan
dropped from $280 to $230, and for FCA, from $280 to $80.
While considering MY 2021 allows for comparision to the NPRM analysis, not all
baseline costs are incurred in MY 2021. Figure 10-11 shows the baseline total technology costs,
and Figure 10-12, the average technology costs, by manufacturer for all model years. Like the
NPRM analysis assumes manufacturers will likely apply most technologies as part of vehicle
redesign or freshening; as a result their technology application comes in discrete blocks. GM
applies $20 million in total technolgy for their MY 2016 fleet, and an additional $60 million in
for MY 2018—their total technology costs vary slightly after this point with the projection of
their fleet size and with the effects of technology learning. Similarly, Ford applies $30 million
for MY 2017 and an additional $80 million in 2027. Chrysler/Fiat, Daimler, and Nissan apply
technology in only one year—Chrysler/Fiat applies $11 million in MY 2018, Daimler $3 million
for MY 2020, and Nissan $3 million for MY 2021. While the total technology costs vary
between manufacturers, the per-vehicle baseline costs range between $0-350 for all
manufacturers and model years.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Daimler
General Motors
Nissan
Model Year
Figure 10-11 Total Annual Baseline Technology Costs (million 2013$) by Model Year and Manufacturer
Daimler	FCA
•Ford
-General Motors
Nissan
400
350
300
V)
O
U
> -C/V
PO250
o i
O O
E. =200
u o
|2f =
a, =150
60 S,
ro
ai
>
<
100
50
lO
ID

00
CD
O
*—1
fsj
ro
•3-
lO
iX)

00
CT)
O
*—1
*—1
*—i
*—1
*—1
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
ro
O
o
o
o
O
O
O
O
O
O
O
O
o
O
O
O
fsj
fsj

fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
Model Year
Figure 10-12 Per Vehicle Baseline Technology Costs (million 2013$) by Model Year and Manufacturer

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
10.2.2 Relevant Model Updates
There are changes to model that help explain the decrease in baseline technology costs for
the current analysis. The current analysis uses the synergies simulated by Argonne for the light-
duty fleet, while the NPRM analysis uses a limited set of synergy values (also initially estimated
for the light-duty fleet. The changes in these synergy factors could impact which technologies
are chosen, and how effective the model calculates them to be." Changes to the model input
costs from the NPRM to the current analysis could also change which technologies get picked by
the model, and the projected costs. One of the major changes to costs is a switch from the ICM
cost mark-up methodology used in the NPRM to the RPE cost mark-up methodology of the
current analysis.00 A more specific change to the input costs is a change to the mass reduction
curve to be based off of the newer 2014 Silverado study, which suggests that 5 percent and 10
percent mass reduction is significantly more expensive than was assumed in the NPRM.™
The final major input change is that the current model uses the 2015 fleet as its reference
point, while the NPRM uses the 2014 fleet. This affects the starting point of each manufacturer
in the model, and could change their predicted standard (through changes in sales mix and work
factor). In order to consider the impacts of using the 2015 reference fleet it is helpful to consider
the sales-weighted fuel economy and work factor distributions across the two reference fleets.
Figure 10-13 shows the sales-weighted empirical cumulative distribution function (CDF)
for GM's work factor and fuel economy for the two reference fleets. The dashed line shows the
values for the 2014 reference fleet, and the solid, for the 2015 reference fleet. The y-axis shows
the cumulative share of the manufacturer's fleet against the two measures. For GM, the work
factor CDF shifted to the right for work factors between 3500 and 5500, suggesting that the
proportion of the fleet with work factors in this range increased in the GM fleet. Since increases
in work factor will decrease the target value for individual vehicles, this average change in work
factor decreases GM's initial CAFE standard.
It should also be noted that some methods of increasing work factor (mainly, decreasing
curb weight) can increase the fuel efficiency of a vehicle, while others (increasing the power) can
decrease fuel efficiency. The empirical CDF for GM's sales-weighted fuel consumption shows
GM's 2015 fleet as having more vehicles with fuel consumption below 6.3 gal/100 mi, fewer
with fuel consumption around 6.3 gal/100 mi, significantly more vehicles with fuel consumption
around 7.0 gal/100 mi. The average fuel consumption of GM's 2014 fleet was 6.27 gal/100 mi,
where the average fuel consumption of GM's 2015 fleet is 6.52 gal/100 mi. The overall increase
in GM's average fuel consumption diminishes the effect of the increase in work factor from MY
2014 to MY 2015 at improving their starting position in MY 2015 relative to MY 2014—their
MY 2015 standard using the 2014 fleet was 6.36, and using the 2014 fleet and is 6.59.
Considering this, their initial shortfall is about the same using either reference fleet.
FF For a more complete discussion of the changes to the Argonne simulation synergies see Chapter 6, Section C.
GG For further discussion on the switch from ICM to RPE for the final analysis see Chapter 6, Section C.
1111 More discussion of the change in mass reduction curves is present in Chapter 6, Section C.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
2014 Fleet
2015 Fleet
2014 Fleet
2015 Fleet
£
o
oo
o
CD
o
o
CM
O
O
O
E
o
oo
o
CD
o
o
CN
o
o
o
i	1	1	1	1	r
3000 4000 5000 6000 7000 8000 9000	6.0	6.5	7.0	7.5
Work Factor	Fuel Consumption (gal/100 mi.)
Figure 10-13 2014 vs. 2015 Reference Fleet Work Factor and Fuel Efficiency for General Motors.
Figure 10-14 shows the same for Ford. There is a similar pattern of a higher proportion
of heavy duty vehicles in Ford's fleet with work factors between 3500 and 5000. This will
decrease Ford's initial standard in the model. Ford also shows a decrease in the proportion of
heavy duty vehicles with higher fuel consumption, which will result in an overall lower fuel
consumption for the 2015 fleet. The result is that Ford will start with a lower standard by using
the 2015 fleet rather than the 2014 fleet, and start with a higher fuel efficiency level—both of
which will work in the same direction to decrease Ford's shortfall to MY 2018 standards. This
suggests that Ford will not need to apply as much technology to comply, and helps to explain
their lower baseline technology costs in the current analysis.
2014 Fleet — 2015 Fleet
-- - 2014 Fleet — 2015 Fleet
E
3
o
CO
o
o
o
o
E
Z3
o
CO
o
o
o
o
2000
4000 6000
Work Factor
8000
4	6
Fuel Consumption (gal/100 mi.)
Figure 10-14 2014 vs. 2015 Reference Fleet Work Factor and Fuel Economy for Ford

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Figure 10-15 shows the cumulative distribution function for the work factor of
Fiat/Chrysler. Although there is some increase in the left tail of the distribution of FCA's work
factor for MY 2015 relative to MY 2014, it is smaller than for the Ford and GM fleets. The CDF
of fuel efficiency also shows that Fiat/Chrysler shows nearly identical distribution of fuel
consumption between the 2014 and 2015 fleets. These two factors combine to explain why
Fiat/Chrysler did not show increases in costs from the NPRM to the current analysis—they did
not have as much of a change in shortfall to MY 2018 standards as both GM and Ford.
-" * 2014 Fleet — 2015 Fleet	-' - 2014 Fleet — 2015 Fleet
O
P
03
Li_

:=-
03
O
o
o
4000
6000
8000
10000
Work Factor	Fuel Consumption (gal/100 mi.)
Figure 10-15 2014 vs. 2015 Reference Fleet Work Factor and Fuel Economy for Fiat/Chrysler
Figure 10-16 shows the same empirical distribution functions for Nissan. Both the
distribution of work factor and fuel consumption are comparable for Nissan's 2014 and 2015
fleets. This helps explain the small change in Nissan's baseline costs between the two analyses.

-------
r E. O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review
¦¦¦ 2014 Fleet
2015 Fleet
-- - 2014 Fleet
2015 Fleet
w
OO
o
CD
O
o
O CN
O
o
r
r
3000 3400 3800
Work Factor
4200
CO
2 o
(U
_£Z
 ¦
Qj O
£ ^
o
O CM
O
O
O
r
5.6
~r
J

~T
5.8 6.0 6.2 6.4
Fuel Consumption (gal/100 mi.)
Figure 10-16 2014 vs. 2015 Reference Fleet Work Factor and Fuel Economy for Nissan
Figure 10-17 shows the cumulative distribution function for work factor and fuel
consumption for Daimler for both the 2014 and 2015 fleets. The distribution of work factor
shifted right for work factors above 3500. The fuel consumption curve shifted right for all fuel
consumptions. This suggests that Daimler will face a lower standard using the 2015 reference
fleet, but that they may also start with a lower initial fuel efficiency level. The change to the
2015 reference fleet does not have clear implications on the relative starting point of Daimler in
the analysis relative to the NPRM analysis.
2014 Fleet
2015 Fleet
2014 Fleet
2015 Fleet
E
U
o
oo
o
CD
O
O
CN
O
O
O
3000 3500 4000 4500
Work Factor
5000
E
3
o
00
o
CD
O
o
CN
o
o
o
T
T
~~T~
4.4
~T
~r
4.0 4.2 4.4 4.6 4.8
Fuel Consumption (gal/100 mi.)
~r~
5.0
Figure 10-17 2014 vs. 2015 Reference Fleet Work Factor and Fuel Economy for Daimler

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
10.2.3 Industry-Level Results of Regulatory Alternatives
Table 10-32, below, summarizes the stringency of standards, the estimated required fuel
efficiency the estimated achieved fuel efficiency, as well as the impacts of each alternative for
the overall industry for MY 2030. Using the updated fleet and analysis, the MY 2030 stringency
is slightly less that in the NPRM (4.91 gallons/100 mile in today's analysis compared to 4.86
gallons/100 mile in the NPRM for the preferred alternative). As has been noted, the standards
are set based in part on the work factor of vehicles; by changing the average work factor of their
fleet, manufacturers can change the average stringency of their standard. While the model does
not simulate changes to work factor which would increase the power or GVWR, it does simulate
changes in work factor due to mass reduction. By lowering the curb weight and holding power
constant, manufacturers can increase the payload of a vehicle; since payload is a component in
calculating the work factor, by lowering curb weight manufacturers can increase their work
factor for a vehicle model and reduce its target. However, the average absolute and proportional
curb weight reduction in the current analysis is less than it was in the NPRM analysis across all
alternatives, which can be explained by the higher mass reduction costs under the current curve.
This suggests that the change in the average overall industry standard in today's analysis is likely
due in major part to changes in the work factor between the 2014 and 2015 reference fleet, and
not to changes in the work factor simulated within the model runs.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-32 Summary of Impacts on the MY2030 HD Industry Fleet (vs. Alternative lb)
Alternative
2
3
4
5
Stringency of Standards
Annual Increase in
Stringency Beginning in
MY 2021
2.0%
2.5%
3.5%
4.0%
Increases Until
MY2025
MY2027
MY2025
MY2025
Total Increase in MY2030
Stringency Relative to
Final Phase 1 Standardsa
9.6%
15.6%
15.6%
17.9%
Estimated Average Fuel Economy (miles per gallon)
Required in MY 2030
19.03
20.37
20.38
20.95
Achieved in MY 2030
19.20
20.47
20.45
20.98
Average Fuel Consumption (gallons/100 miles)
Required in MY 2030
5.25
4.91
4.91
4.77
Achieved in MY 2030
5.21
4.88
4.89
4.77
Estimated Average Greenhouse Gas Emissions (grams per mile)
CO 2 Required in MY 2030
494
462
462
450
CO 2 Achieved in MY
2030
490
460
460
449
Technology Penetration in MY 2030 (percent)
VVT and/or VVL
56
56
56
56
Cylinder Deactivation
4
4
4
4
Direct Injection Engine
17
27
26
29
Turbo Charged Engine
59
69
68
68
8 Speed Auto. Trans.
77
95
94
95
EPS, Accessories
52
80
80
96
12V Stop-start
0
0
3
11
Strong Hybrid
0
2
2
7
Aero. Improvements
46
80
80
98
Mass Reduction (vs. No-Action)
Mass Reduction (lb.)
28
240
24
289
Mass Reduction (percent
of curb weight)
0.43
3.6
3.7
4.3
Technology Costs (vs. No-Action)
Average Vehicle ($)
$500
$1470
$1480
1890
Payback Period (m)b
19
30
31
33
Notes:
a This increase in stringency is based on the estimated percentage change in fuel consumption (gal/lOOmi)
stringency projected by the model for the MY2030 fleet under the final Phase 2 standards relative to the
continuation of Phase 2 standards. Note that if manufacturers' have applied mass reduction to an individual

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
vehicle model in the CAFE model that this will increase the work factor of that vehicle in the model, and
make the individual target less stringent. Thus, where any mass reduction is applied in the model, the total
increase in stringency of the fleet presented here will be lower than the total stringency increase of the fleet
if no mass reduction were applied.
b Here payback period is calculated using estimated undiscounted retail fuel savings and the initial
technology costs for MY2030.
Today's Method A analysis using the updated version of the CAFE model and updated
inputs shows that regulatory Alternatives 3 and 4 could be met with a small application of strong
(P2) HEVs. However, Alternative 5 could be met with the considerably greater application of
strong HEVs. Although there is some increase in the penetration rates between alternatives as
stringency increases, the current analysis suggests that under all alternatives, nearly all of the
MY 2030 heavy-duty fleet could use 8-speed transmissions, VVT/VVL improvements and turbo-
charged engines with application across more than half of the fleet, direct injection could be
present in a quarter of the fleet, and cylinder deactivation could play a minor part in the HD fleet.
EPS and improved electrical accessories vary more between alternatives; present in 52 percent of
the fleet in Alterative 2, 80 percent in Alternatives 3 and 4, and 96 percent in Alternative 5.
Aerodynamic improvements and mass reduction follow a similar pattern; with a larger
penetration of these technologies with Alternative 3 than with Alternative 2, a similar penetration
under Alternatives 3 and 4, and a higher in penetration in Alternative 5.
A way to measure the cost-effectiveness of the technologies on consumers is to look at
the payback period. In this context, the payback period is defined as the number of months of
driving it will take a consumer to earn back the increased technology costs by the amount they
save in fuel by driving a more fuel efficient vehicle. Under the current analysis, the average
additional technology cost will payback in fuel savings in under 17 months for Alternative 2, 27
months for Alternatives 3 and 4, and 30 months for Alternative 5. It is important to note that
there are inputs other than the cost and effectiveness of technologies which could affect the
payback period; the fuel prices and mileage accumulation schedules will affect how quickly the
cost of a fuel-saving technology pays back.
The current analysis uses updated fuel price estimates from AEO 2015 that are lower than
in the NPRM analysis. Lower fuel prices will decrease the absolute amount of fuel savings
(assuming the same number of gallons is consumed) and increase the payback period if the
technologies, their cost, and their effectiveness are unchanged. Further, we have updated the
vehicle use schedule (vehicle miles traveled, or VMT) based on actual vehicle odometer readings
from IHS/Polk data as shown in Figure 13.9. While the overall survival-weighted schedules
show 6.5 percent fewer lifetime miles for heavy-duty vehicles, they show more annual miles
driven for the first 5-years of use for heavy-duty vehicles. The result is that the overall lifetime
fuel savings will decrease, but the fuel savings will be higher for the first 5 years. Since the
payback periods under both analyses are shorter than 5 years, using the updated vehicle
schedules will show a shorter payback period (if other factors are unchanged) than in the NPRM
analysis. The changes in fuel prices and the change in the mileage accumulation schedule work
in opposite directions on the payback period; the total change in payback period is attributable to

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
both of these input changes as well as to the changes in the cost11 and effectiveness" of the
different technology inputs, and the changes in the reference fleet.
Industry costs in MY 2030 provide one perspective on technology costs. Industry cost in
each model year provides additional perspective on the timing, pace and the amount of resources
and spending that would need to be allocated to implement technologies and is important in the
consideration of the feasibility of the alternatives. Figure 10-18 and Figure 10-19 show the total
and average additional and total additional technology costs for the industry by model year and
alternative. Note that the trend of the total and average costs are very similar, this is because the
fleets size the AEO projections suggest a relatively constant fleet size during the considered
MY's. The total and average technology costs increase with alternative stringency. It is
important to note that Alternatives 3 and 4 both increase total stringency for the MY2030
industry fleet by 15.6 percent. Also note that these estimations of stringency increases include
the model projections of how the application of mass reduction will alter work factor and
individual vehicle targets.^ The annual average and total technology costs of Alternative 3
approach those of Alternative 4 by MY 2029 when both alternatives have reached maximum
stringency. If manufacturers are to reach the same stringency level over a longer horizon, they
will likely make similar technology choices, but be given longer to implement them. This will
make the total technology costs lower, but should unsurprisingly make the marginal technology
costs for model years where both standards have matured very similar.
11 The costs now use RPE rather than ICM, and we updated the mass reduction curve to the 2014 Silverado.
11 Nominal effectiveness input values are as for the NPRM analysis. Synergy factors applied to adjust fuel
consumption impacts for specific combinations of technologies reflect current vehicle simulation work conducted
for NHTSA by Argonne National Laboratory.
KK The final Phase 2 standard target curves increase in stringency by 16.2 percent compared to final Phase 1
standards, as discussed in Section VLB.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
1.8
¦Alternative 2
¦Alternative 3
Alternative 4
• Alternative 5
ai
IS)
ro
0)
1.6
1.4
u
ts 1-2
o «
u o 1
> PM
M c
£ O 0.8
o :=
¦5 £.0.6
£
ro 0.4
0.2

Model Year
00
fsj (N
O O
fsj (N
CD	O
fsj	ro
O	O
fsj	fsj
Figure 10-18 Industry Total Technology Cost Increase by Model Year and Alternative
2500
•Alternative 2
•Alternative 3
Alternative 4
• Alternative 5
(0

<
1000
500
	' . • ¦
Model Year
lo	id
fsj fsj	fsj
O O	O
fsj fsj	fsj
oo
fsj fsj
o o
fsj fsj
CD	O
fsj	ro
o	o
fsj	fsj
Figure 10-19 Industry Average Technology Cost Increase by Model Year and Alternative
The average incremental industry technology costs mature to around $500 under
Alternative 2, $1500 under Alternatives 3 and 4, and $1900 under Alternative 5. Figure 10-20

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
shows the cumulative total industry costs by model year fleet. $4.2 billion in additional
technology costs for model years 2016-2030 are associated with Alternative 2, $9.9 billion with
Alternative 3, $11.4 billion with Alternative 4, and $14.9 billion with Alternative 5. While the
marginal technology costs of Alternative 3 approach those of Alternative 4 as the total
stringencies converge, the total costs of Alternative 4 are $1.5 billion more by MY2030. It is
particularly noteworthy that costs and the rate of increase in costs would be significantly
different in the MYs 2017 - 2021 timeframe among the alternatives. This identifies the
significant differences in the resources and capital that would be required to implement the
technologies required to comply with each of the alternatives during this period, as well as the
reduction in lead time to implement the technologies which increases reliability risk. These
differences are an important consideration for the feasibility of the alternatives and for the
selection of the final standards, as discussed further below.
Alternative 2
Alternative 3
Alternative 4
Alternative 5
Model Year
Figure 10-20 Industry Cumulative Total Technology Cost Increase by Model Year and Alternative
10.2.4 Manufacturer-Specific Results of Regulatory Alternatives
In addition to varying across scenario and model year, the impacts of the standards vary
across manufacturers. Manufacturers will have different compliance strategies based on which
technologies they have already invested in, in both their heavy-duty and light-duty fleets, and
based on the effectiveness of new technology applications specific to the vehicles in their heavy
duty fleets. Table 10-33 summarizes the initial technology utilization in the 2015 fleet by
manufacturer. Ford uses direct injection for 8 percent of their fleet, cylinder deactivation for 13
percent of their fleet, and turbo-charged engines for 8 percent of their fleet. Daimler has already
invested to equip all of its fleet with 8-speed automatic transmissions. These differences in

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
initial technology levels affect the new investments each manufacturer would need to further
improve the fuel efficiency of their fleets.
Table 10-33 Summary of MY2015 Reference Fleet Technology Penetration
Technology
GM
Ford
FCA
Daimler
Nissan
Industry
Technology Penetration (percent)
Cylinder Deactivation
0
0
13
0
0
2
Direct Injection
Engine
0
8
0
0
0
4
Turbo Charged Engine
0
8
0
0
0
4
8 Speed Auto. Trans.
0
0
0
100
0
3
EPS, Accessories
0
0
0
0
0
0
12V Stop-start
0
0
0
0
0
0
Strong Hybrid
0
0
0
0
0
0
Aero. Improvements
0
0
0
0
0
0
10.2.4.1 General Motors
Table 10-34 summarizes the alternatives, and a technology pathway General Motors
could use to comply with each of the alternatives. The pathway includes implementing 8 speed
automatic transmissions across its entire fleet. For Alternatives 2 and 3, no stop-start or HEVs
are added to GM's fleet, for Alternative 4, 1 percent of GM's fleet uses stop-start, and for
Alternative 5, 2 percent uses stop-start and 13 percent are HEVs. For all alternatives, nearly all
of the GM's fleet would use electric power steering and improved electric accessories.
For all alternatives, VVT/VVL is applied to 65 percent of its engines. For Alternative 2,
none of its engines get direct injection and 43 percent get turbocharging and downsizing, while
for Alternatives 3-5, direct injection is applied to 28 percent of its engines and turbocharging and
downsizing is applied to 61 percent of its engines. For all alternatives, all of GM's fleet gets
aerodynamic improvements. The average mass reduction is 52 lbs. (0.78 percent of the average
curb weight) under Alternative 2, and 350-380 lbs. (5.2-5.7 percent of the average curb weight)
under Alternatives 3-5. Similar technology is applied for Alternatives 3 and 4 in MY 2030, but
there are significantly more strong hybrids under Alternative 5.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-34 Summary Impacts on General Motors HD Fleet by Alternative (vs. Alternative lb)
Alternative
2
3
4
5
Alternative Stringency
Annual Increase in
Stringency Beginning
in MY 2021
2.0%
2.5%
3.5%
4.0%
Increases Until
MY2025
MY2027
MY2025
MY2025
Total Increase in
MY2030 Stringency
Relative to Final Phase
1 Standards a
9.6%
15.2%
15.4%
17.7%
Estimated Average Fuel Economy (miles per gallon)
Required in MY 2030
18.69
19.92
19.96
20.53
Achieved in MY 2030
18.70
20.04
20.04
20.6
Average Fuel Consumption (gallons/100 miles)
Required in MY 2030
5.35
5.02
5.01
4.87
Achieved in MY 2030
5.35
4.99
4.99
4.85
Estimated Average Greenhouse Gas Emissions (grams per mile)
CO 2 Required in MY
2030
498
467
466
453
CO 2 Achieved in MY
2030
496
464
464
452
Technology Penetration in MY 2030 (percent)
WT and/or VVL
65
65
65
65
Cylinder Deactivation
0
0
0
0
Direct Injection Engine
0
28
28
28
Turbo Charged Engine
33
61
61
61
8 Speed Auto. Trans.
100
100
100
100
EPS, Accessories
100
100
100
100
12V Stop-start
0
0
2
2
Strong Hybrid
0
0
0
13
Aero. Improvements
100
100
100
100
Mass Reduction (vs. No-Action)
Curb Weight Mass
Reduction (lb.)
52
384
384
340

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Mass Reduction
(percent of curb weight)
0.78
5.7
5.7
5.1
Note:
a This increase in stringency is based on the estimated percentage change in fuel consumption (gal/lOOmi)
stringency projected by the model for the MY2030 fleet under the final Phase 2 standards relative to the
continuation of Phase 1 standards. Note that if manufacturers' have applied mass reduction to an individual
vehicle model in the CAFE model that this will increase the work factor of that vehicle in the model, and
make the individual target less stringent. Thus, where any mass reduction is applied in the model, the total
increase in stringency of the fleet presented here will be lower than the total stringency increase of the fleet
if no mass reduction were applied.
Figure 10-21 and Figure 10-22 show the total and average incremental technology costs
by alternative. Under Alternative 2 General Motors' incremental technology cost is $140M in
MY 2019, increasing to $180M in MY2021. The pathways for Alternatives 3 and 4 are very
similar, which again should not be surprising given that the standards result in the same total
stringency increase in MY 2027 and beyond and the long redesign cycles in the segment. GM's
incremental technology cost is $190M in MY 2019, increasing to $400M in MY 2021, and
$530M in MY2028. Under Alternative 5 GM could have a similar compliance strategy as
Alternative 3 and 4, but incremental technology cost is$650M in MY2028. The highest annual
average technology cost for GM is: $750 under Alternative 2, $1940 under Alternatives 3 and 4,
and $2370 under Alternative 5. In the case of GM, the added lead time of Alternative 4 does not
significantly change the cost of their compliance strategy.
700
¦Alternative 2 — — — Alternatives
Alternative 4 — • -Alternatives
Q)



c



•*-*
¦to-

rsj
cuo
c
o
o
o

c

-C
£_
u

0)

H

"ro



o

H

600
500
/ ^ . .
7 —
200
100
o
o
fsj
O
fsj
LO
fsj
O
fsj
ID
fsj
O
fsj
fsj
o
fsj
00
fsj
O
fsj
CD
fsj
O
fsj
O
CO
O
fsj
Model Year
Figure 10-21 Total Technology Cost Increase for General Motors by Model Year and Alternative

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Q)

\A

ro

0)



u

c

4-*

\A

O

U

> w
CUD
ro
O

o
O
c
rsj
.c

u

Q)

H



<

2500
2000
1500
1000
500
¦Alternative 2 - — -Alternatives
Alternative 4 — • -Alternatives
//
•-«/
lO


00
CD
O
*—1
fsj
ro
•3-
lO


00
CD
O
*—1
*—1
*—i
*—1
*—1
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
ro
O
O
o
o
o
O
O
O
O
O
O
O
O
O
O
O
fsj
fsj

fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
Model Year
Figure 10-22 Average Technology Cost Increase for General Motors by Model Year and Alternative
Figure 10-23 shows the cumulative total incremental costs for GM under all alternatives.
The total costs to comply with Alternative 2 for GM for MY's 2016-2030 is $2.1 billion, for
Alternatives 3 and 4 it is $4.8 billion, and for Alternative 5 it is $5.2 billion.
Alternative 2
Alternative 3
Alternative 4
Alternative 5
Model Year
Figure 10-23 General Motors Cumulative Total Technology Cost Increase by Model Year and Alternative

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
10.2.4.2 Ford
Table 10-35 gives the same summary of a potential compliance strategy for Ford's
heavy-duty fleet. Similar to GM, to reach compliance Ford uses 8 speed automatic transmissions
in their entire fleet. For Alternatives 3 and 4, Ford uses hybrid technologies in 4 percent of their
fleet, and for Alternative 5, they use hybrid technologies in 7 percent of their fleet. In addition to
strong hybrids, Ford uses 12v stop-start in 4 percent of their fleet in Alternative 4, and 12v stop-
start in 19 percent of their fleet in Alternative 5. The compliance strategy in the NPRM analysis
shows Ford using significantly more hybrids and 12v stop-start systems in Alternatives 4 and 5
than the current analysis which likely explains part of the lowered cost for Ford in the current
analysis.
Under the current analysis possible compliance strategy, the application of engine
technologies for Ford come in discrete chunks, as with GM. Ford uses VVT/VVL in 58 percent
of their fleet under all alternatives by MY2030; they started with 8 percent direct-injection
engines, and end with 27 percent; they also started with 8 percent turbo-charged engines, but end
with 69 percent for all scenarios. The application of EPS and improved accessories vary across
the compliance strategies of different regulatory alternatives; under Alternative 2, only 13
percent of Ford's fleet improves these electrical features, while under Alternatives 3-4, 64
percent, and Alternative 5, 96 percent.
For body-platform technologies, Ford applies in discrete chunks to the same platforms
across some Alternatives. They apply an average of 77 lb. (1.2 percent) mass reduction across
their fleet in Alternative 2 and 132-142 lb. (2.0-2.2 percent) in Alternative 3-5. Progressively
less mass reduction is applied under Alternatives 4 and 5—this is likely because more of the fleet
was hybridized and mass reduction to small platforms was no longer necessary to comply.
Aerodynamic improvements are not applied in Alternative 2, but are applied to 64 percent of the
fleet in Alternative 3 and 4, and to all of the fleet in Alternative 5.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-35 Summary of Impacts on Ford HD Fleet by Alternative (vs. Alternative lb)
Alternative
2
3
4
5
Alternative Stringency
Annual Increase in
Stringency Beginning
in MY 2021
2.0%
2.5%
3.5%
4.0%
Increases Until
MY2025
MY2027
MY2025
MY2025
Total Increase in
MY2030 Stringency
Relative to Final Phase
1 Standards a
9.6%
15.7%
15.7%
18.1%
Estimated Average Fuel Economy (miles per gallon)
Required in MY 2030
19.23
20.62
20.62
21.23
Achieved in MY 2030
19.36
20.61
20.63
21.21
Average Fuel Consumption (gallons/100 miles)
Required in MY 2030
5.2
4.85
4.85
4.71
Achieved in MY 2030
5.16
4.85
4.85
4.71
Estimated Average Greenhouse Gas Emissions (grams per mile)
CO 2 Required in MY
2030
488
456
455
443
CO 2 Achieved in MY
2030
485
455
455
443
Technology Penetration in MY 2030 (percent)
WT and/or VVL
58
58
58
58
Cylinder Deactivation
0
0
0
0
Direct Injection Engine
27
27
27
27
Turbo Charged Engine
69
69
69
69
8 Speed Auto. Trans.
64
100
100
100
EPS, Accessories
13
64
64
96
12V Stop-start
0
0
4
19
Hybridization
0
4
4
7
Aero. Improvements
0
64
64
100
Mass Reduction (vs. No-Action)
Curb Weight Mass
Reduction (lb.)
77
142
140
132

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Mass Reduction
(percent of curb weight)
Note:
a This increase in stringency is based on the estimated percentage change in fuel consumption (gal/lOOmi)
stringency projected by the model for the MY2030 fleet under the final Phase 2 standards relative to the
continuation of Phase 1 standards. Note that if manufacturers' have applied mass reduction to an individual
vehicle model in the CAFE model that this will increase the work factor of that vehicle in the model, and
make the individual target less stringent. Thus, where any mass reduction is applied in the model, the total
increase in stringency of the fleet presented here will be lower than the total stringency increase of the fleet
if no mass reduction were applied.
Figure 10-24 and Figure 10-25 show the total and average incremental technology costs
for Ford by alternative and model year. Ford adds $80 million in technology costs for MY 2017
and an additional $40 million in MY 2026 in Alternative 2. For the Preferred Alternative, Ford
adds $130 million in MY2017 and an additional $300 million in MY 2026. Under Alternative 4,
Ford adds $260 million in MY 2017 and $180 million in MY 2026. Similar to the industry
pattern, Ford's compliance strategy involves less annual technology costs early in Alternative 3
than Alternative 4, but their technology costs converge under the two alternatives as the final
stringency level is reached under Alternative 3 in MY 2027.
It is important to note that the increase in costs and rate of the increase in costs is
significantly different for MY 2017 among the alternatives—with the incremental total cost
increase for MY 2017 being double those of Alternative 3 for Alternative 4, and more than
double for Alternative 5. MY 2017 is the first redesign year and Ford does not have another
scheduled redesign until MY 2026. Under the additional lead time of Alternative 3, the majority
of Ford's cost increases occur in the MY 2026 redesign, while Alternatives 4 and 5 put most of
the cost burden to reach compliance on the MY 2017 redesign (or would require an additional
redesign be added between MY 2017 and 2026). NHTS A judges the lack of lead time would
make Alternatives 4 and 5 beyond maximum feasibility for Ford because its designs for MY
2017 are essentially complete and substantial resources and very high costs would be required to
add another vehicle redesign between MY 2017 and MY 2026 to implement the technologies
that would be needed to comply with those alternatives.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
700
¦Alternative 2
¦Alternative 3
Alternative 4
• Alternative 5
600
01
u
(0
01
= ^ 500
4-> VV
o 00
u o 400
> 
o
200
100
3
•7
:i
.*/
•
•• /
/
/
lO
ID

00
CD
O
*—1
fsj
ro
•3-
lO


00
CD
O
*—1
*—1
*—i
*—1
*—1
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
ro
O
o
o
o
O
O
O
O
O
O
O
O
O
O
O
o
fsj
fsj

fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
Model Year
Figure 10-24 Total Technology Cost Increase for Ford by Model Year and Alternative
¦Alternative 2 - — -Alternatives	Alternative 4 — • -Alternatives
1800
01
£ 1600
01
1400
g 1200	I ^ •	/ y
<3	• ^ _	
§5 £ 1000	I
O
° S 800
— £ •/
O onn	•
cu 600	* ;		*'
I;	/
So 400	•; 			I

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Alternative 2
Alternative 3
Alternative 4
Alternative 5
Model Year
Figure 10-26 Ford Cumulative Technology Cost Increase by Model Year and Alternative
10.2.4.3 Fiat/Chrysler
Table 10-36shows the MY 2030 summary for Fiat/Chrysler. Fiat/Chrysler is the only
manufacturer which uses cylinder deactivation in their reference fleet, and they are the only
manufacturer to use cylinder deactivation as a part of their possible compliance strategy. Under
all scenarios, FCA increases their initial cylinder deactivation utilization of 13 percent to 24
percent. Under all scenarios turbo-charged engines are applied to 76 percent of FCA's fleet by
MY 2030. Other technologies are applied to the FCA equally across all scenarios; 37 percent of
their fleet uses VVT and/or VVL, and 64 percent uses 8-speed automatic transmissions under all
scenarios.
The additional stringency from Alternative 2 to Alternatives 3-5 results in other increased
technology applications in the FCA fleet. Under Alternatives 3-5, the presence of EPS/electrical
accessories increases from the 82 percent to the entirety of the FCA fleet. Similarly, increased
aerodynamic improvements increase from 84 percent of the fleet to all of it. Finally, 12v stop-
start enters 3 percent of the fleet under Alternatives 3-5. Alternatives 3 and 4 look much the
same, except that Alternative 3 is the only alternative to use any (1 percent) SHEV-P2 hybrids.
Alternative 5 uses twice as much mass reduction than Alternatives 3-4; it uses 37 percent direct
injection versus the 24 percent in Alternatives 2-4. The resulting costs are comparable under
Alternatives 3 and 4, and almost 50 percent higher under Alternative 5.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-36 Summary of Impacts on Fiat/Chrysler HD Fleet by Alternative (vs. Alternative lb)
Alternative
2
3
4
5
Alternative Stringency
Increases Until
MY2025
MY2027
MY2025
MY2025
Total Increase in
MY2030 Stringency
Relative to Final Phase
1 Standards a
9.6%
15.8%
15.8%
17.6%
Total Increase in
Stringency Relative to
Final Phase 1 Standards
MY2025
MY2027
MY2025
MY2025
Estimated Average Fuel Economy (miles per gallon)
Required in MY 2030
18.59
19.96
19.96
20.41
Achieved in MY 2030
18.97
20.06
20.04
20.42
Average Fuel Consumption (gallons/100 miles)
Required in MY 2030
5.38
5.01
5.01
4.9
Achieved in MY 2030
5.27
4.99
4.99
4.9
Estimated Average Greenhouse Gas Emissions (grams per mile)
CO 2 Required in MY
2030
520
485
485
474
CO 2 Achieved in MY
2030
509
482
482
474
Technology Penetration in MY 2030 (percent)
WT and/or VVL
37
37
37
37
Cylinder Deactivation
24
24
24
24
Direct Injection Engine
24
24
24
37
Turbo Charged Engine
76
76
76
76
8 Speed Auto. Trans.
64
64
64
64
EPS, Accessories
82
100
100
100
12V Stop-start
0
3
3
3
Hybridization
0
1
0
0
Aero. Improvements
84
100
100
100
Mass Reduction (vs. No-Action)
Curb Weight Mass
Reduction (lb.)
29
330
333
694

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Mass Reduction
(percent of curb weight)
0.4
4.6
4.6
9.6
Note:
a This increase in stringency is based on the estimated percentage change in fuel consumption (gal/lOOmi)
stringency projected by the model for the MY2030 fleet under the final Phase 2 standards relative to the
continuation of Phase 1 standards. Note that if manufacturers' have applied mass reduction to an individual
vehicle model in the CAFE model that this will increase the work factor of that vehicle in the model, and
make the individual target less stringent. Thus, where any mass reduction is applied in the model, the total
increase in stringency of the fleet presented here will be lower than the total stringency increase of the fleet
if no mass reduction were applied.
Figure 10-27and Figure 10-28 show the incremental total and average technology costs
for Chrysler/Fiat by model year and regulatory stringency. Chrysler/Fiat shows more technology
costs for higher stringency alternatives, with annual technology costs of Alternative 3
approaching Alternative 4 annual technology costs as the Alternative 3 approaches the final
stringency level in MY 2027. Under all alternatives Chrysler/Fiat incurs increased technology
costs starting in MY 2018 and MY2025, because they are estimated redesign years. The
maximum annual technology costs for Chrysler are $92M in Alternative 2, $213M in Alternative
3, $227M in Alternative 4, and $330M in Alternative 5. This results in average technology costs
of: $680, $1640, $1690, and $2460, respectively.
As with Ford, the costs and the rate of increase in costs are significantly different in the
MY 2018 timeframe among the alternatives, because MY 2018 is the first estimated model year
for redesign, and the next estimated redesign opportunity is in MY 2025. Figure 10-27identifies
the significant differences in the resources and capital that would be required to implement the
technologies required to comply with each of the alternatives—with the estimated MY 2018
technology cost increases being 48M under Alternative 3, 78M under Alternative 4, and 112M
under Alternative 5. NHTS A judges the short lead time would make Alternatives 4 and 5
beyond maximum feasible for FCA because its designs for MY 2018 are nearing completion and
substantial resources and very high costs would be required to add another vehicle redesign
between MY 2018 and MY 2025 to implement the technologies that would be needed to comply
with those alternatives.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
¦Alternative 2 - — -Alternatives
350
Alternative 4
¦ Alternative 5
01
re
01
4-> W
« fo
U o
>  r-
O o
o —
u
01
(0
4-"
o
300
250
200
= 150
100
50
0
lO
ID

00
CD
O
*—1
fsj
ro
•3-
lO


00
CD
O
*—1
*—1
*—i
*—1
*—1
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
ro
O
o
o
o
O
O
O
O
O
O
O
O
O
O
O
O
fsj
fsj

fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
Model Year
Figure 10-27 Total Technology Cost Increase for Fiat/Chrysler by Model Year and Alternative
•Alternative 2 - — -Alternatives 	Alternative 4 — • -Alternatives
 • /¦
' 		r
y
mior^ooaiOiHiNcofmior^ooaio
iHiHiHiHiHfMfMfMfMfMfMfMfMfMfMCO
oooooooooooooooo
fMfMfMfMfMfMfMfMfMfMfMfMfMfMfslfM
Model Year
Figure 10-28 Average Technology Cost Increase for Fiat/Chrysler by Model Year and Alternative
The cumulative technology costs attributable to the action alternatives for FCA are
represented in Figure 10-29, below. The total costs for MY's 2016-2030 under alter Alternative
2 are $750 million, under Alternative 3, they are $1.5 billion, for Alternative 4, $1.8 billion, and
for Alternative 5 they are $2.6 billion.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Alternative 2
Alternative 3
Alternative 4
Alternative 5
Model Year
Figure 10-29 Fiat/Chrysler Cumulative Technology Cost Increase by Model Year and Alternative
10.2.4.4 Nissan
Table 10-37 shows the manufacturer-specific MY2030 summary for Nissan. Nissan's
2015 reference fleet uses VVT and/or VVL on all of their heavy-duty vehicles. Their fleet uses
two engines on only one body-style platform. As a result, technologies applied to Nissan's fleet
are applied to large proportions of their fleet. Under all scenarios, their entire fleet gains 8-speed
automatic transmissions. Under Alternatives 3-5, all of their fleet gets level-2 body-level
aerodynamic improvements and all of their fleet gets electric accessory and/or EPS
improvements. Under Alternatives 2, 4, and 5, one of Nissan's two heavy-duty engines gets
direct-injection, while under Alternative 3, both engines get the technology. Direct injection of
their entire fleet is the most cost-effective way to reach compliance under Alternative 2, applying
5 percent mass reduction to their entire fleet and direct injection of one of their engines is the
most cost-effective strategy under Alternative 4, and applying 10 percent mass reduction to their
entire fleet, direct injection to one of their engines, and making their other engine hybrid is the
most cost-effective strategy under Alternative 5.
Note that without a change in the work factor or fleet mix, a manufacturer will face the
same MY2030 standard under Alternatives 3 and 4, and a more stringent standard under
Alternative 5. However, by applying 5 percent mass reduction in Alternative 4, Nissan is able to
reduce their standard by .27 MPG, and by applying 10 percent mass reduction in Alternative 5 to
have the same MY2030 standard under Alternatives 3 and 5. The result is that the CAFE level
for Nissan is highest under Alternative 2, where direct injection of their entire fleet is the most
cost-effective compliance strategy. We assume that manufacturers are able to make technologies
more cost-effectively the longer they are on the market—this is called "learning." A likely

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
reason that the model prefers direct injection in Alternative 3 but not in Alternatives 4 and 5, is
that the longer horizon of the stringency increase (until MY2027) results in direct injection that
is more cost-effective than the shorter time span of Alternatives 4 and 5.
Table 10-37 Summary of Impacts on Nissan HD Fleet by Alternative (vs. Alternative lb)
Alternative
2
3
4
5
Alternative Stringency
Annual Increase in
Stringency Beginning
in MY 2021
2.0%
2.5%
3.5%
4.0%
Increases Until
MY2025
MY2027
MY2025
MY2025
Total Increase in
MY2030 Stringency
Relative to Final Phase
1 Standards a
9.6%
16.2%
15.1%
16.2%
Estimated Average Fuel Economy (miles per gallon)
Required in MY 2030
19.65
21.19
20.92
21.19
Achieved in MY 2030
19.63
23.12
21.05
21.46
Average Fuel Consumption (gallons/100 miles)
Required in MY 2030
5.09
4.72
4.78
4.72
Achieved in MY 2030
5.09
4.32
4.75
4.66
Estimated Average Greenhouse Gas Emissions (grams per mile)
CO 2 Required in MY
2030
452
419
425
420
CO 2 Achieved in MY
2030
453
384
422
414
Technology Penetration in MY 2030 (percent)
WT and/or VVL
100
100
100
100
Cylinder Deactivation
0
0
0
0
Direct Injection Engine
51
100
51
51
Turbo Charged Engine
51
100
51
51
8 Speed Auto. Trans.
100
100
100
100
EPS, Accessories
37
100
100
100
12V Stop-start
0
0
0
49
Hybridization
0
0
0
0
Aero. Improvements
0
100
100
100
Mass Reduction (vs. No-Action)

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
Curb Weight Mass
Reduction (lb.)
0
0
307
615
Mass Reduction
(percent of curb weight)
0
0
5
10
Note:
a This increase in stringency is based on the estimated percentage change in fuel consumption (gal/lOOmi)
stringency projected by the model for the MY2030 fleet under the final Phase 2 standards relative to the
continuation of Phase 1 standards. Note that if manufacturers' have applied mass reduction to an individual
vehicle model in the CAFE model that this will increase the work factor of that vehicle in the model, and
make the individual target less stringent. Thus, where any mass reduction is applied in the model, the total
increase in stringency of the fleet presented here will be lower than the total stringency increase of the fleet
if no mass reduction were applied.
Figures Figure 10-30and Figure 10-31show the total and average incremental technology
costs for Nissan across the different regulatory alternatives. Nissan applies technology in all
alternatives in MY2021; this is a redesign year for much of their fleet. As might be expected,
they incur less technology cost in less stringent scenarios at this redesign. However, under
Alternative 3 they apply more technology in MY2029, making their marginal technology costs
under Alternative 3 for MY2029 and after higher than the marginal technology costs under
Alternative 4. They incur less technology costs in the early years and more in MY's 2029 and
beyond. In order to explain why the model predicts this action of Nissan it is useful to look at the
cumulative total incremental costs in Figure 10-32.
Alternative 2 - — -Alternatives 	Alternative 4 — • -Alternatives
50
I			
• * _
I
r - -
	 . /
			
'/
o
oooooooooooooooo
NfMfMfMfMfMfMfMNfMfMfMfMfMfMfM
Model Year
Figure 10-30 Total Technology Cost Increase for Nissan by Model Year and Alternative
45
Q)
ro 40
 vv
o 3 30
u o
>  r-
O O
o —
u
01
(0
4->
o
25
- 20
15
10
5

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
¦Alternative 2	Alternatives
Alternative 4
-Alternative 5
3000
01
re
01
2500
g 2000
u
> w
cuo ro
O rH 1500
o °
= fM
i;
t - -
u
01
1000
01
00
(0
01
>
<
500
lO
ID

00
CD
O
*—1
fsj
ro
•3-
lO
lO

00
CD
O
*—1
*—1
*—i
*—1
*—1
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
ro
O
o
o
o
O
O
O
O
O
O
O
O
o
O
O
O
fsj
fsj

fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
Model Year
Figure 10-31 Average Technology Cost Increase for Nissan by Model Year and Alternative
By incurring less technology cost early, and more technology cost later, Nissan has a
lower cumulative total cost for MY's 2016-2030 under Alternative 3 than Alternative 4. The
total cumulative cost for MY's 2016-2030 of Alternative 2 is $86 million, $178 million for
Alternative 3, $258 for Alternative 4, and $387 for Alternative 5. Since Nissan is trying to
minimize their total cost under all model years, and not their marginal cost under any single
model year, the model chooses a compliance strategy in this case which shows higher marginal
costs for Nissan in Alternative 3 than 4 for some model years, but lower cumulative total costs
over all model years.

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
• Alternative 2	Alternative 3
Alternative 4
-Alternative 5
Q)



c

4-*

l/>

o

U

>

CUO
¦to-
_o
rn


o
O
c
CM
-C
u
C
Q)
o
H
—


(O
E
o

H

m

>



_ro

3

E

3

U

450
400
350
300
250
200
150
100
50
lO
ID

00
CD
O
*—1
fsj
ro
•3-
lO
lO

00
CT)
O
*—1
*—1
*—i
*—1
*—1
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
ro
O
o
o
o
O
O
O
O
O
O
O
O
O
O
O
O
fsj
fsj

fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
fsj
Model Year
Figure 10-32 Nissan Cumulative Technology Cost Increase by Model Year and Alternative
Nissan's first redesign is in MY 2020, and they do not have another redesign scheduled
until 2029. Under Alternative 4 and 5 all of their technological application is done in MY 2020,
but under Alternative 3 the application can be spread out between the two redesign cycles.
NHTS A judges the short lead time to apply technology would make Alternatives 4 and 5 beyond
maximum feasibility for Nissan because it puts the burden of all technological application on the
MY 2020 redesign. Substantial resources and costs would be required to do so or to add another
vehicle redesign between MY 2020 and MY 2029. Since manufacturers must spread out their
capital for such deployment endeavors between the light and heavy duty fleets, the ability to
spread costs between model years is important to consider.
10.2.4.5 Daimler
Table 10-38 shows a MY2030 summary for Daimler. Daimler came into the analysis with
all of their fleet using 8-speed automatic transmissions. Their initial CAFE level in MY2020 of
25.68 was sufficient to meet their standard under Alternatives 2-5. Their only action to turbo-
charge all the engines in their fleet occurs in the dynamic baseline. As a result, no additional
actions or costs are incurred under any of the alternatives. For this reason, a figure of their annual
technology costs, nor their cumulative total technology costs has not been provided—if it were, it
would be a horizontal line showing zero costs for all model years.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-38 Summary of Impacts on Daimler HD Fleet by Alternative (vs. Alternative lb)
Alternative
2
3
4
5
Alternative Stringency
Annual Increase in
Stringency Beginning in
MY 2021
2.0%
2.5%
3.5%
4.0%
Increases Until
MY2025
MY2027
MY2025
MY2025
Total Increase in
Stringency Relative to Final
Phase 1 Standards a
9.7%
16.3%
16.3%
18.4%
Estimated Average Fuel Economy (miles per gallon)
Required in MY 2030
22.88
24.69
24.69
25.32
Achieved in MY 2030
25.68
25.68
25.68
25.68
Average Fuel Consumption (gallons/100 miles)
Required in MY 2030
4.37
4.05
4.05
3.95
Achieved in MY 2030
3.89
3.89
3.89
3.89
Estimated Average Greenhouse Gas Emissions (grams per mile)
CO 2 Required in MY 2030
445
413
412
402
CO 2 Achieved in MY 2030
396
396
396
396
Technology Penetration in MY 2030 (percent)
VVT and/or VVL
0
0
0
0
Cylinder Deactivation
0
0
0
0
Direct Injection Engine
0
0
0
0
Turbo Charged Engine
100
100
100
100
8 Speed Auto. Trans.
100
100
100
100
EPS, Accessories
0
0
0
0
12V Stop-start
0
0
0
0
Hybridization
0
0
0
0
Aero. Improvements
0
0
0
0
Mass Reduction (vs. No-Action)
Curb Weight Mass
Reduction (lb.)
0
0
0
0
Mass Reduction (percent of
curb weight)
0
0
0
0

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Note:
a This increase in stringency is based on the estimated percentage change in fuel consumption (gal/lOOmi)
stringency projected by the model for the MY2030 fleet under the final Phase 2 standards relative to the
continuation of Phase I standards. Note that if manufacturers' have applied mass reduction to an individual
vehicle model in the CAFE model that this will increase the work factor of that vehicle in the model, and
make the individual target less stringent. Thus, where any mass reduction is applied in the model, the total
increase in stringency of the fleet presented here will be lower than the total stringency increase of the fleet
if no mass reduction were applied.
10.2.5 Summary of Consumer/Operator Impacts
Table 10-39 summarizes the impacts of the regulation on the consumer/operator of the
heavy-duty vehicles. Consumers of more fuel efficient vehicles will benefit in several ways:
they will spend less on fuel to operate vehicles for the same amount of travel, some will drive
more because their per-mile travel costs less, and they will spend less time refueling vehicles. In
order to estimate the fuel savings for each regulatory alternative, future gasoline prices must be
predicted and the rebound effect (per-mile elasticity of operating a vehicle) must be assumed to
account for the cost of additional driving. In the main analysis, the rebound effect is assumed to
be 10 percent, so that, for example, a 10 percent reduction in the per-mile travel costs will result
in a lpercent increase in the amount of miles driven. Since the literature has also supported other
rebound effects, NHTSA tests several sensitivity cases assuming different rebounds: 5 percent,
15 percent, and 20 percent. Based on the average miles driven of 2b/3 vans and trucks, the
expected lifetime fuel savings for a heavy-duty vehicle under the preferred scenario is $3636.
The other benefits of to the consumer of increasing fuel economy are increased mobility
and a decreased amount of time spent refueling the vehicle. Because increasing the efficiency of
a vehicle makes per-mile travel cheaper to the operator, consumers of these vehicles can travel
more, at less than the total amount they are willing to pay—this increase in welfare that is not
accounted for by the cost of travel is the consumer surplus. The estimated mobility benefit is
$394 under the preferred alternative. The avoided time refueling also has a value. In order to
estimate this value we make several assumptions outlined in more detail of the NPRM
description of the model assumptions (Section E). Over the lifetime of a MY2030 vehicle, we
estimate the refueling surplus at $94 under the preferred alternative.
It is also important to note that the average manufacturer costs will not be spread
proportionally across the fleet—some vehicles will have incurred more technology costs than
others. How manufacturers distribute costs among models will largely depend on the elasticity
of particular models and the importance of fleet mix in meeting standards and on total profits.
Without privy to this sort of information, we use average technology cost increase as a proxy for
measuring the industry and consumer costs across different scenarios. The average technology
cost increase is $1472 under the preferred alternative. We assume that all of this cost will be
passed onto the consumer in the form of an increase in price. However, we also consider that an
increase in price will have other costs to the operator of the vehicle.
More expensive vehicles will have higher taxes/fees associated with their purchase, will
be more expensive to insure (these costs are related to the purchase price or value of a vehicle)
and will be more expensive to finance (higher loan values will be taken out which result in

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
higher amounts paid in total interest). The total additional costs to the average consumer from
the sum of these sources is $589 under the preferred alternative. It is important to keep in mind
that the additional cost to finance a more expensive vehicle will have different effects depending
on the budget constraint of the consumer. For consumers who are budget-constrained, they will
finance more of the vehicle and the costs of financing will be higher for these already-
constrained consumers. For consumers who do not have to finance the vehicle, there will be no
costs—and therefore, no additional costs—to finance the vehicle. Since budget-constrained
consumers likely have a more elastic demand for new vehicles, the increase in price and the
heterogeneous increase in financing might work in the same direction to price proportionally
more of the most budget-constrained consumers out of the new vehicle market.
Considering all the costs and benefits the standards will have to the consumer, the result
is a net benefit to the consumer under all the considered alternatives. The net benefit to the
consumer is $2063 under the preferred alternative, higher than the net benefit under alternative
4. The payback period is another measure of the effect of the rule on consumers-for all
alternatives the payback period is under 3 years—suggesting that consumers that own vehicles
for at least 3 years will receive a net benefit from the preferred regulatory action.
Table 10-39 Summary of Consumer/Operator Impacts for MY 2030 (vs. Alternative lb)
Alternative
2
3
4
5
A
ternative Stringency
Annual Increase
2.0%
2.5%
3.5%
4.0%
Increases Until
MY2025
MY2027
MY2025
MY2025
Average Value of Lifetime Fuel Savings, $2013 (vs. No-Action)
Pretax
$1713
$3256
$3229
$3804
Tax
$200
$381
$377
$448
Total
$1913
$3636
$3607
$4252
Average Value of Additional Economic Benefits, $2013 (vs.
^Jo-Action)
Mobility Increase
$220
$394
$390
$453
Avoided Refueling
$49
$94
$93
$112
Average New Vehicle Purchase
'vs. No-Action)
Price Increase ($)
$496
$1472
$1481
$1893
Additional Costs ($)a
$198
$589
$592
$757
Payback (months) b
20
33
33
38
Net Lifetime Consumer/Operator Benefits (vs. No-Action)
Total Net Benefit ($) ||$ 1488
$2063 $1989 $2167
Notes:
a Additional Costs include additional taxes, fees, maintenance costs, financing costs, and insurance costs
incurred under the regulatory alternatives.
b The payback period from the consumer perspective uses a 7% discount rate of retail fuel savings starting
at the time of purchase. The cost increases paid back include: technology costs, maintenance costs, taxes,
and fees.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
10.2.6 Summary of Societal Impacts
Table 10-40 summarizes the overall societal impacts of the regulation under different
scenarios (relative to the lb baseline). Net social benefits increase with the stringency of the
standards. The net benefits for the preferred alternative are $18.8 billion. The largest benefit of
the program comes in the form of fuel savings. The fuel savings reported above do not include
fuel tax savings, as taxes are considered a transfer, and not a loss, of societal well-being. The
fuel savings are associated with a fuel security externality, which monetizes the economic risk
associated with potential fuel price spikes—as fewer gallons of oil are necessary for
transportation, this risk decreases. The carbon externality represents the reduced cost of carbon
damage when fuel economy increases (and carbon emissions decrease), and is also related
directly with fuel savings.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-40 Summary of Lifetime Total Societal Impacts of MY's 2015-2029 (vs. Alternative lb)
Alternative
2
3
4
5
Alternative Stringency
Annual Increase
2.0%
2.5%
3.5%
4.0%
Increases Until
MY2025
MY2027
MY2025
MY2025
Fuel Purchases vs. No-Action (billion 2013$)
Pretax Savings
$11.1
$17.8
$20.2
$22.7
Fuel-Related Externalities vs. No-Action (billion 2013$)
Energy Security
$0.7
$1.2
$1.4
$1.5
CO 2 Emissions
$2.4
$3.8
$4.4
$4.9
VMT-Related Externalities vs. No-Action (billion 2013$)
Driving Surplus
$1.3
$2.0
$2.3
$2.5
Refueling Surplus
$0.3
$0.6
$0.6
$0.7
Congestion
-$0.3
-$0.5
-$0.5
-$0.6
Crashes
-$0.2
-$0.2
-$0.3
-$0.3
Noise
$0.0
$0.0
$0.0
$0.0
Fatalities
-$0.7
-$0.3
-$0.4
$0.7
Criteria Emissions
$0.7
$1.2
$1.4
$1.5
Vehicle Purchase/Operating Costs vs. No-Action (billion 2013$)
Technology Costs
$2.9
$6.5
$7.7
$10.2
Maintenance Costs
$0.1
$0.3
$0.3
$0.5
Cost-Benefit Summary vs. No-Action (billion 2013$)
Total Social Cost
$4.2
$7.8
$9.2
$11.6
Total Social Benefit
$16.5
$26.6
$30.3
$34.5
Net Social Benefit
$12.3
$18.8
$21.1
$22.9
Increasing fuel economy decreases the cost of per-mile travel. Since this reduction in the
cost of travel results in an increase of total travel, it also results in an increase of externalities
associated with increased total VMT. Of these, the driving surplus represents the societal net
increase in benefit from increased mobility consumer surplus—the sum of the benefit to all
operators of increased travel which is not captured by the total cost of travel. Defined from the
societal perspective, the refueling benefit is the sum of all the value of the time saved on
refueling by increasing the average fuel efficiency of the heavy duty fleet. Congestion represents
the societal cost of increases in congestion on the roads—the lost value of additional time spent
in traffic. The crash externality is the cost of the damage done by the additional crashes that will

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
happen with more VMT exposure, and the noise externality represents the cost of a change in
noise related to increases in vehicle travel (in this analysis, it is negligible for all alternatives).
Some VMT-related externalities are not always positive or negative, but depend on the
stringency of the standards. For this analysis the criteria pollutant externality is always a benefit,
but this need not be the case. Reduction in overall fuel consumed reduces emissions associated
with production and distribution of fuels. Increases in VMT will result in more emission of
vehicle criteria pollutants and more associated damages. However, increasing fuel-economy
though vehicle technologies, such as aerodynamics, mass reduction and improved tire rolling
resistance, will result in a decrease in vehicle emissions of and damages from criteria pollutants.
Shifts in technologies towards electric and hybrid-electric alternatives can increase the emissions
of certain pollutants, and reduce the emissions of others. The stringency increases considered in
the heavy-duty analysis do not require these technologies to penetrate the market at such a level
that this is visible in the results. For these reasons the externality associated with changes in
criteria pollutant emissions is always positive for this analysis.
The vehicle mass reduction in HD pickup and vans is estimated to reduce the net
incidence of highway fatalities. By reducing mass on some HD pickup and vans, the fatality rate
associated with crashes involving at least one HD pickup or van vehicles decreases. However,
the analysis anticipates that the indirect effect of the proposed standards, by reducing the
operating costs, would lead to increased travel by HD pickups and vans and, therefore, more
crashes involving these vehicles. The sign of the fatality externality varies with the stringency of
the standards. Over the lifetime of MY's 2016-2029, for Alternatives 2 it is estimated
approximately 120 additional fatalities could occur relative to the 30,200 heavy-duty crash-
related fatalities in the baseline. For Alternatives 3 and 4 we estimate approximately 50
additional fatalities relative to the no-action alternative. The additional risk of fatality is
represented as a social cost in Alternatives 2-4. For Alternative 5 we estimate approximately 110
fewer fatalities (represented as a positive externality). For Alternatives 2-4, the effect of
removing mass from the heavier vehicles is less than the effect of increased VMT-exposure; for
Alternative 5, it is larger, and the alternative could result in a decrease of fatalities.
The major direct costs of the program are increased technology costs and costs associated
with the resultant increase in new vehicle prices and changes in technologies. The sum of
technology costs across the industry increase under all increases of stringency, as do the
increases in associated additional costs. Additional costs include: additional costs of
maintenance associated with certain technologies. These costs will mostly be borne by the
consumer, and paid back in the form of fuel savings.
10.2.7 Summary of Environmental Impacts
In addition to modeling the societal impacts from a monetary standpoint, the CAFE
model also considers the absolute change in the physical emissions of various criteria pollutants
across the Alternatives. Table 10-41 summarizes the total environmental impacts from increased
fuel efficiency of MYs 2016-2030, taking into consideration the reduction in emissions from
increased efficiency, the additional emissions associated with the increased VMT from cheaper
per-mile travel, and changes in emissions due to the production and distribution of heavy-duty
vehicles. Across all scenarios, the absolute reduction in emissions increases. For context, the

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
percentage change of emissions relative to the baseline emission levels is also provided. The
proportional reduction in criteria pollutants greatly varies; the greenhouse gases—carbon
dioxide, methane, and nitrous oxide—as well as the criteria pollutants—sulfur dioxide and diesel
particulate matter—show the largest proportional reductions across all scenarios.
Table 10-41 Summary of Lifetime Emission Impacts of MY's 2015-2029 (vs. Alternative lb)
Alternative
2
3
4
5
Annual Increase
2.0%
2.5%
3.5%
4.0%
Increases Until
MY2025
MY2027
MY2025
MY2025
Greenhouse Gas Emissions Reductions vs. No-Action
CO 2 (mmt)
66
107
120
135
CH4 and N2O (tons)
97,925
160,044
180,557
202,666
Greenhouse Gas Emissions Percent Reduction vs. No-Action
O
O
3.8%
6.1%
6.9%
7.7%
CH4 andN20
0.7%
1.2%
1.3%
1.5%
Other Emissions Absolute Reduction vs. No-Action
CO (tons)
13,747
22,828
26,375
29,589
VOCandNOx
(tons)
33,324
56,100
63,237
70,957
PM25 (tons)
1,320
2,213
2,498
2,806
SO 2 (tons)
10,713
17,877
20,172
22,669
Air Toxics (tons)
53
75
84
94
Diesel PM10 (tons)
2,357
3,944
4,450
5,004
Other Emissions Percent Reduction vs. No-Action
CO
0.2
0.4
0.4
0.5
VOCandNOx
1.6
2.8
3.1
3.5
PM25
1.9
3.3
3.7
4.1
S02
3.7
6.2
6.9
7.8
Air Toxics
0.2
0.2
0.2
0.3
Diesel PM10
3.5
5.8
6.5
7.3
10.2.8 Sensitivity Analysis Evaluating Different Inputs to the NHTSA CAFE Model
This section describes some of the principal sensitivity results, obtained by running the
various scenarios describing the policy alternatives with alternative inputs. OMB Circular A-4

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
indicates that "it is usually necessary to provide a sensitivity analysis to reveal whether, and to
what extent, the results of the analysis are sensitive to plausible changes in the main assumptions
and numeric inputs."LL Considering this guidance, a number of sensitivity analyses were
performed using analysis Method A to examine important assumptions and inputs, including the
following, all of which are discussed in greater detail in the accompanying RIA:
1.	Payback Period: In addition to the 0 and 6 month payback periods discussed above, also
evaluated cases involving payback periods of 12, 18, and 24 months.
2.	Fuel Prices: Evaluated cases involving fuel prices from the AEO 2015 low and high oil
price scenarios. (See AEO-Low and AEO-High in the tables).
3.	Fuel Prices and Payback Period: Evaluated one side case involving a 0 month payback
period combined with fuel prices from the AEO 2015 low oil price scenario, and one side
case with a 24 month payback period combined with fuel prices from the AEO 2014 high
oil price scenario.
4.	Benefits to Vehicle Buyers: The main Method A analysis assumes there is no loss in
value to owner/operators resulting from vehicles that have an increase in price and higher
fuel economy. NHTSA performed this sensitivity analysis assuming that there is a 25, or
50 percent loss in value to owner/operators - equivalent to the assumption that
owner/operators will only value the calculated benefits they will achieve at 75, or 50
percent, respectively, of the main analysis estimates. (These are labeled as
75pctOwner/Operator Benefit and 50pctOwner/Operator Benefit).
5.	7 Pet Discount Rate: The main analysis results are considered using either a 0 or 3
percent discount rate. We also considered an alternative case where future savings/costs
are discounted 7 percent annually.
6.	Value of Avoided GHG Emissions: Evaluated side cases involving lower and higher
valuation of avoided CO2 emissions, expressed as the social cost of carbon (SCC).
7.	Rebound Effect: Evaluated side cases involving rebound effect values of 5 percent, 15
percent, and 25 percent. (These are labeled as 05PctReboundEffect, 15PctReboundEffect
and 25PctReboundEffect).
8.	ICM-based Markup: Evaluated a side case using a retail price equivalent (ICM) markup
factor.
9.	Mass-Safety Effect: Evaluated side cases with the mass-safety impact coefficient at the
values defining the 5th and 95th percent points of the confidence interval estimated in the
underlying statistical analysis. (These are labeled MassFatalityCoeff05pct and
MassFatalityCoeff95pct.).
10.	VMT Schedules: Evaluated side cases considering the NHTS considered in the NPRM
analysis as a high-VMT case, and another considered schedule as a low-VMT case.
11.	Strong HEVs: Evaluated a side case in which strong HEVs were excluded from the set of
technology estimated to be available for HD pickups and vans through model year 2030.
As in Section VIC. (8), this "no SHEV" case allowed turbocharging and downsizing on
all GM vans to provide a lower-cost path for compliance.
Table 10-42, below, summarizes key metrics for each of the cases included in the
sensitivity analysis using Method A for the alternative. The table reflects the percent change in
the metrics (columns) relative to the main analysis, due to the particular sensitivity case (rows)
LL Available at http://www.whitehouse.gov/omb/circulars a004 a-4/.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
for the alternative 3. For each sensitivity run, the change in the metric can we described as the
difference between the baseline and the preferred alternative for the sensitivity case, minus the
difference between the preferred alternative and the baseline in the main analysis, divided by the
difference between the preferred alternative and the baseline in the main analysis. Or,
„ , , ,, ^ . Adit sen case ~ ^Alt main run . „ „
Table Metric = 			 ¦ 100
A<4it main run
Each metric represents the sum of the impacts of the preferred alternative over the model
years 2015-2029, and the percent changes in the table represent percent changes to those sums.
More detailed results for all alternatives are available in the accompanying RIA Chapter 10.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-42 Sensitivity Analysis Results from CAFE Model in the HD Pickup and Van Market Segment
using Method A and versus the Dynamic Baseline, Alternative lb (2.5 % growth in stringency: Cells are
percent change from base case)a
Sensitivity Case
Fuel
O
O
Fuel
Social
Social
Social Net

Savings
(gallons)
Savings
(MMT)
Savings
($)
Costs
($billion)
Benefits
($billion)
Benefits
($billion)
0 Month Payback
8.4%
8.0%
7.7%
8.0%
7.8%
7.7%
12 Month Payback
-13%
-14%
-15%
-2.8%
-14%
-19%
18 Month Payback
-30%
-31%
-32%
-16%
-31%
-38%
24 Month Payback
-47%
-47%
-48%
-32%
-48%
-54%
AEO-Low
-5.4%
-5.8%
-31%
-19%
-26%
-29%
AEO-High
-27%
-28%
18%
-2.8%
13%
20%
AEO-Low, 0






Month Payback
35%
33%
33%
42%
34%
30%
AEO-High, 24






Month Payback
-50%
-50%
-51%
-37%
-51%
-57%
7pct Discount Rate
0.0%
0.0%
-41%
-31%
-35%
-37%
50pct
Owner/Operator
Benefit
0.0%
0.0%
-50%
0.0%
-34%
-48%
75pct
Owner/Operator
Benefit
0.0%
0.0%
-25%
0.0%
-17%
-24%
Low SCC
0.0%
0.0%
0.0%
0.0%
-11%
-16%
High SCC
0.0%
0.0%
0.0%
0.0%
8.2%
12%
Very High SCC
0.0%
0.0%
0.0%
0.0%
30%
43%
5pct Rebound
4.6%
4.6%
4.6%
-13%
0.37%
5.5%
15pct Rebound
-4.6%
-4.6%
-4.6%
12%
-0.37%
-5.5%
25pct Rebound
-14%
-14%
-14%
37%
-1.1%
-17%
5th Percentile Mass






Fatality Coefficient
0.0%
0.0%
0.0%
-11%
0.0%
4.6%
95th Percentile






Mass Fatality
Coefficient
0.0%
0.0%
0.0%
15%
0.0%
-6.0%
No SHEV-P2's
0.18%
0.29%
0.29%
-1.3%
0.26%
0.88%
Non-C02eq GHG






Values
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
ICM-Based Mark-






up
-5.7%
-6.0%
-6.1%
-16%
-6.0%
-1.8%
High VMT
8.6%
7.4%
5.9%
0.11%
6.2%
8.7%
Low VMT
-7.7%
-8.3%
-8.0%
-14%
-7.8%
-5.4%
Note:
aFor an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the less dynamic
baseline, la, and more dynamic baseline, lb, please see Section X.A.I.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
For some of the cases for which results are presented above, the sensitivity of results to
changes in inputs is simple, direct, and easily observed. For example, changes to valuation of
avoided GHG emissions impact only this portion of the estimated economic benefits;
manufacturers' responses and corresponding costs are not impacted. Similarly, a higher discount
rate does not affect physical quantities saved (gallons of fuel and metric tons of CO2 in the
table), but reduces the value of the costs and benefits attributable to these standards in an
intuitive way. Higher rebound results in fewer volumetric fuel savings and social net benefits, as
drivers are assumed to be more responsive in their driving habits to changes in the cost per mile
of travel. Some other cases warrant closer consideration:
First, cases involving alternatives to the reference case involving voluntary
overcompliance of technologies that pay back in six-months involve different degrees of fuel
consumption improvement. Increasing the length of the payback period assumption for voluntary
overcompliance amounts to increasing fuel economy improvements in the absence of the rule
(the baseline), and manufacturers are compelled to add less technology in order to comply with
the standards (in the regulatory alternatives). Because all estimated impacts of these standards
are shown as incremental values relative to this baseline, longer voluntary overcompliance
payback periods correspond to smaller estimates of incremental impacts.
Table 10-43 shows the effect of varying the voluntary overcompliance assumption from
the consumer perspective. The baseline over-compliance payback period is as described above—
the number of months within which a technology must pay back to the consumer in the form of
undiscounted retail fuel savings for a manufacturer to voluntarily apply that technology without
regulatory action. The incremental per-vehicle technology cost is the average additional cost of
technology applied to MY 2030 vehicles under the final regulation (incremental to the baseline)
of each sensitivity case. The per-vehicle lifetime fuel savings is the average lifetime retail value
of fuel savings under each sensitivity case discounted at 7 percent annually starting at the time of
purchase (MY 2030). Compliance payback period is the number of months of ownership it
would take the average consumer to recoup the additional technology costs in discounted fuel
savingsMM.
As can be seen, the baseline voluntary overcompliance assumption changes how much of
the technology costs and fuel savings are attributed to the regulation; both fewer fuel savings and
fewer technology costs are attributed to the regulatory alternative as the payback period defining
voluntary overcompliance increases. Further, because the model only applies the technologies
with the shortest payback periods (the most cost-effective technologies) in the baseline, the fuel
savings decrease at a greater proportion than the technology costs. The result is that the payback
period of the regulatory alternative increases (and at an increasing rate) as manufacturers are
assumed to apply more technology in the baseline.
1414 This is based on the VMT schedules of average miles driven by age of MDHD pickups and vans and AEO fuel
price projections.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-43 Sensitivity Analysis of the Voluntary Overcompliannce Assumption on Compliance Payback
Period and Key Consumer Impacts for the MY 2030 MDHD Fleet
Baseline Over-
compliance
Payback (months)
Incremental
Per-Vehicle
Technology Cost
Per-Vehicle
Lifetime
Fuel Savings
Technology Cost
Payback Period
(months)a
0
$1,471
$3,966
28
6
$1,472
$3,636
31
12
$1,317
$3,031
33
18
$1,214
$2,556
38
24
$944
$1,684
45
Note:
a Here the payback period uses a 7% discount rate of retail fuel savings starting at the time of purchase and
only considers the additional costs of technology application.
Cases involving different fuel prices similarly involve different degrees of fuel economy
improvement in the absence of the standard, as more, or less, improvement occurs as a result of
more, or fewer, technologies appearing cost effective to owner/operators. Low fuel prices
change the amount of fuel savings for each technology, since the choice in technology
application also involves both the size of the cost and the fuel savings, lower fuel prices can
change the rank of the technologies. Under low fuel prices, the model applies fewer SHEV-P2's.
The result is a reduction in volumetric fuel savings, and an even larger reduction in monetary
fuel savings, because the fuel savings are worth less. There is also a reduction in social costs,
and social net benefits. Higher fuel prices correspond to reductions in the volumetric fuel
savings attributable to these standards as, but lead to increases in the value of fuel saved (and net
social benefits) because each gallon saved is worth more when fuel prices are high.
The low price and 0-month payback case leads to a significant increase in volumetric
savings compared to the main analysis. Note that the fuel savings are higher than in the 0-month
payback case alone. Part of the reason for this is that the lower fuel price case takes into
consideration that when fuel prices are lower, consumers buy more heavy-duty vehicles (this is
estimated from the AEO2015 low fuel price case). Another piece of the explanation is that the
lower fuel prices result in a different technology cost-effectiveness ranking of technologies, and
that the 0 month payback baseline results in no voluntary over compliance in the baseline.
Different technologies are picked than in the 0 month pay back sensitivity alone, and the most
cost effective that would have been applied in the baseline, are now attributed to the preferred
alternative. Similarly, the high price and 24-month payback case results in large reductions to
volumetric savings that can be attributed to these standards because more is applied in the
baseline. Further, the presence of high fuel prices is not sufficient to lead to increases in either
the dollar value of fuel savings or net social benefits.
The case which involves the VlUS-based VMT schedules (the high VMT case) results in
greater volumetric fuel and GHG-savings attributable to the standards. Under this case the
higher estimate of VMT results in more fuel consumption in the baseline, and a higher absolute
change in fuel consumption when fuel-saving technologies are applied in the preferred
alternative. These higher amount of gallons saved, results in more monetary fuel savings,
comparable social costs, and an increase in overall net social benefits attributed to the standards.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The low-VMT schedule, developed as an alternative to the adopted VMT-schedule from the
IHS/Polk odometer readings, results in lower volumetric fuel consumption and GHG reductions
under the preferred alternative. Lower VMT estimates result in less fuel consumption in the
baseline, and a lower absolute change in fuel consumption under the preferred alternative. This
schedule attributes lower costs to the standards—the lower fuel savings under the low-VMT
schedule changes the technology application decisions of the model, since fewer fuel savings are
considered in measure the cost-effectiveness of technologies. The result is lower absolute
technology costs, but also lower social net benefits.
The case which makes SHEV-P2's unavailable involves relatively small increases to
volumetric fuel savings and CO2 reductions—not surprising, since SHEV-P2's play only a minor
role in the compliance strategy of the preferred alternative in the central analysis. These small
increases in fuel savings are associated with small increases in social benefits, slightly larger
proportional increases in social costs, but still result in a small increase in social net benefit.
The case that uses the ICM mark-up methodology rather than the RPE methodology
results in a reduction of volumetric fuel savings and GHG reductions. The reduction in fuel
savings is accompanied by a reduction in monetary fuel savings, social benefits, social costs, and
social net benefits. This is likely due to shifts in technology applications due to different costs
mark-ups associated with different types of technologies under the ICM mark-up methodology.
If, instead of using the values in the main analysis, each sensitivity case were itself the
main analysis, the costs and benefits attributable to the final rule will be as they appear in Table
10-44, below.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-44 Costs and Benefits of Standards for MY 2015-2029 HD Pickups and Vans under Alternative
Assumptions
Sensitivity Case
Fuel
Savings
(billion
gallons)
C02
Reduction
(MMT)
Fuel
Savings
($billion)
Social
Costs
($billion)
Social
Benefits
($bi 1 lion)
Net Social
Benefits
($billion)
6 Month Payback
9.2
110
18
7.8
27
19
0 Month Payback
10
120
19
8.2
28
20
12 Month Payback
8.0
92
15
7.3
22
15
18 Month Payback
6.4
74
12
6.4
18
12
24 Month Payback
4.9
56
9.3
5.2
14
8.5
AEO-Low
8.7
100
12
6.1
19
13
AEO-High
6.7
77
21
7.3
30
22
AEO-Low, 0 Month
Payback
12
140
24
11
35
24
AEO-High, 24 Month
Payback
4.7
53
8.8
4.8
13
8.0
7pct Discount Rate
9.2
110
11
5.2
17
12
50pct Owner/Operator
Benefit
9.2
110
8.9
7.5
17
9.7
75pct Owner/Operator
Benefit
9.2
110
13
7.5
22
14
Low SCC
9.2
110
18
7.5
23
16
High SCC
9.2
110
18
7.5
28
21
Very High SCC
9.2
110
18
7.5
34
27
5pct Rebound
9.7
110
19
6.6
26
20
15pct Rebound
8.8
100
17
8.5
26
18
25pct Rebound
8.0
92
15
10
26
16
5th Percentile Mass
Fatality Coefficient
9.2
110
18
6.7
26
19
95th Percentile Mass
Fatality Coefficient
9.2
110
18
8.7
26
18
No SHEV-P2's
9.3
110
18
7.5
26
19
Non-C02eq GHG
Values
9.2
110
18
7.5
26
19
ICM-Based Mark-Up
8.7
100
17
6.3
25
18
High-VMT
10
110
19
7.6
28
20
Low-VMT
8.5
98
16
6.5
24
18
10.2.9 Discussion of the Maximum Feasibility of the Adopted Standards
As noted above, EPCA and EISA require NHTSA to "implement a commercial medium-
and heavy-duty on-highway vehicle and work truck fuel efficiency improvement program
designed to achieve the maximum feasible improvement" and to establish corresponding fuel

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
consumption standards "that are appropriate, cost-effective, and technologically feasible.',NN In
order to determine which of the regulatory alternatives meets the requirements of the statute
NHTSA has considered both the modeling results of "Method A" and comments offered on the
proposed rulemaking.
10.2.9.1 Consideration of Modeling Results
For both the NPRM and the current analysis of potential standards for HD pickups and
vans, NHTSA applied NHTSA's CAFE Compliance and Effects Modeling System (sometimes
referred to as "the CAFE model" or "the Volpe model"), which DOT's Volpe National
Transportation Systems Center (Volpe Center) developed, maintains, and applies to support
NHTSA CAFE analyses and rulemakings. NHTSA used this model in its Method A analysis to
evaluate regulatory alternatives for Phase 2 standards applicable to HD pickups and vans, and
used results of this analysis to inform its selection of the regulatory alternative that will achieve
the maximum feasible improvement in HD pickup and van fuel efficiency. This analysis,
includes several updates to the model and to accompanying inputs, as discussed above in this
section.
In the proposal, the agencies proposed to adopt Alternative 3 from among the five
regulatory alternatives under consideration.00 As discussed in the NPRM, the agencies found
that Alternative 2 would unduly forego significant fuel savings and avoided GHG emissions, and
that Alternative 5 could involve rapid and early cost increases and necessitate significant
application of the most advanced technologies considered by the agencies, 80 FR 40494-95. The
agencies have estimated the cost and efficacy of fuel-saving technologies assuming performance
and utility will be held constant or improved. In particular, we have assumed payload will be
preserved (and possibly improved via reduced vehicle curb weight); however, some fuel-saving
technologies, such as hybrid electric vehicles, could reduce payload via increased curb weight
(due to the added electrical machine, batteries and controls, and because of the physical size of
those components). If the increase in weight from the hybrid system is not offset with a weight
reduction elsewhere in the vehicle, the payload capability will be reduced resulting in lost utility
but also an increase in stringency due to changes in work factor. Further, it is also possible that
applications such as vans where the advanced technologies of downsized gasoline and diesel
engines could be used in conjunction with strong hybridization, extended high power demand
resulting from a vehicle at full payload or towing, certain types of hybrid powertrains could
experience a temporary loss of towing capacity if the capacity of the hybrid's energy storage
device (e.g., batteries, hydraulic accumulator) is insufficient for the extended power demand
required to maintain expected vehicle speeds.
The Method A analysis shows in the short term, MY 2017 - 2021 timeframe, that there
are significant differences in the rate at which technologies would need to be applied among the
alternatives. NHTSA believes the rates of technology application require for Alternatives 4 and
5 are beyond maximum feasible when considering the availability of manufacturers' resources
m 49 USC 32902(k)(2).
00 These Alternatives are defined in Section C(6).

-------
*** E.O. 12866 Review — Revised —Do Not Cite, Quote, or Release During Review ***
and capital to implement the technologies in that timeframe, and that Alternatives 4 and 5 would
not provide adequate lead time for the industry to fully address reliability considerations.
Like the NPRM analysis (i.e. the Method B analysis), Method A indicates Alterative 4
would achieve little benefit beyond that achieved by Alternative 3. For example, as shown in the
following graph of estimated total fuel consumed by HD pickups and vans over time under the
various regulatory alternatives, outcomes under Alternative 4 are nearly indistinguishable from
those under Alternative 3. By 2030, the two are less than 0.5 percent apart.
11
10
m 8
GO
-Q
C
o
4-»
CL
E
D
C
o
u
D
4
ro
3
c
c
<
/ \
/V
I
/\
I \
/ \
/ \
/
\ I
\l

^Xx
— Historical and No Action
	Alternative 2
—Alternative 3
° Alternative 4
x Alternative 5
1970 1980 1990 2000 2010 2020
Calendar Year
2030
2040
2050
Figure 10-33 Method A Annual Fuel Consumption across Regulatory Alternatives
Weighing against the small additional benefit estimated to be potentially available under
Alternative 4, NHTSA also considered the estimated additional costs. Method A analysis shows
overall incremental costs (i.e., costs beyond the No Action Alternative) under Alternative 4 to be
about 12 percent more than under Alternative 3.
As mentioned above, these estimated differences were mostly small on a relative basis.
Averaged over all model years included in the analysis, estimated incremental costs are $106
higher under Alternative 4 than under Alternative 3. For Daimler and General Motors, there is
little or no estimated difference in costs under these two Alternatives. For FCA, Ford, and
Nissan, differences are somewhat larger, averaging $120, $173, and $272, respectively.
However, as explained in greater detail above NHTSA's method A analysis shows considerably

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
greater total and average additional costs in earlier model years under Alternative 4 than under
Alternative 3.
Although NHTSA's Method Analysis also indicates that some manufacturers could need
to apply additional technology as soon as MY2016 under baseline standards defining the No-
Action Alternative, average estimated costs (versus continuation today's technology) in MY2017
are two thirds more under Alternative 4 than under the No Action Alternative.
Beyond these directly-estimated costs, the agencies also considered factors beyond those
addressed quantitatively in either the NPRM analysis or the updated analysis. In general, these
other factors reflect risk and uncertainty involved with standards for HD pickups and vans.
These risks and uncertainty appear considerably greater than for light-duty vehicles. The HD
pickup and van market has significantly fewer vehicle models than the light-duty market making
forecasting uncertainty a greater risk to compliance. All current manufacturers of HD pickups
and vans also produce light-duty vehicles. These manufacturers' light-duty offerings span wide
ranges of models, configurations, shared vehicle platforms, engines, transmissions, and design
schedules. As a result, if some specific aspects of production do not progress as initially planned
for light-duty vehicles (e.g., if mass reduction on some platform does not achieve as much
benefit as planned, or if a new engine does not perform as well as projected, or if limited
engineering resources make it necessary to delay a redesign), these manufacturers should have
ample opportunity to comply with light-duty CAFE and GHG standards by making adjustments
among other models, platforms, engines, and transmissions. This is not the case for HD pickups
and vans. Current HD PUV manufacturers offer products spanning only 1-3 platforms, at most
half a dozen engines or transmissions, and only 1-3 schedules for redesigns. As summarized
below, this provides 5-10 times less flexibility than for light-duty vehicles.
Table 10-45 MY 2015 Body and Engine Platforms by Manufacturer for Light- and Heavy-Duty Pickups

PLATFORMS
ENGINES
TRANSMISSIONS
DESIGN SCHEDULES

Light-Duty
FID PUV
Light-Duty
FID PUV
Light-Duty
HD PUV
Light-Duty
HD PUV
Daimler
12
1
29
2
20
2
18
1
FCA
15
3
24
5
21
6
24
3
Ford
9
2
22
5
27
3
18
2
General Motors
17
2
26
5
39
3
21
2
Nissan
6
1
13
2
21
2
23
1
Considering further that credits from other manufacturers are not potentially available as
for light-duty vehicles (e.g., Honda, Toyota, and some other manufacturers currently have excess
light-duty CAFE credits that could be traded to other OEMs), this means that overestimating the
industry's capability to improve fuel efficiency and reduce GHG emissions, and consequently
setting standards at too high of a level, poses a much greater compliance risk for HD PUV fleets
than for light-duty fleets. If the factors discussed here, for which the agencies are currently
unable to account in our analysis, lead manufacturers to fail to comply with the standards, then
the additional benefits of setting standards slightly higher would be lost. In the agencies'
judgment, even setting aside the somewhat higher estimated costs under Alternative 4, the very

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
small additional benefit that could be achieved under Alternative 4 do not warrant the increased
exposure to this risk.
Regarding Alternative 5, the Method A analysis shows somewhat greater benefits than
under Alternatives 3 or 4, but Alternative 5 entails considerably greater costs and dependence on
strong hybrid technology, as well as even greater exposure to the above-mentioned uncertainties
and risks. Under the Method A analysis for Alternative 5, incremental costs averaged across all
model years considered are estimated to be about $400 higher (about 46 percent) than under
Alternative 3, and that analysis shows an overall fleet application of approximately 7 percent
strong hybrids, with General Motors applying approximately 13 percent and Ford approximately
7 percent.
We have also assumed that fuel-saving technologies will be no more or less reliable than
technologies already in production. However, if there is insufficient lead-time to fully develop
new technologies, they could prove to be less reliable, perhaps leading to increased repair costs
and out-of-service time. If the fuel-saving technologies considered here ultimately involve
reliability problems, overall costs will be greater than we have estimated. Method A analysis
shows in the short term, MYs 2017 - 2021 timeframe, there are significant differences in the rate
at which technologies would need to be applied among the alternatives. Figure 10-18and Figure
10-19, above, shows the progression in average and total technology costs and the rate of
increase in those costs among the alternatives using Method A. They highlight the increases in
resources and capital that would be required to implement the technologies required to comply
with each of the alternatives, as well as the reduction in lead time to implement the technologies
which increases reliability risk. As discussed further above in the manufacturer-specific effects,
Ford and FCA are estimated to redesign vehicles in MYs 2017 and 2018 respectively, and
vehicle designs for those model years are complete or nearly complete. The next estimated
redesign for Ford is in MY 2026, and for FCA in MY 2025, and substantial resources and very
high costs would be required to add another vehicle redesign between the estimated redesign
model years to implement the technologies that would be needed to comply with those
alternatives.
10.2.9.2 Consideration of Comments
NHTSA proposed that Alternative 3 represented the maximum feasible alternative under
EISA, and EPA proposed that Alternative 3 reflected a reasonable consideration of the statutory
factors of technology effectiveness, feasibility, cost, lead time, and safety for purposes of CAA
sections 202 (a)(1) and (2). Although the agencies and commenters also found that Alternative 4
merited serious consideration, the agencies noted that Alternative 3 was generally designed to
achieve the levels of fuel consumption and GHG stringency that Alternative 4 would achieve,
but with several years of additional lead time, meaning that manufacturers could, in theory, apply
new technology at a more gradual pace, with greater reliability and flexibility.
Some comments on the proposal called for adoption of standards more stringent and/or
more rapidly advancing in stringency than those defining Alternative 3. For example, CARB
argued that Alternative 4 would, compared to Alternative 3, achieve greater benefits comparably
attractive in terms of cost effectiveness and while remaining less stringent than CAFE standards

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
for light-duty trucks.pp UCS provided similar comments, indicating further that the standards
should be technology forcing and therefore more aggressive than Alternative 4, they specifically
suggested that gasoline vehicles could achieve up to a 23.6 percent improvement in MY 2027
while diesel vehicles can achieve an 18 percent improvement.^ ACEEE similarly
recommended increasing the stringency by 7 percent in MY 2027 and that standards should
reflect increased use of cylinder deactivation, cooled EGR, and GDI and turbo downsizing in
pickups. For diesels, ACEEE commented that additional reductions were possible, based on an
estimate of 10 percent penetration of engine downsizing for pickups and 30 percent penetration
for vans in 2027, and also assuming 6 percent penetration of hybrids in diesel vans.
Citing the potential for fuel-saving technology to migrate from light-duty pickups and
vans to heavy-duty pickups and vans, CBD also called for more stringent HD pickup and van
standards that would "close the gap" with light-duty standards, as any gap allows manufacturers
to essentially choose to classify a pickup as heavy-duty to avoid more stringent requirements if it
was classified as a light-duty vehicle.1® ICCT likewise commented that the proposed standards
represent only a 2.2 and 1.6 percent year-over-year improvement for the gasoline and diesel
fleets, respectively, from MYs 2014-2025 compared to an almost 3 percent per year
improvement for light-duty trucks in the same time frame. ICCT recommended that the
agencies' analysis incorporate the full analysis and inputs from the light-duty rulemaking and
that the result would be improvements in the range of 35 percent over the MYs 2014-2025 rather
than the proposed 23 percent improvement over this time frame.
On the other hand, some other reviewers commented that the proposed standards could be
unduly aggressive considering the products and technologies involved. GM commented that any
attempt to force more stringent regulations than proposed, such as Alternative 4, would be
extremely detrimental to manufacturers, consumers, the U.S. economy, and the millions of
transportation-related jobs. Daimler similarly commented that the proposed standards would be
a challenge for automotive manufacturers. Under certain conditions, such a standard may
necessitate hybridization of the affected vehicle fleet, which would require substantial
development and material costs. All technologies taken into account for the class 2b/3
stringencies should reflect cost effectiveness calculations, especially alternative powertrains such
as hybrids, battery, and fuel cell driven electric vehicles. Daimler recommends that the agencies
adopt the proposed standard over Alternative 4, as the additional two years of lead-time will be
critical for automotive manufacturers in developing the necessary technologies to achieve
compliance. Nissan commented that the Alternative 4 3.5 percent per stringency level is simply
not feasible, as it does not provide the necessary lead-time to enable manufacturers to balance
competitive market constraints with the cost of applying new technologies to a limited product
offering. Nissan further commented that to the extent that the more stringent alternative is
predicated on the adoption of hybrid and electric powertrain technology, Nissan does not believe
that such technology is feasible for this market segment.
The American Automotive Policy Council (AAPC, representing FCA, Ford, and General
Motors) further commented that proposals for greater stringency than Alternative 3 are not
pp CARB, Docket No. NHTSA-2014-0132-0125at pages 52-53.
QQ UCS, Docket No. EPA-HQ-OAR-2014-0827-1329, at pages 23-25.
1111 CBD, Docket No. NHTSA-2014-0132-0101, at pages 8-9.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
supportable given the required early introduction of unproven technologies with their associated
consumer acceptance risk, as well as the many implicit risks that impact stringency. AAPC
commented that the proposed standards are aggressive and will challenge industry. AAPC noted
that the baseline fleet includes a high percentage of advanced diesel technology such as SCR,
making additional improvements considerably more challenging. In the light-duty fleet, diesel
technology accounts for 3 percent of fleet whereas the heavy-duty fleet consists of over 50
percent diesel.
AAPC also noted that Phase 2 technologies are being used today. For example, FCA's
modern gasoline engine has robust combustion with multiple spark plugs, variable cam phasing,
cylinder deactivation, and cooled EGR. AAPC commented that even with this level of gasoline
engine technology, FCA is challenged by the early year Phase 1 standards and will need to look
at adding even more technology for Phase 2. AAPC also provided data showing that while
smaller displacement boosted gasoline engine technology may be applicable in some variants of
commercial vans, this technology is not suited for the pickup truck variants in this segment
because of customer demands for towing capability. AAPC commented that concurrent
stringency increases in Tier 3/LEV III criteria emission requirements will negatively impact CO2
and fuel consumption. As an alternative to the standards proposed in the NPRM, the American
Automotive Policy Council (AAPC, representing FCA, Ford, and General Motors) proposed
standards that would achieve the stringency by model year 2027, but that would do so at a more
gradual pace.ss As means of providing flexibility in complying with these standards, AAPC also
commented that the agencies should allow credits to be banked for longer than 5 years, and
should allow credits to be transferred between the light- and heavy-duty fleets.TT
10.2.9.3 Determination
Having considered these comments as well as the updated analysis summarized above,
NHTSA is adopting standards under which the stringency of fuel consumption standards for HD
pickups and vans advance at an annual rate of 2.5 percent during model years 2021-2027 relative
to the 2018 MY Phase 1 standard level. In NHTSA's judgment, this pace of stringency increase
will appropriately accommodate manufacturers' redesign workload and product schedules,
especially in light of this sector's limited product offerings1111 and long product cycles. Given the
provided flexibility to carry credits forward (and back) between model years, this approach
strikes a balance between, on one hand, meaningful early fuel efficiency improvements and, on
the other, providing manufacturers appropriate lead time.
Compared to Alternative 3, Alternative 2 would forego significant cost-efficient
opportunities to apply conventional and moderately advanced technology in order to reduce fuel
consumption and emissions. Also, although the updated analysis summarized above shows costs
for Alternative 3 (as costs incremental to the No Action Alternative) somewhat higher than
estimated in the NPRM analysis, the agencies find that under either the Method A or Method B
ss AAPC, Docket No. NHTSA-2014-0132-0103 ], at pages 12-13.
TT AAPC, Docket No. NHTSA-2014-0132-0103 at pages 13-16.
1X1 Manufacturers generally have only one pickup platform and one van platform in this segment.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
analyses, AAPC's proposed more gradual progression leading up to MY 2027 would also forego
cost-effective improvements which are readily feasible in the lead time provided.
Furthermore, the Method A analysis indicates that the standards defining Alternative 3
can likely be met with minimal reliance on hybrid technologies. Considering this, NHTSA also
find it unnecessary to extend the lifespan of banked credits or adopt other credit related
flexibilities to mitigate the stringency increases under Alternative 3.
10.3 What Industry Impacts Did EPA's "Method B" Analysis Show for
Regulatory Alternatives?
The analysis fleet provides a starting point for estimating the extent to which
manufacturers might add fuel-saving (and, therefore, C02-avoiding) technologies under various
regulatory alternatives, including the no-action alternative that defines a baseline against which
to measure estimated impacts of new standards. The analysis fleet is a forward-looking
projection of production of new HD pickups and vans, holding vehicle characteristics (e.g.,
technology content and fuel consumption levels) constant at model year 2014 levels, and
adjusting production volumes based on recent DOE and commercially-available forecasts. This
analysis fleet includes some significant changes relative to the market characterization that was
used to develop the of Phase 1 standards applicable starting in model year 2014; in particular, the
analysis fleet includes some new HD vans (e.g., Ford's Transit and Fiat/Chrysler's Promaster)
that are considerably more fuel-efficient than HD vans these manufacturers have previously
produced for the U.S. market.
While the Phase 2 standards are scheduled to begin in model year 2021, the requirements
they define are likely to influence manufacturers' planning decisions several years in advance.
This is true in light-duty planning, but accentuated by the comparatively long redesign cycles
and small number of models and platforms offered for sale in the 2b/3 market segment.
Additionally, manufacturers will respond to the cost and efficacy of available fuel consumption
improvements, the price of fuel, and the requirements of the Phase 1 standards that specify
maximum allowable average fuel consumption and GHG levels for MY2014-MY2018 HD
pickups and vans (the final standard for MY2018 is held constant for model years 2019 and
2020). The forward-looking nature of product plans that determine which vehicle models will be
offered in the model years affected by these standards lead to additional technology application
to vehicles in the analysis fleet that occurs in the years prior to the start of these standards. From
the industry perspective, this means that manufacturers will incur costs to comply with these
standards in the baseline and that the total cost of the regulations will include some costs that
occur prior to their start, and represent incremental changes over a world in which manufacturers
will have already modified their vehicle offerings compared to today.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-46 MY2021 Method B Baseline Costs for Manufacturers in 2b/3 Market Segment in the Dynamic
Baseline, or Alternative lb
MANUFACTURER
AVERAGE
TOTAL COST

TECHNOLOGY
INCREASE ($M)

COST ($)

Chrysler/Fiat
275
27
Daimler
18
0
Ford
258
78
General Motors
782
191
Nissan
282
3
Industry
442
300
As Table 10-46 shows, the industry as a whole is expected to add about $440 of new
technology to each new vehicle model by 2021 under the no-action alternative defined by the
Phase 1 standards. Reflecting differences in projected product offerings in the analysis fleet,
some manufacturers (notably Daimler) are significantly less constrained by the Phase 1 standards
than others and face lower cost increases as a result. General Motors (GM) shows the largest
increase in average vehicle cost, but results for GM's closest competitors (Ford and
Chrysler/Fiat) do not include the costs of their recent van redesigns, which are already present in
the analysis fleet (discussed in greater detail below).
The above results reflect the assumption that manufacturers having achieved compliance
with standards might act as if buyers are willing to pay for further fuel consumption
improvements that "pay back" within 6 months (i.e., those improvements whose incremental
costs are exceeded by savings on fuel within the first six months of ownership). It is also
possible that manufacturers will choose not to migrate cost-effective technologies to the 2b/3
market segment from similar vehicles in the light-duty market. Resultant technology costs in
model year 2021 results for the no-action alternative, summarized in Table 10-47 below, are
quite similar to those shown above for the 6-month payback period. Due to the similarity
between the two baseline characterizations, results in the following discussion represent
differences relative to only the 6-month payback baseline.
Table 10-47 MY2021 Method B Baseline Costs for HD Pickups and Vans in the Flat Baseline, or Alternative
la
MANUFACTURER
AVERAGE
TECHNOLOGY COST
($)
TOTAL COST
INCREASE ($M)
Chrysler/Fiat
268
27
Daimler
0
0
Ford
248
75
General Motors
767
188
Nissan
257
3
Industry
431
292
The results below represent the impacts of several regulatory alternatives, including those
defined by the Phase 2 standards, as incremental changes over the baseline, where the baseline is

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
defined as the state of the world in the absence of this regulatory action (but, of course, including
the Phase 1 standards). Large-scale, macroeconomic conditions like fuel prices are constant
across all alternatives, including the baseline, as are the fuel economy improvements under the
no-action alternative defined by the Phase 1 rule that covers model years 2014 - 2018 and is
constant from model year 2018 through 2020. In the baseline scenario, the Phase 1 standards are
assumed to remain in place and at 2018 levels throughout the analysis (i.e. MY 2030). The only
difference between the definitions of the alternatives is the stringency of these standards starting
in MY 2021 and continuing through MY 2025 or MY 2027, and all of the differences in
outcomes across alternatives are attributable to differences in the standards.
The standards vary in stringency across regulatory alternatives (1 - 5), but as discussed
above, all of the standards are based on the curve developed in the Phase 1 standards that relate
fuel economy and GHG emissions to a vehicle's work factor. The alternatives considered here
represent different rates of annual increase in the curve defined for model year 2018, growing
from a 0 percent annual increase (Alternative 1, the baseline or "no-action" alternative) up to a 4
percent annual increase (Alternative 5). Table 10-48 shows a summary of outcomes by
alternative incremental to the baseline (Alternative lb) for Model Year 2030vv, with the
exception of technology penetration rates, which are absolute.
The technologies applied as inputs to the CAFE model (in either its Method B or A
iterations) have been grouped (in most cases) to give readers a general sense of which types of
technology are applied more frequently than others, and are more likely to be offered in new
class 2b/3 vehicles once manufacturers are fully compliant with the standards in the alternative.
Model year 2030 was chosen to account for technology application that occurs once the
standards have stabilized, but manufacturers are still redesigning products to achieve compliance
- generating technology costs and benefits in those model years. The summaries of technology
penetration are also intended to reflect the relationship between technology application and cost
increases across the alternatives. The table rows present the degree to which specific
technologies are predicted to be present in new class 2b and class 3 vehicles in 2030, and
correspond to: variable valve timing (VVT) and/or variable valve lift (VVL), cylinder
deactivation, direct injection, engine turbocharging, 8-speed automatic transmissions, electric
power-steering and accessory improvements, micro-hybridization (which reduces engine idle,
but does not assist propulsion), full hybridization (integrated starter generator or strong hybrid
that assists propulsion and recaptures braking energy), and aerodynamic improvements to the
vehicle shape. In addition to the technologies in the following tables, there are some lower-
complexity technologies that have high market penetration across all the alternatives and
manufacturers; low rolling-resistance tires, low friction lubricants, and reduced engine friction
are examples.
vv As noted above, the CAFE model estimates that redesign schedules will "straddle" model year 2027, the latest
year for which the agencies are increasing the stringency of fuel consumption and GHG standards. Considering also
that today's analysis estimates some earning and application of "carried forward" compliance credits, the model was
run extending the analysis through model year 2030.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-48 Summary of HD Pickup and Van Alternatives' Impact on Industry versus the Dynamic
Baseline, Alternative lb
ANNUAL STRINGENCY INCREASE
2.0%/Y
2.5%/Y
3.5%/Y
4.0%/Y
Stringency Increase Through MY
2025
2027
2025
2025
Total Stringency Increase
9.6%
16.2%
16.3%
18.5%
Average Fuel Economy (miles per gallon)
Required
19.04
20.57
20.57
21.14
Achieved
19.14
20.61
20.83
21.27
Average Fuel Consumption (gallons /100 mi.)
Required
5.25
4.86
4.86
4.73
Achieved
5.22
4.85
4.80
4.70
Average Greenhouse Gas Emissions (g/mi)
Required
495
458
458
446
Achieved
491
458
453
444
Technology Penetration (%)
WT and/or VVL
46
46
46
46
Cylinder Deac.
29
21
21
21
Direct Injection
17
25
31
32
Turbocharging
55
63
63
63
8-Speed AT
67
96
96
97
EPS, Accessories
54
80
79
79
Stop Start
0
0
10
13
Hybridization3
0
8
35
51
Aero. Improvements
36
78
78
78
Mass Reduction (vs. No-Action)
CW (lb.)
239
243
325
313
CW (%)
3.7
3.7
5.0
4.8
Technology Cost (vs. No-Action)
Average ($)b
578
1,348
1,655
2,080
Total ($m)0
437
1,019
1,251
1,572
Payback period (m)0
25
31
34
38
Notes:
" Includes mild hybrids (ISG) and strong HEVs.
b Values used in Methods A & B
0 Values used in Method A, calculated using a 3% discount rate.
In general, as stated above, the Method B model projected that the standards will cause
manufacturers to produce HD pickups and vans that are lighter, more aerodynamic, and more
technologically complex across all the alternatives. As Table 10-48 shows, there is a difference
between the relatively small increases in required fuel economy and average incremental
technology cost between the alternatives, suggesting that the challenge of improving fuel
consumption and CO2 emissions accelerates as stringency increases (i.e., that there may be a
"knee" in the relationship between technology cost and reductions in the fuel consumption/GHG
emissions.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
The contrast between alternatives 3 and 4 is even more prominent, with an identical
required fuel economy improvement projected to lead to price increases greater than 20 percent
based on the more rapid rate of increase and shorter time span of Alternative 4, which achieves
all of its increases by MY 2025 while Alternative 3 continues to increase at a slower rate until
MY 2027. Despite these differences, the increase in average payback period when moving from
Alternative 3 to Alternative 4 to Alternative 5 is fairly constant at around an additional three
months for each jump in stringency.
Manufacturers offer few models, typically only a pickup truck and/or a cargo van, and
while there are a large number of variants of each model, the degree of component sharing across
the variants can make diversified technology application either economically impractical or
impossible. This forces manufacturers to apply some technologies more broadly in order to
achieve compliance than they might do in other market segments (passenger cars, for example).
This difference between broad and narrow application - where some technologies must be
applied to entire platforms, while some can be applied to individual model variants - also
explains why certain technology penetration rates decrease between alternatives of increasing
stringency (cylinder deactivation or mass reductions in Table 10-48, for example). For those
cases, narrowly applying a more advanced (and costly) technology can be a more cost effective
path to compliance and lead to reductions in the amount of lower-complexity technology that is
applied.
One driver of the change in technology cost between Alternative 3 and Alternative 4 in
the Method B analysis is the amount of hybridization projected to result from the implementation
of the standards. While only about 5 percent full hybridization (defined as either integrated
starter-generator or strong hybrid) is expected to be needed to comply with Alternative 3, the
higher rate of increase and compressed schedule moving from Alternative 3 to Alternative 4 is
enough to increase the percentage of the fleet adopting full hybridization by a factor of two. To
the extent that manufacturers are concerned about introducing hybrid vehicles in the 2b and 3
market, it is worth noting that new vehicles subject to Alternative 3 achieve the same fuel
economy as new vehicles subject to Alternative 4 by 2030, with less full hybridization projected
under this Method B analysis as being needed to achieve the improvement.
The alternatives also lead to important differences in outcomes at the manufacturer level,
both from the industry average and from each other. General Motors, Ford, and Fiat Chrysler,
are expected to have approximately 95 percent of the 2b/3 new vehicle market during the years
that these standards are being phased in. Due to their importance to this market and the
similarities between their model offerings, these three manufacturers are discussed together and a
summary of the way each is impacted by the standards appears below in Table 10-49, Table
10-50, and Table 10-51 for General Motors, Ford, and Chrysler/Fiat, respectively.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-49 Summary of Impacts on General Motors by 2030 in the HD Pickup and Van Market versus the
Dynamic Baseline, Alternative lb
ANNUAL
STRINGENCY
INCREASE
2.0%/Y
2.5%/Y
3.5%/Y
4.0%/Y
Stringency Increase
Through MY
2025
2027
2025
2025
Average Fuel Economy (miles per gallon)
Required
18.38
19.96
20
20.53
Achieved
18.43
19.95
20.24
20.51
Average Fuel Consumption (gallons /100 mi.)
Required
5.44
5.01
5
4.87
Achieved
5.42
5.01
4.94
4.87
Average Greenhouse Gas Emissions (g/mi)
Required
507
467
467
455
Achieved
505
468
461
455
Technology Penetration (%)
WT and/or VVL
64
64
64
64
Cylinder Deac.
47
47
47
47
Direct Injection
18
18
36
36
Turbocharging
53
53
53
53
8-Speed AT
36
100
100
100
EPS, Accessories
100
100
100
100
Stop Start
0
0
2
0
Hybridization0
0
19
79
100
Aero. Improvements
100
100
100
100
Mass Reduction (vs. No-Action)
CW (lb.)
325
161
158
164
CW (%)
5.3
2.6
2.6
2.7
Technology Cost (vs. No-Action)
Average ($)a
785
1,706
2,244
2,736
Total ($m,
undiscounted)b
214
465
611
746
Notes:
a Values used in Methods A & B
b Values used in Method A, calculated at a 3% discount rate
0 Includes mild hybrids (ISG) and strong HEVs.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-50 Summary of Impacts on Ford by 2030 in the HD Pickup and Van Market versus the Dynamic
Baseline, Alternative lb
ANNUAL
STRINGENCY
INCREASE
2.0%/Y
2.5%/Y
3.5%/Y
4.0%/Y
Stringency Increase
Through MY
2025
2027
2025
2025
Average Fuel Economy (miles per gallon)
Required
19.42
20.96
20.92
21.51
Achieved
19.5
21.04
21.28
21.8
Average Fuel Consumption (gallons /100 mi.)
Required
5.15
4.77
4.78
4.65
Achieved
5.13
4.75
4.70
4.59
Average Greenhouse Gas Emissions (g/mi)
Required
485
449
450
438
Achieved
482
447
443
433
Technology Penetration (%)
WT and/or VVL
34
34
34
34
Cylinder Deac.
18
0
0
0
Direct Injection
16
34
34
34
Turbocharging
51
69
69
69
8-Speed AT
100
100
100
100
EPS, Accessories
41
62
59
59
Stop Start
0
0
20
29
Hybridization0
0
2
14
30
Aero. Improvements
0
59
59
59
Mass Reduction (vs. No-Action)
CW (lb.)
210
202
379
356
CW (%)
3.2
3
5.7
5.3
Technology Cost (vs. No-Action)
Average ($)a
506
1,110
1,353
1,801
Total ($m,
undiscounted)b
170
372
454
604
Notes:
a Values used in Methods A & B
b Values used in Method A, calculated at a 3% discount rate
0 Includes mild hybrids (ISG) and strong HEVs.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-51 Summary of Impacts on Fiat Chrysler by 2030 in the HD Pickup and Van Market versus the
Dynamic Baseline, Alternative lb
ANNUAL
STRINGENCY
INCREASE
2.0%/Y
2.5%/Y
3.5%/Y
4.0%/Y
Stringency Increase
Through MY
2025
2027
2025
2025
Average Fuel Economy (miles per gallon)
Required
18.73
20.08
20.12
20.70
Achieved
18.83
20.06
20.10
20.70
Average Fuel Consumption (gallons /100 mi.)
Required
5.34
4.98
4.97
4.83
Achieved
5.31
4.99
4.97
4.83
Average Greenhouse Gas Emissions (g/mi)
Required
515
480
479
466
Achieved
512
481
480
467
Technology Penetration (%)
WT and/or VVL
40
40
40
40
Cylinder Deac.
23
23
23
23
Direct Injection
17
17
17
17
Turbocharging
74
74
74
74
8-Speed AT
65
88
88
88
EPS, Accessories
0
100
100
100
Stop-Start
0
0
0
0
Hybridization0
0
3
3
10
Aero. Improvements
0
100
100
100
Mass Reduction (vs. No-Action)
CW (lb.)
196
649
648
617
CW (%)
2.8
9.1
9.1
8.7
Technology Cost (vs. No-Action)
Average ($)a
434
1,469
1,486
1,700
Total ($m,
undiscounted) b
48
163
164
188
Notes:
a Values used in Methods A & B
b Values used in Method A, calculated at a 3% discount rate
0 Includes mild hybrids (ISG) and strong HEVs.
The fuel consumption and GHG standards require manufacturers to achieve an average
level of compliance, represented by a sales-weighted average across the specific targets of all
vehicles offered for sale in a given model year, such that each manufacturer will have a unique

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
required consumption/emissions level determined by the composition of its fleet, as illustrated
above. However, there are more interesting differences than the small differences in required
fuel economy levels among manufacturers. In particular, the average incremental technology
cost increases with the stringency of the alternative for each manufacturer, but the size of the
cost increase from one alternative to the next varies among them, with General Motors showing
considerably larger increases in cost moving from Alternative 3 to Alternative 4 than from either
Alternative 2 to Alternative 3 or Alternative 4 to Alternative 5. Ford is estimated to have more
uniform cost increases from each alternative to the next, in increasing stringency, though still
benefits from the reduced pace and longer period of increase associated with Alternative 3
compared to Alternative 4..
The Method B simulation results show all three manufacturers facing cost increases when
the stringency of the standards move from 2.5 percent annual increases over the period from MY
2021 - 2027 to 3.5 percent annual increases from MY 2021 - 2025, but General Motors has the
largest at 75 percent more than the industry average price increase for Alternative 4. GM also
faces higher cost increases in Alternative 2, about 50 percent more than either Ford or
Fiat/Chrysler. And for the most stringent alternative considered, EPA estimates that General
Motors will face average cost increases of more than $2,700, in addition to the more than $700
increase in the baseline - approaching nearly $3,500 per vehicle over today's prices.
Technology choices also differ by manufacturer, and some of those decisions are directly
responsible for the largest cost discrepancies. For example, in this Method B analysis, GM is
estimated to engage in the least amount of mass reduction among the Big 3 after Phase 1, and
much less than Chrysler/Fiat, but reduces average vehicle mass by over 300 pounds in the
baseline - suggesting that some of GM's easiest Phase 1 compliance opportunities can be found
in lightweighting technologies. Similarly, Fiat Chrysler is projected to apply less hybridization
than the others, and much less than General Motors, which is simulated in Alternative 4 to have
full hybrids (either integrated starter generator or complete hybrid system) on all of its fleet by
2030, nearly 20 percent of which will be strong hybrids. . Because the analysis applies the same
technology inputs and the same logic for selecting among available opportunities to apply
technology, the unique situation of each manufacturer determined which technology path is
projected as the most cost-effective.
In order to understand the differences in incremental technology costs and fuel economy
achievement across manufacturers in this market segment, it is important to understand the
differences in their starting position relative to these standards. One important factor, made more
obvious in the following figures, is the difference between the fuel economy and performance of
the recently redesigned vans offered by Fiat Chrysler and Ford (the Promaster and Transit,
respectively), and the more traditionally-styled vans that continue to be offered by General
Motors (the Express/Savannah). In MY 2014, Ford began the phase-out of the Econoline van
platform, moving those volumes to the Euro-style Transit vans. The Transit platform represents
a significant improvement over the existing Econoline platform from the perspective of fuel
economy, and for the purpose of complying with the standards, the relationship between the
Transit's work factor and fuel economy is a more favorable one than the Econoline vans it
replaces. Since the redesign of van offerings from both Fiat Chrysler and Ford occur in (or prior
to) the 2014 model year, the costs, fuel consumption improvements, and reductions of vehicle
mass associated with those redesigns are included in the analysis fleet, meaning they are not

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
carried forward as part of the compliance modeling exercise. By contrast, General Motors is
simulated to redesign their van offerings after 2014, such that there is a greater potential for these
vehicles to incur additional costs attributable to new standards, unlike the costs associated with
the recent redesigns of their competitors. The inclusion of these new Ford and Fiat Chrysler
products in the analysis fleet is the primary driver of the cost discrepancy between GM and its
competitors in both the baseline and Alternative 2 in this Method B analysis, when Ford and Fiat
Chrysler have to apply considerably less technology to achieve compliance.
The remaining 5 percent of the 2b/3 market is attributed to two manufacturers, Daimler
and Nissan, which, unlike the other manufacturers in this market segment, only produce vans.
The vans offered by both manufacturers currently utilize two engines and two transmissions,
although both Nissan engines are gasoline engines and both Daimler engines are diesels. Despite
the logical grouping, these two manufacturers are projected to be impacted much differently by
these standards. For the least stringent alternative considered, Daimler is projected to add no
technology and incurs no incremental cost in order to comply with the standards. At stringency
increases greater than or equal to 3.5 percent per year, Daimler only really improves some of
their transmissions and improves the electrical accessories of its Sprinter vans. By contrast,
Nissan's starting position is much weaker and their compliance costs closer to the industry
average in Table 10-48. This difference could increase if the analysis fleet supporting the final
rule includes forthcoming Nissan HD pickups.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-52 Summary of Impacts on Daimler by 2030 in the HD Pickup and Van Market versus the
Dynamic Baseline, Alternative lb
ANNUAL
STRINGENCY
INCREASE
2.0%/Y
2.5%/Y
3.5%/Y
4.0%/Y
Stringency Increase
Through MY
2025
2027
2025
2025
Average Fuel Economy (miles per gallon)
Required
23.36
25.19
25.25
25.91
Achieved
25.23
25.79
25.79
26.53
Average Fuel Consumption (gallons /100 mi.)
Required
4.28
3.97
3.96
3.86
Achieved
3.96
3.88
3.88
3.77
Average Greenhouse Gas Emissions (g/mi)
Required
436
404
404
393
Achieved
404
395
395
384
Technology Penetration (%)
WT and/or VVL
0
0
0
0
Cylinder Deac.
0
0
0
0
Direct Injection
0
0
0
0
Turbocharging
44
44
44
44
8-Speed AT
0
44
44
100
EPS, Accessories
0
0
0
0
Stop-Start
0
0
0
0
Hybridization0
0
0
0
0
Aero. Improvements
0
0
0
0
Mass Reduction (vs. No-Action)
CW (lb.)
0
0
0
0
CW (%)
0
0
0
0
Technology Cost (vs. No-Action)
Average ($)a
0
165
165
374
Total ($m,
undiscounted) b
0
4
4
9
Notes:
a Values used in Methods A & B
b Values used in Method A, calculated at a 3% discount rate
0 Includes mild hybrids (ISG) and strong HEVs.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
Table 10-53 Summary of Impacts on Nissan by 2030 in the HD Pickup and Van Market versus the Dynamic
Baseline, Alternative lb
ANNUAL
STRINGENCY
INCREASE
2.0%/Y
2.5%/Y
3.5%/Y
4.0%/Y
Stringency Increase
Through MY
2025
2027
2025
2025
Average Fuel Economy (miles per gallon)
Required
19.64
21.19
20.92
21.46
Achieved
19.84
21.17
21.19
21.51
Average Fuel Consumption (gallons /100 mi.)
Required
5.09
44.72
4.78
4.66
Achieved
5.04
4.72
4.72
4.65
Average Greenhouse Gas Emissions (g/mi)
Required
452
419
425
414
Achieved
448
419
419
413
Technology Penetration (%)
WT and/or VVL
100
100
100
100
Cylinder Deac.
49
49
49
49
Direct Injection
51
51
51
100
Turbocharging
51
51
51
50
8-Speed AT
0
51
51
51
EPS, Accessories
0
100
100
100
Stop-Start
0
0
0
0
Hybridization0
0
0
0
28
Aero. Improvements
0
100
100
100
Mass Reduction (vs. No-Action)
CW (lb.)
0
0
307
303
CW (%)
0
0
5
4.9
Technology Cost (vs. No-Action)
Average ($)11
378
1,150
1,347
1,935
Total ($m,
undiscounted)b
5
15.1
17.7
25.4
Notes:
a Values used in Methods A & B
b Values used in Method A, calculated at a 3% discount rate
0 Includes mild hybrids (ISG) and strong HEVs.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
As Table 10-52 and Table 10-53 show, Nissan is projected to apply more technology than
Daimler in the less stringent alternatives and significantly more technology with increasing
stringency. The Euro-style Sprinter vans that comprise all of Daimler's model offerings in this
segment put Daimler in a favorable position. However, those vans are already advanced -
containing downsized diesel engines and advanced aerodynamic profiles. Much like the Ford
Transit vans, the recent improvements to the Sprinter vans occurred outside the scope of the
compliance modeling so the costs of the improvements are not captured in the analysis.
Although Daimler's required fuel economy level is much higher than Nissan's (in miles
per gallon), Nissan starts from a much weaker position than Daimler and must incorporate
additional engine, transmission, platform-level technologies (e.g. mass reduction and
aerodynamic improvements) in order to achieve compliance. In fact, more than 25 percent of
Nissan's van offerings are projected to contain integrated starter generators by 2030 in
Alternative 5.

-------
*** E.O. 12866 Review — Revised — Do Not Cite, Quote, or Release During Review ***
References
1	RTI International, "Automobile Industry Retail Price Equivalent and Indirect Cost Multipliers," February 2009;
EPA-420-R-09-003; http://www3.epa.gov/otaq/ld-hwy/420r09003.pdf.
2	Rogozhin, A., et al., "Using indirect cost multipliers to estimate the total cost of adding new technology in the
automobile industry," International Journal of Production Economics (2009), doi: 10.1016/i.iipe.2009.11.031.
3	80 FR 40137.

-------
Chapter 11: Results of the Preferred and Alternative
Standards
The heavy-duty truck segment is very complex. The sector consists of a diverse group of
impacted parties, including engine manufacturers, chassis manufacturers, truck manufacturers,
trailer manufacturers, truck fleet owners and the public. The standards are largely shaped to
optimize the environmental and fuel savings benefits of the program, while balancing the
relevant statutory factors and respecting the unique and varied nature of the sector. In
developing this rulemaking, we considered a number of alternatives that could result in fewer or
potentially greater GHG and fuel consumption reductions than the preferred alternative. This
section summarizes the alternatives we considered and presents assessments of CO2 reductions
and fuel savings associated with each alternative. See the Preamble for a discussion of how the
agencies balanced the relevant statutory factors to select the preferred alternative (in this action,
the "final program").
For this final rulemaking, the agencies used two analyses by employing both DOT's
CAFE model and EPA's MOVES model. These models were used to project fuel consumption
and GHG emissions impacts resulting from the standards. The agencies used EPA's MOVES
model to estimate fuel consumption and emissions impacts for tractor-trailers (including the
engines which power the vehicle), and vocational vehicles (including the engine which powers
the vehicle). For heavy-duty pickups and vans, the agencies performed separate analyses using
the CAFE model ("Method A") and the MOVES model ("Method B") to estimate fuel
consumption and emissions from these vehicles. For these methods, the agencies analyzed the
impact of the final rules, relative to two different reference cases - "flat" (Alternative la) and
"dynamic" (Alternative lb). The flat baseline projects very little improvement in new vehicles
in the absence of new Phase 2 standards. In contrast, the dynamic baseline projects more
improvements in vehicle fuel efficiency. See Chapter 5 for the discussion of the EPA's MOVES
model (which was used for both methods) and Chapter 10 for the discussion of the DOT's CAFE
model (which was used for Method A).
11.1 What Are the Alternatives that the Agencies Considered?
The five alternatives below represent a broad range of potential stringency levels, and
thus a broad range of associated technologies, costs and benefits for a HD vehicle fuel efficiency
and GHG emissions program.
In developing alternatives, NHTSA must consider EISA's requirement for the MD/HD
fuel efficiency program noted above. 49 U.S.C. 32902(k)(2) and (3) contain the following three
requirements specific to the MD/HD vehicle fuel efficiency improvement program: (1) The
program must be "designed to achieve the maximum feasible improvement;" (2) the various
required aspects of the program must be appropriate, cost-effective, and technologically feasible
for MD/HD vehicles; and (3) the standards adopted under the program must provide not less than
four model years of lead time and three model years of regulatory stability. In considering these
various requirements, NHTSA will also account for relevant environmental and safety
considerations.

-------
Each of the alternatives presented by NHTSA and EPA represents, in part, a different
way the agencies could establish a HD program pursuant to EISA and the CAA. The agencies
are finalizing the Alternative 3 standards. The alternatives below represent a broad range of
approaches under consideration for finalizing the HD vehicle fuel efficiency and GHG emissions
standards.
Chapters 11.1.1 through 11.2 summarize the alternatives that were analyzed and how
they were modeled.
11.1.1 Alternative 1: No Action (the Baseline for Phase 2)
OMB guidance regarding regulatory analysis indicates that proper evaluation of the
benefits and costs of regulations and their alternatives requires agencies to identify a baseline:
"You need to measure the benefits and costs of a rule against a baseline. This
baseline should be the best assessment of the way the world would look absent the
proposed action. The choice of an appropriate baseline may require
consideration of a wide range ofpotential factors, including:
•	evolution of the market
•	changes in external factors affecting expected benefits and costs
•	changes in regulations promulgated by the agency or other government entities
•	degree of compliance by regulated entities with other regulations
It may be reasonable to forecast that the world absent the regulation will
resemble the present. If this is the case, however, your baseline should reflect the
future effect of current government programs and policies. For review of an
existing regulation, a baseline assuming no change in the regulatory program
generally provides an appropriate basis for evaluating regulatory alternatives.
When more than one baseline is reasonable and the choice of baseline will
significantly affect estimated benefits and costs, you should consider measuring
benefits and costs against alternative baselines. In doing so you can analyze the
effects on benefits and costs of making different assumptions about other
agencies' regulations, or the degree of compliance with your own existing rules.
In all cases, you must evaluate benefits and costs against the same baseline. You
should also discuss the reasonableness of the baselines used in the sensitivity
analyses. For each baseline you use, you should identify the key uncertainties in
your forecast. w
A no-action alternative is also required as a baseline against which to measure
environmental impacts of the standards and alternatives. NHTSA, as required by the National
Environmental Policy Act, is documenting these estimated impacts in the EIS published with this
FRM.2

-------
The No Action Alternative for today's analysis, alternatively referred to as the "baseline"
or "reference case," assumes that the agencies do not issue new rules regarding MD/HD fuel
efficiency and GHG emissions. That is, this alternative assumes that the Phase 1 MD/HD fuel
efficiency and GHG emissions program's model year 2018 standards would be extended
indefinitely and without change.
The agencies recognize that there are a number of factors that create uncertainty in
projecting a baseline against which to compare the future effects of the alternatives. The
composition of the future fleet—such as the relative position of individual manufacturers and the
mix of products they each offer—cannot be predicted with certainty at this time. As reflected, in
part, by the market forecast underlying the agencies' analysis, we anticipate that the baseline
market for medium- and heavy-duty vehicles will continue to evolve within a competitive market
that responds to a range of factors. Additionally, the heavy-duty vehicle market is diverse, as is
the range of vehicle purchasers.
Heavy-duty vehicle manufacturers have reported that their customers' purchasing
decisions are influenced by their customers' own determinations of minimum total cost of
ownership, which can be unique to a particular customer's circumstances. For example, some
customers (e.g., less-than-truckload or package delivery operators) operate their vehicles within a
limited geographic region and typically own their own vehicle maintenance and repair centers
within that region. These operators tend to own their vehicles for long time periods, and
sometimes for the entire service life of the vehicle. Their total cost of ownership is influenced by
their ability to better control their own maintenance costs, and thus they can afford to consider
fuel efficiency technologies that have longer payback periods, outside of the vehicle
manufacturer's warranty period. Other customers (e.g. truckload or long-haul operators) tend to
operate cross-country, and thus must depend upon truck dealer service centers for repair and
maintenance. Some of these customers tend to own their vehicles for about four to seven years,
so that they typically do not have to pay for repair and maintenance costs outside of either the
manufacturer's warranty period or some other extended warranty period. Many of these
customers tend to require seeing evidence of fuel efficiency technology payback periods on the
order of 18 to 24 months before seriously considering evaluating a new technology for potential
adoption within their fleet (NAS 2010, Roeth et al. 2013, Klemick et al. 2014). Purchasing
decisions, however, are not based exclusively on payback period, but also include the
considerations discussed in this section. For the baseline analysis, the agencies use payback
period as a proxy for all of these considerations, and therefore the payback period used for the
baseline analysis may be shorter than the payback period industry uses as a threshold for the
further consideration of a technology.
Some owners accrue relatively few vehicle miles traveled per year, such that they may be
less likely to adopt new fuel efficiency technologies, while other owners who use their vehicle(s)
with greater intensity may be even more willing to pay for fuel efficiency improvements.
Regardless of the type of customer, their determination of minimum total cost of ownership
involves the customer balancing their own unique circumstances with a heavy-duty vehicle's
initial purchase price, availability of credit and lease options, expectations of vehicle reliability,
resale value and fuel efficiency technology payback periods. The degree of the incentive to
adopt additional fuel efficiency technologies also depends on customer expectations of future
fuel prices, which directly impacts customer expectations of the payback period.

-------
Another factor the agencies considered is that other federal and state-level policies and
programs are specifically aimed at stimulating fuel efficiency technology development and
deployment. Particularly relevant to this sector are DOE's 21st Century Truck Partnership,
EPA's voluntary SmartWay Transport program, and California's AB32 fleet requirements.3'4'5
The future availability of more cost-effective technologies to reduce fuel consumption could
provide manufacturers an incentive to produce more fuel-efficient medium- and heavy-duty
vehicles, which in turn could provide customers an incentive to purchase these vehicles. The
availability of more cost-effective technologies to reduce fuel consumption could also lead to a
substitution of less cost-effective technologies, where overall fuel efficiency could remain fairly
flat if buyers are less interested in fuel consumption improvements than in reduced vehicle
purchase prices and/or improved vehicle performance and/or utility.
We have also applied the EIA's AEO estimates of future fuel prices; however, heavy-
duty vehicle customers could have different expectations about future fuel prices, and could
therefore be more inclined or less inclined to apply new technology to reduce fuel consumption
than might be expected based on EIA's forecast. We expect that vehicle customers will be
uncertain about future fuel prices, and that this uncertainty will be reflected in the degree of
enthusiasm to apply new technology to reduce fuel consumption.
Considering all of these factors, the agencies have approached the definition of the No
Action Alternative separately for each vehicle and engine category covered by today's rule.
Except as noted below, these baselines are largely the same as the proposed Alternatives la and
lb.
For trailers, the agencies considered two No Action alternatives to cover a nominal range
of uncertainty. The trailer category is unique in the context of this rulemaking because it is the
only heavy-duty category not regulated under Phase 1. The agencies project that in 2018, about
half of new 53' dry van and reefer trailers will have technologies qualifying for the SmartWay
label for aerodynamic improvements and about 90 percent would have the lower rolling
resistance tires. About half also have automatic tire inflation systems to maintain optimal tire
pressure. For Alternative la as presented in this action (referred to as the "flat" baseline), this
technology adoption remains constant after 2018. In the second case, Alternative lb, the
agencies projected that the combination of EPA's voluntary SmartWay program, DOE's 21st
Century Truck Partnership, California's AB32 trailer requirements for fleets, and the potential
for significantly reduced operating costs should result in continuing improvement to new trailers.
The agencies projected that the fraction of the in-use fleet qualifying for SmartWay will continue
to increase beyond 2027 as older trailers are replaced by newer trailers. We projected that these
improvements will continue until 2040 when 75 percent of new trailers will be assumed to
include skirts.
For vocational vehicles, the agencies considered one No Action alternative. For the
vocational vehicle category the agencies recognized that these vehicles tend to operate over
fewer vehicle miles travelled per year. Therefore, the projected payback periods for fuel
efficiency technologies available for vocational vehicles are generally longer than the payback
periods the agencies consider likely to lead to their adoption based solely on market forces. This
is especially true for vehicles used in applications in which the vehicle operation is secondary to
the primary business of the company using the vehicle. For example, since the fuel consumption

-------
of vehicles used by utility companies to repair power lines would generally be a smaller cost
relative to the other costs of repairing lines, fuel saving technologies would generally not be as
strongly demanded for such vehicles. Thus, the agencies project that fuel-saving technologies
will either not be applied or will only be applied as a substitute for more expensive fuel
efficiency technologies, except as necessitated by the Phase 1 fuel consumption and GHG
standards.
For tractors, the agencies considered two No Action alternatives to cover a nominal range
of uncertainty. For Alternative la the agencies project that fuel-saving technologies will either
not be applied or will only be applied as a substitute for more expensive fuel efficiency
technologies to tractors (thereby enabling manufacturers to offer tractors that are less expensive
to purchase), except as necessitated by the Phase 1 fuel consumption and GHG standards. In
Alternative lb the agencies estimated that some available technologies will save enough fuel to
pay back fairly quickly - within the first six months of ownership. The agencies considered a
range of information to formulate these two baselines for tractors.
Both public6 and confidential historical information shows that tractor trailer fuel
efficiency improved steadily through improvements in engine efficiency and vehicle
aerodynamics over the past 40 years, except for engine efficiency which decreased or was flat
between 2000 and approximately 2007 as a consequence of incorporating technologies to meet
engine emission regulations. Today vehicle manufacturers, the Federal Government, academia
and others continue to invest in research to develop fuel efficiency improving technologies for
the future.
In public meetings and in meetings with the agencies, the trucking industry stated that
fuel cost for tractors is the number one or number two expense for many operators, and therefore
is a very important factor for their business. However, the pre-Phase 1 market suggests that,
tractor manufacturers and operators could be slow to adopt some new technologies, even where
the agencies have estimated that the technology would have paid for itself within a few months
of operation. Tractor operators have told the agencies they generally require technologies to be
demonstrated in their fleet before widespread adoption so they can assess the actual fuel savings
for their fleet and any increase in cost associated with effects on vehicle operation, maintenance,
reliability, mechanic training, maintenance and repair equipment, stocking unique parts and
driver acceptance, as well as effects on vehicle resale value. Tractor operators often state that
they would consider conducting an assessment of technologies when provided with data that
show the technologies may payback costs through fuel savings within 18 to 24 months, based on
their assumptions about future fuel costs. In these cases, an operator may first conduct a detailed
paper study of anticipated costs and benefits. If that study shows likely payback in 18 to 24
months for their business, the fleet may acquire one or several tractors with the technology to
directly measure fuel savings, costs and driver acceptance for their fleet. Small fleets may not
have resources to conduct assessments to this degree and may rely on information from larger
fleets or observations of widespread acceptance of the technology within the industry before
adopting a technology. This uncertainty over the actual fuel savings and costs and the lengthy
process to assess technologies significantly slows the pace at which fuel efficiency technologies
are adopted.

-------
The agencies believe that using the two baselines addresses the uncertainties we have
identified for tractors. The six-month payback period of Alternative lb reflects the agencies'
consideration of factors, discussed above, that could limit—yet not eliminate—manufacturers'
tendencies to voluntarily improve fuel consumption. In contrast, Alternative la reflects a
baseline for vehicles other than trailers wherein manufacturers either do not apply fuel efficiency
technologies or only apply them as a substitute for more expensive fuel efficiency technologies,
except as necessitated by the Phase 1 fuel consumption and GHG standards.
For HD pickups and vans, the agencies considered two No Action alternatives to cover a
nominal range of uncertainty: a flat baseline (designated Alternative la) where no improvements
are modeled beyond those needed to meet Phase 1 standards and a dynamic baseline (designated
Alternative lb) where certain cost-effective technologies (i.e., those that payback within a 6
month period) are assumed to be applied by manufacturers to improve fuel efficiency beyond the
Phase 1 requirements in the absence of new Phase 2 standards. In Alternative lb the agencies
considered additional technology application, which involved the explicit estimation of the
potential to add specific fuel-saving technologies to each specific vehicle model included in the
agencies' HD pickup and van fleet analysis, as discussed in Section VI of the Preamble.
Estimated technology application and corresponding impacts depend on the modeled inputs.
Also, under this approach a manufacturer that has improved fuel consumption and GHG
emissions enough to achieve compliance with the standards is assumed to apply further
improvements, provided those improvements reduce fuel outlays by enough (within a specified
amount of time, the payback period) to offset the additional costs to purchase the new vehicle.
These calculations explicitly account for and respond to fuel prices, vehicle survival and mileage
accumulation, and the cost and efficacy of available fuel-saving technologies. Therefore, all else
being equal, more technology is applied when fuel prices are higher and/or technology is more
cost-effective. However, considering factors discussed above that could limit manufacturers'
tendency to voluntarily improve HD pickup and van fuel consumption, Alternative lb applies a
6-month payback period. In contrast, for Alternative la the agencies project that fuel-saving
technologies would either not be applied or only be applied as a substitute for more expensive
fuel efficiency technologies, except as necessitated by the Phase 1 fuel consumption and GHG
standards. In terms of impacts under reference case fuel prices, the payback period input plays a
more significant role under the No-Action Alternatives (defined by a continuation of model year
2018 standards) than under the more stringent regulatory alternatives for HD pickups and vans
described next.
11.1.1.1 Alternative la
For an explanation of analytical Methods A and B identified in some of the following
tables, please see Section I.D of the Preamble; for an explanation of the flat baseline, la, and
dynamic baseline, lb, please see Section X.A.I of the Preamble. The estimated reductions in
energy ratesA used in MOVES for Alternative la are presented in Table 11-1.
The projected use of diesel-powered auxiliary power units (APUs) during extended idling
for Alternative la is presented in Table 11-2. The reductions in aerodynamic and tire rolling
A Note that the "reductions in energy rates" for tractors and vocational vehicles reflect changes in CO2 emissions not
represented by tire rolling resistance, aerodynamic drag, or vehicle weight.

-------
resistance coefficients, and the absolute changes in average vehicle weight modeled in MOVES
for Alternative la are presented in Table 11-3. Chapter 10 discusses the agencies' use of the
CAFE model in greater detail.
Table 11-1 Estimated Reductions in Energy Rates for Alternative la using Analysis Method B a
VEHICLE TYPE
FUEL
MODEL
YEARS
FUEL/CO2
REDUCTION
Long-haul
Tractor-Trailer
and HHD
Vocational
Diesel
2018+
0.4%
Short-haul
Tractor-Trailer
and HHD
Vocational
Diesel
2018+
0%
Single-Frame
Vocational13
Diesel and CNG
2021-2023
0%
2024+
0%
Gasoline
2021-2023
0%
2024+
0%
HD Pickup Trucks
and Vans
Diesel and
Gasoline
2021
0%
2022
0%
2023
0%
2024
0%
2025+
0%
Notes:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the
flat baseline, la, and dynamic baseline, lb, please see Section X.A.I.
b Vocational vehicles modeled in MOVES include heavy heavy-duty, medium heavy-duty, and light
heavy-duty vehicles. However, for light heavy-duty vocational vehicles, class 2b and 3 vehicles are not
included in the inventories for the vocational sector. Instead, all vehicles with GVWR of less than
14,000 lbs were modeled using the energy rate reductions described below for HD pickup trucks and
vans. In practice, many manufacturers of these vehicles choose to average the lightest vocational
vehicles into chassis-certified families (i.e., heavy-duty pickups and vans).
Table 11-2 Assumed Diesel APU Use during Extended Idling for Combination Long-haul Tractor-Trailers
for Alternative la
VEHICLE TYPE
MODEL YEARS
DIESEL APU
PENETRATION
Combination Long-Haul
Tractors
2010+
9%

-------
Table 11-3 Estimated Reductions in Road Load Factors for Alternative la
VEHICLE TYPE
MODEL
REDUCTION IN
REDUCTION IN
WEIGHT

YEARS
TIRE ROLLING
AERODYNAMIC
REDUCTION


RESISTANCE
DRAG
(LB)a


COEFFICIENT
COEFFICIENT

Combination Long-haul
Tractor-Trailers
2018-2020
5.7%
2.8%
-69
2021-2023
5.7%
2.8%
-69
2024-2026
5.7%
2.8%
-69
2027+
5.7%
2.8%
-69
Combination Short-
haul Tractor-Trailersb
2018-2020
0.9%
0%
0
2021-2023
0.9%
0%
0
2024-2026
0.9%
0%
0
2027+
0.9%
0%
0
Intercity Buses
2021-2023
0%
0%
0
2024+
0%
0%
0
Transit and School
Buses
2021-2023
0%
0%
0
2024+
0%
0%
0
Refuse Trucks
2021-2023
0%
0%
0
2024+
0%
0%
0
Single Unit Short-haul
Trucks
2021-2023
0%
0%
0
2024+
0%
0%
0
Single Unit Long-haul
Trucks
2021-2023
0%
0%
0
2024+
0%
0%
0
Motor Homes
2021-2023
0%
0%
0
2024+
0%
0%
0
Notes:
a Negative weight reductions reflect an expected weight increase as a byproduct of the other vehicle and engine
improvements.
b Vocational tractors are included in the short-haul tractor segment.
11.1.1.2 Alternative lb
The estimated reductions in energy rates used in MOVES and the projected use of
auxiliary power units (APUs) during extended idling for Alternative lb are presented in Table
11-4 and Table 11-5, respectively. The reductions in aerodynamic and tire rolling resistance
coefficients, and the absolute changes in average vehicle weight are presented in Table 11-6.

-------
Table 11-4 Estimated Reductions in Energy Rates for Alternative lb using Analysis Method B a
VEHICLE TYPE
FUEL
MODEL
YEARS
FUEL/CO2
REDUCTION
Long- and Short-
Haul Tractor-
Trailer and HHD
Vocational
Diesel
2018+
0%
Single-Frame
Vocational13
Diesel and CNG
2021+
0%

Gasoline
2021+
0%
HD pickup trucks
and vans
Diesel and
Gasoline
2017
0.60%
2018
0.79%


2019
1.11%


2020
1.13%


2021
1.74%


2022
2.41%


2023
2.44%


2024
2.47%


2025
2.48%


2026
2.43%


2027
2.51%


2028
2.53%


2029
2.55%
Notes:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the
flat baseline, la, and dynamic baseline, lb, please see Section X. A. 1.
b Vocational vehicles modeled in MOVES include heavy heavy-duty, medium heavy-duty, and light
heavy-duty vehicles. However, for light heavy-duty vocational vehicles, class 2b and 3 vehicles are not
included in the inventories for the vocational sector. Instead, all vehicles with GVWR of less than
14,000 lbs were modeled using the energy rate reductions described below for HD pickup trucks and
vans. In practice, many manufacturers of these vehicles choose to average the lightest vocational
vehicles into chassis-certified families (i.e., heavy-duty pickups and vans).
Table 11-5 Assumed Diesel APU Use during Extended Idling for Combination Long-haul Tractor-Trailers
for Alternative lb
VEHICLE TYPE
MODEL
YEARS
DIESEL APU
PENETRATION
Combination Long-Haul
Tractors
2010+
9%

-------
Table 11-6 Estimated Reductions in Road Load Factors for Alternative lb
TRUCK TYPE
MODEL
REDUCTION IN TIRE
REDUCTION IN
WEIGHT

YEARS
ROLLING RESISTANCE
AERODYNAMIC
REDUCTION


COEFFICIENT
DRAG COEFFICIENT
(LB)a
Combination
2018
5.7%
2.8%
-69
Long-haul
2019
5.9%
3.5%
-71
Tractor-Trailers
2020
6.2%
4.1%
-74

2021
6.4%
4.8%
-76

2022
6.7%
5.4%
-78

2023
6.9%
6.1%
-81

2024
7.1%
6.7%
-83

2025
7.4%
7.4%
-88

2026
7.6%
8.1%
-92

2027
7.9%
8.7%
-97

2028
8.1%
10.1%
-105
Combination
2018
0.9%
0.0%
0
Short-haul
2019
1.2%
0.5%
0
Tractor-
2020
1.5%
1.0%
0
Trailers'3
2021
1.8%
1.4%
0

2022
2.1%
1.9%
0

2023
2.4%
2.4%
0

2024
2.7%
2.9%
0

2025
3.0%
3.4%
0

2026
3.2%
3.8%
0

2027
3.5%
4.3%
0

2028
3.8%
4.8%
0
Intercity Buses
2021-2023
0%
0%
0

2024+
0%
0%
0
Transit and
2021-2023
0%
0%
0
School Buses
2024+
0%
0%
0
Refuse Trucks
2021-2023
0%
0%
0

2024+
0%
0%
0
Single Unit
2021-2023
0%
0%
0
Short-haul
2024+
0%
0%
0
Trucks




Single Unit
2021-2023
0%
0%
0
Long-haul
2024+
0%
0%
0
Trucks




Motor Homes
2021-2023
0%
0%
0

2024+
0%
0%
0
Notes:
a Negative weight reductions reflect an expected weight increase as a byproduct of the other vehicle and engine
improvements.
b Vocational tractors are included in the short-haul tractor segment.

-------
11.1.2 Alternative 2: Less Stringent than the Preferred Alternative
For vocational vehicles and combination tractor-trailers, Alternative 2 represents a
stringency level which is approximately half as stringent overall as the final standards. The
agencies developed Alternative 2 to reflect a continuation of the Phase 1 approach of- applying
off-the-shelf technologies rather than requiring the development of new technologies or
fundamental improvements to existing technologies. For tractors and vocational vehicles, this
also involved less integrated optimization of the vehicles and engines. Alternative 2 would not
set standards for MY 2027.
The agencies' decisions regarding which technologies could be applied to comply with
Alternative 2 considered not only assuming the use of off-the shelf technologies, but also
considered other factors, as well, such as how broadly certain technologies fit in-use applications
and regulatory structure. The resulting Alternative 2 could be met with fewer technologies and
lower penetration rates than those the agencies project will be used to meet the final Phase 2
standards. Alternative 2 is estimated to be achievable without the application of some
technologies, at any level. These and other differences are described below by category.
Overall, Alternative 2 for the final rules is conceptually similar to Alternative 2 in the NPRM.
However, some changes have been made to reflect new information provided in public
comments.
The agencies project that Alternative 2 combination tractor standards could be met by
applying lower adoption rates of the projected technologies for Alternative 3. This includes a
projection of slightly lower per-technology effectiveness for Alternative 2 versus 3. Alternative
2 also assumes that there would be little optimization of combination tractor powertrains.
The Alternative 2 for vocational vehicles assessed for these final rules does differ
somewhat from the proposal because it reflects new duty cycles that weight idle emissions more
heavily. The agencies project that the Alternative 2 vocational vehicle standard could be met
without any use of strong hybrids or any other type of transmission technology. Rather, it could
be met with off-the-shelf idle reduction technologies, low rolling resistance tires, and axle
efficiency improvements.
The Alternative 2 trailer standards would apply to only 53-foot dry and refrigerated box
trailers and could be met through the use of less effective aerodynamic technologies and higher
rolling resistance tires versus what the agencies projected could be used to meet Alternative 3
(i.e. the final standards).
The HD pickup truck and van alternatives are characterized by an annual required
percentage change (decrease) in the functions defining attribute-based targets for per-mile fuel
consumption and GHG emissions. Under the HD pickup and van standards in Alternative 2 and
each other alternative, a manufacturer's fleet would, setting aside any changes in production mix,
be required to achieve average fuel consumption/GHG levels that increase in stringency every
year relative to the standard defined for MY2018 (and held constant through 2020) that
establishes fuel consumption/GHG targets for individual vehicles. A manufacturer's specific
fuel consumption/GHG requirement is the sales-weighted average of the targets defined by the
work-factor curve in each year. Therefore, although the alternatives involve steady increases in

-------
the functions defining the targets, stringency increases faced by any individual manufacturer may
not be steady if changes in the manufacturer's product mix cause fluctuations in the average fuel
consumption and GHG levels required of the manufacturer. See Section VI of the Preamble for
additional discussion of this topic. Alternative 2 represents a 2.0 percent annual improvement in
the target curve through 2025 in fuel consumption/GHG emissions relative to the work-factor
curve in 2020. This would be 0.5 percent less stringent per year compared to the standards of
Alternative 3 and would not increase in stringency for MYs 2026 or 2027. For HD pickups and
vans the agencies project that most manufacturers could comply with the standards defining
Alternative 2 by applying technologies similar to those that could be applied in order to comply
with the final standards, but at lower application rates than could be necessitated by the
standards.
The analytical inputs for Alternative 2 are shown in the following tables. The estimated
reductions in energy rates used in MOVES and the projected use of auxiliary power units
(APUs) during extended idling are presented in Table 11-7 and Table 11-8, respectively. The
reductions in aerodynamic and tire rolling resistance coefficients, and the absolute changes in
average vehicle weight are presented in Table 11-9.

-------
Table 11-7 Estimated Reductions in CO2 Emission Rates for Alternative 2 using Analysis Method Ba
VEHICLE TYPE
FUEL
MODEL
YEARS
FUEL/CO2
REDUCTION
Long-haul
Tractor-Trailer
and HHD
Vocational
Diesel
2018-2020
0.4%
2021-2023
3.5%
2024+
6.8%
Short-haul
Tractor-Trailer
and HHD
Vocational
Diesel
2018-2020
0%
2021-2023
3.0%
2024+
6.5%
Single-Frame
Vocational13
Diesel and CNG
2021-2023
4.0%
2024+
7.6%
Gasoline
2021-2023
2.9%
2024+
4.7%
HD pickup trucks
and vans
Diesel and
Gasoline
2021
2.0%
2022
3.96%
2023
5.88%
2024
7.76%
2025+
9.61%
Notes:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the
flat baseline, la, and dynamic baseline, lb, please see Section X.A.I.
b Vocational vehicles modeled in MOVES include heavy heavy-duty, medium heavy-duty, and light
heavy-duty vehicles. However, for light heavy-duty vocational vehicles, class 2b and 3 vehicles are not
included in the inventories for the vocational sector. Instead, all vehicles with GVWR of less than
14,000 lbs were modeled using the energy rate reductions described below for HD pickup trucks and
vans. In practice, many manufacturers of these vehicles choose to average the lightest vocational
vehicles into chassis-certified families (i.e., heavy-duty pickups and vans).
Table 11-8 Assumed Diesel APU Use during Extended Idling for Combination
Long-haul Tractor-Trailers for Alternative 2
VEHICLE TYPE
MODEL
YEARS
DIESEL APU
PENETRATION
Combination Long-Haul Tractors
2010-2020
9%
2021-2023
30%
2024+
40%

-------
Table 11-9 Estimated Reductions in Road Load Factors for Alternative 2
VEHICLE
MODEL
REDUCTION IN TIRE
REDUCTION IN
WEIGHT
TYPE
YEARS
ROLLING RESISTANCE
AERODYNAMIC
REDUCTION


COEFFICIENT
DRAG COEFFICIENT
(LB)a
Combination
Long-haul
Tractor-Trailers
2018-2020
6.0%
5.6%
-140
2021-2023
7.5%
7.1%
-140
2024+
8.5%
8.9%
-140
Combination
Short-haul
Tractor-Trailersb
2018-2020
0.9%
0%
0
2021-2023
4.8%
0.9%
0
2024+
5.5%
3.6%
0
Intercity Buses
2021-2023
6.5%
0%
0
2024+
7.6%
0%
0
Transit and
School Buses
2021-2023
0%
0%
0
2024+
2.7%
0%
0
Refuse Trucks
2021-2023
0%
0%
0
2024+
2.7%
0%
0
Single Unit
Short-haul Trucks
2021-2023
4.8%
0%
0
2024+
5.6%
0%
0
Single Unit
Long-haul Trucks
2021-2023
6.5%
0%
0
2024+
7.6%
0%
0
Motor Homes
2021-2023
3.0%
0%
0
2024+
5.9%
0%
0
Notes:
a Negative weight reductions reflect an expected weight increase as a byproduct of the other vehicle and engine
improvements.
b Vocational tractors are included in the short-haul tractor segment.
11.1.3	Alternative 3: Preferred Alternative and Standards
Alternative 3 represents the agencies' final program. This alternative consists of the final
fuel efficiency and GHG standards for HD engines, HD pickup trucks and vans, Class 2b
through Class 8 vocational vehicles, and Class 7 and 8 combination tractors. Details regarding
modeling of this final program are included in Chapter 5 of this RIA as the control case (Chapter
5.3.2.3). Note that the impacts of the final program can be found in RIA Chapters 5, 6 and 8.
11.1.4	Alternative 4: Achieving Proposed Standards with Less Lead-Time
As indicated by its description in the title above, Alternative 4 represents standards that
are effective on a more accelerated timeline in comparison to the timeline of in the proposed
Alternative 3 standards. This alternative is unchanged from Alternative 4 in the proposal. The
agencies believe that reanalyzing the same Alternative 4 provides a useful context for
commenters who supported the proposed Alternative 4.
In the NPRM, Alternatives 3 and 4 were both designed to achieve similar fuel efficiency
and GHG emission levels in the long term but with Alternative 4 being accelerated in its
implementation timeline. Specifically, Alternative 4 reflects the same or similar standard
stringency levels as the proposed Alternative 3, but 3 years sooner (2 years for heavy-duty

-------
pickups and vans), so that the final phase of the standards would occur in MY 2024, or (for
heavy-duty pickups and vans) 2025.
The estimated reductions in energy rates used in MOVES and the projected use of
auxiliary power units (APUs) during extended idling for Alternative 4 are presented in Table
11-10 and Table 11-11, respectively. The reductions in aerodynamic and tire rolling resistance
coefficients, and the absolute changes in average vehicle weight are presented in Table 11-12.
Table 11-10 Estimated Reductions in Energy Rates for Alternative 4 using Analysis Method B a
VEHICLE TYPE
FUEL
MODEL
YEARS
FUEL/CO 2
REDUCTION
Long-haul Tractor-Trailer
and HHD Vocational
Diesel
2018-2020
1.3%
2021-2023
6.6%
2024+
10.4%
Short-haul Tractor-Trailer
and HHD Vocational
Diesel
2018-2020
0.9%
2021-2023
6.9%
2024+
10.4%
Single-Frame Vocational13
Diesel and CNG
2021-2023
7.7%
2024+
13.3%
Gasoline
2021-2023
5.2%
2024+
10.3%
HD pickup trucks and vans
Diesel and
Gasoline
2021
3.50%
2022
6.88%
2023
10.14%
2024
13.28%
2025+
16.32%
Notes:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Section X.A.I.
b Vocational vehicles modeled in MOVES include heavy heavy-duty, medium heavy-duty, and light heavy-
duty vehicles. However, for light heavy-duty vocational vehicles, class 2b and 3 vehicles are not included
in the inventories for the vocational sector. Instead, all vehicles with GVWR of less than 14,000 lbs were
modeled using the energy rate reductions described below for HD pickup trucks and vans. In practice,
many manufacturers of these vehicles choose to average the lightest vocational vehicles into chassis-
certified families (i.e., heavy-duty pickups and vans).

-------
Table 11-11 Assumed Diesel APU Use during Extended Idling for Combination Long-haul Tractor-Trailers
for Alternative 4
VEHICLE TYPE
MODEL
YEARS
DIESEL APU
PENETRATION
Combination
Long-Haul
Tractors
2010-2020
9%
2021-2023
80%
2024+
90%
Table 11-12 Estimated Reductions in Road Load Factors for Alternative 4
VEHICLE
MODEL
REDUCTION IN TIRE
REDUCTION IN
WEIGHT
TYPE
YEARS
ROLLING
AERODYNAMIC
REDUCTION


RESISTANCE
DRAG COEFFICIENT
(LB)a


COEFFICIENT

Combination
2018-2020
5.5%
5.1%
-131
Long-haul
2021-2023
12.6%
19.3%
-246
Tractor-Trailers
2024+
17.9%
26.9%
-304
Combination
2018-2020
4.0%
1.6%
-41
Short-haul
2021-2023
13.0%
11.6%
-100
Tractor-Trailersb
2024+
17.6%
15.9%
-127
Intercity Buses
2021-2023
6.5%
0%
0

2024+
16.5%
0%
0
Transit Buses
2021-2023
0%
0%
0

2024+
3.0%
0%
0
School Buses
2021-2023
0%
0%
0

2024+
4.0%
0%
0
Refuse Trucks
2021-2023
0%
0%
20

2024+
3.0%
0%
25
Single Unit
2021-2023
4.8%
0%
5.8
Short-haul
2024+
13.0%
0%
7
Trucks




Single Unit
2021-2023
6.5%
0%
20
Long-haul
2024+
16.5%
0%
25
Trucks




Motor Homes
2021-2023
3.0%
0%
0

2024+
7.4%
0%
0
Notes:
a Negative weight reductions reflect an expected weight increase as a byproduct of the other vehicle and
engine improvements.
b Vocational tractors are included in the short-haul tractor segment.

-------
11.1.5 Alternative 5: More Stringent Standards
Alternative 5 represents even more stringent standards compared to Alternatives 3 and 4,
as well as the same implementation timeline as Alternative 4. As discussed in the feasibility
discussions in the Preamble, we are not adopting Alternative 5 because we cannot project that
manufacturers can develop and introduce in sufficient quantities the technologies that could be
used to meet Alternative 5 standards. We believe that for some or all of the categories, the
Alternative 5 standards are technically infeasible within the lead time allowed. We have not
fully estimated costs for this alternative for tractors and vocational vehicles because we believe
that there would be such substantial additional costs related to pulling ahead the development of
so many additional technologies that we cannot accurately predict these costs. We also believe
this alternative could result in a decrease in the in-use reliability and durability of new heavy-
duty vehicles and that we do not have the ability to accurately quantify the costs that would be
associated with such problems. Instead we merely note that costs would be significantly greater
than the estimated costs for Alternatives 3, assuming (against our view) that such standards
would be feasible at all.
The tractor and vocational vehicle standards would be based on higher adoption rates of
the projected technologies and higher effectiveness. In addition, it assumes some adoption of all-
electric vocational vehicles.
The trailer standards in Alternative 5 are more stringent than Alternatives 3 and 4, but
rely on the same technologies. The greater reductions would be projected to be achieved through
a combination of slightly higher effectiveness and higher adoption rates.
The Alternative 5 HD pickup truck and van standards in Method B are based on more
extensive use of mild and strong hybrid technology and its use by more manufacturers. The
result would be that over half of the HD gasoline pickup fleet would need to incorporate some
form of strong hybrid technology. If achievable, Alternative 5 would require the average pickup
truck or van fuel consumption and GHG emissions to decrease by approximately 4.0 percent per
year relative to Phase 1 for model years 2021, 2022, 2023, 2024 and 2025. This is more
aggressive than Alternative 3 by 1.50 percent per year over the same model years. The estimated
reductions in energy rates used in MOVES and the projected use of auxiliary power units
(APUs) during extended idling for Alternative 5 are presented in Table 11-13 and Table 11-14,
respectively. The reductions in aerodynamic and tire rolling resistance coefficients, and the
absolute changes in average vehicle weight are presented in Table 11-15.

-------
Table 11-13 Estimated Reductions in Energy Rates for Alternative 5 using Analysis Method B a
VEHICLE TYPE
FUEL
MODEL
YEARS
FUEL/CO2
REDUCTION
Long-haul
Tractor-Trailer
and HHD
Vocational
Diesel
2018-2020
1.0%
2021-2023
12.5%
2024+
17.3%
Short-haul
Tractor-Trailer
and HHD
Vocational
Diesel
2018-2020
0.6%
2021-2023
12.7%
2024+
17.2%
Single-Frame
Vocational13
Diesel and CNG
2021-2023
12.7%
2024+
18.3%
Gasoline
2021-2023
10.7%
2024+
15.3%
Urban Buses
Diesel and CNG
2021-2023
11.8%
2024+
14.4%
HD pickup trucks
and vans
Diesel and
Gasoline
2021
4.0%
2022
7.84%
2023
11.53%
2024
15.07%
2025+
18.46%
Notes:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the
flat baseline, la, and dynamic baseline, lb, please see Section X.A.I.
b Vocational vehicles modeled in MOVES include heavy heavy-duty, medium heavy-duty, and light
heavy-duty vehicles. However, for light heavy-duty vocational vehicles, class 2b and 3 vehicles are not
included in the inventories for the vocational sector. Instead, all vehicles with GVWR of less than
14,000 lbs were modeled using the energy rate reductions described below for HD pickup trucks and
vans. In practice, many manufacturers of these vehicles choose to average the lightest vocational
vehicles into chassis-certified families (i.e., heavy-duty pickups and vans).
Table 11-14 Assumed Diesel APU Use during Extended Idling for Combination Long-haul Tractor-Trailers
for Alternative 5
VEHICLE TYPE
MODEL YEARS
DIESEL APU
PENETRATION
Combination Long-
Haul Tractors
2010-2020
9%
2021+
100%

-------
Table 11-15 Estimated Reductions in Road Load Factors for Alternative 5
VEHICLE TYPE
MODEL
REDUCTION IN
REDUCTION IN
WEIGHT

YEARS
TIRE ROLLING
AERODYNAMIC
REDUCTION


RESISTANCE
DRAG
(LB)a


COEFFICIENT
COEFFICIENT

Combination Long-
haul Tractor-Trailers
2018-2020
6.1%
7.2%
-169
2021-2023
16.2%
25.06%
930
2024-2026
19.1%
31.7%
843
2027+
19.1%
33.7%
817
Combination Short-
haul Tractor-Trailersb
2018-2020
5.5%
0.9%
-23
2021-2023
16.4%
11.2%
1069
2024-2026
19.8%
13.9%
1058
2027+
19.8%
13.9%
1058
Intercity Buses
2021-2023
20.8%
0%
0
2024+
24.7%
0%
0
Transit Buses
2021-2023
0%
0%
0
2024+
12.0%
0%
0
School Buses
2021-2023
14.9%
0%
0
2024+
19.0%
0%
0
Refuse Trucks
2021-2023
0%
0%
0
2024+
12.0%
0%
0
Single Unit Short-
haul Trucks
2021-2023
6.4%
0%
9.4
2024+
10.2%
0%
15.2
Single Unit Long-
haul Trucks
2021-2023
13.3%
0%
31.5
2024+
13.3%
0%
39.4
Motor Homes
2021-2023
20.8%
0%
0
2024+
24.7%
0%
0
Notes:
a Negative weight reductions reflect an expected weight increase as a byproduct of the other vehicle and engine
improvements.
b Vocational tractors are included in the short-haul tractor segment.
11.2 How Do These Alternatives Compare in Overall GHG Emissions
Reductions and Fuel Efficiency?
As noted earlier, the agencies analyzed the impact of each alternative on both
downstream and upstream emissions using two separate methods. The results of NHTSA's
Method A are shown in Chapter 11.2.1. The results of EPA's Method B are shown in Chapter
11.2.2.
11.2.1 Comparison of Alternatives Using Method A
The following tables compare the NHTSA estimates of overall fuel consumption and
GHG emissions reductions and benefits and costs of each of the regulatory alternatives the
agencies considered. Note that for tractors, trailers, pickups and vans the agencies compared
overall fuel consumption and GHG emissions reductions and benefits and costs relative to two

-------
different baselines, described above in the section on the No Action alternative. Therefore, for
tractors, trailers, pickups and vans two results are listed; one relative to each baseline, namely
Alternative la and Alternative lb.
For Method A, NHTSA analyzed pickup and van overall fuel consumption and emissions
reductions and benefits and costs using the NHTSA CAFE model. In addition, the agencies used
EPA's MOVES model to estimate pickup and van fuel consumption and emissions and a cost
methodology that applied vehicle costs in different model years. The Method A analysis
extended through MY 2032. The agencies concluded that in these instances the choice of
baseline and the choice of modeling approach (Method A versus Method B) did not impact the
agencies' decision to propose Alternative 3 as the preferred alternative and hence the final
standards for HD pickups and vans.
Table 1 l-16through Table 11-19 summarize the key costs and benefit estimates of the
program using Method A. The first two tables show the costs and benefits using a 3 percent
discount rate under both the flat and dynamic baselines. The third and fourth tables show the
costs and benefits using a 7 percent discount rate for both baselines. Under all possible
combinations of discount rate and baseline the net benefits from highest to lowest are as follows:
Alternative 5; Alternative 3; Alternative 4; Alternative 2.

-------
Table 11-16 MY 2018-2029 Lifetime Summary of Program Benefits and Cost, Discounted at 3% (relative to
Baseline la), Method Aa
Vehicle segment
Alt 2
Alt 3
Alt 4
Alt 5
Discounted pre-tax fuel savings ($billion)
HD pickups and Vans
12.1
18.7
20.3
22.3
Vocational Vehicles
13.5
25.5
23.6
34.6
Tractors/Trailers
50.2
118.8
115.7
169.1
Total
75.7
163.0
159.6
225.9
Discounted Total technology costs ($billion)
HD pickups and Vans
3.1
6.8
8.2
9.9
Vocational Vehicles
1.6
6.6
7.1
9.5
Tractors/Trailers
9.0
11.0
11.6
26.8
Total
13.7
24.4
26.9
46.2
Discounted value of emissions reductions ($billon)
HD pickups and Vans
3.4
5.3
5.7
6.3
Vocational Vehicles
5.2
9.8
9.1
13.3
Tractors/Trailers
21.9
50.9
50.9
73.4
Total
30.5
66.0
65.7
93.0
Total costs($billion)
HD pickups and Vans
4.4
7.9
8.6
10.3
Vocational Vehicles
2.4
7.3
8.8
11.3
Tractors/Trailers
13.2
14.0
15.7
30.8
Total
20.0
29.2
33.1
52.4
Total benefits($billion)
HD pickups and Vans
18.1
28.1
30.4
33.3
Vocational Vehicles
20.2
37.8
35.1
51.2
Tractors/Trailers
78.1
179.8
176.5
255.5
Total
114.1
245.7
242.0
340.0
Net benefits($billion)
HD pickups and Vans
13.7
20.2
21.8
23.0
Vocational Vehicles
17.8
30.5
26.3
39.9
Tractors/Trailers
64.9
165.8
160.9
224.7
Total
94.1
216.5
208.9
287.6
Note:
" For an explanation of analytical Methods A and B, please see Section I.D; for an explanation
of the less dynamic baseline, la, and more dynamic baseline, lb, please see Section X. A. 1.

-------
Table 11-17 MY 2018-2029 Lifetime Summary of Program Benefits and Costs, Discounted at 3% (relative to
Baseline lb), Method Aa
Vehicle segment
Alt 2
Alt 3
Alt 4
Alt 5
Discounted pre-tax fuel savings ($billion)
HD pickups and Vans
10.7
17.4
19.5
21.9
Vocational Vehicles
13.5
25.5
23.6
34.6
Tractors/Trailers
37.6
106.2
103.1
156.5
Total
61.8
149.1
146.2
213.0
Discounted Total technology costs ($billion)
HD pickups and Vans
2.8
6.4
7.5
9.8
Vocational Vehicles
1.6
6.6
7.1
9.5
Tractors/Trailers
8.8
10.7
11.3
26.6
Total
13.2
23.7
25.9
45.9
Discounted value of emissions reductions ($billon)
HD pickups and Vans
3.0
4.9
5.5
6.2
Vocational Vehicles
5.2
9.8
9.1
13.3
Tractors/Trailers
16.4
45.4
45.4
67.9
Total
24.6
60.1
60.0
87.4
Total costs($billion)
HD pickups and Vans
4.0
7.4
8.6
10.0
Vocational Vehicles
2.4
7.3
8.8
11.3
Tractors/Trailers
12.9
13.8
15.5
30.6
Total
19.3
28.5
32.9
51.9
Total benefits($billion)
HD pickups and Vans
16.0
26.0
29.2
32.7
Vocational Vehicles
20.2
37.8
35.1
51.2
Tractors/Trailers
59.2
161.0
157.7
236.7
Total
95.4
224.8
222.0
320.6
Net benefits($billion)
HD pickups and Vans
12.0
18.6
20.6
22.7
Vocational Vehicles
17.8
30.5
26.3
39.9
Tractors/Trailers
46.3
147.2
142.2
206.1
Total
76.1
196.3
189.1
268.7
Note:
" For an explanation of analytical Methods A and B, please see Section I.D; for an explanation
of the less dynamic baseline, la, and more dynamic baseline, lb, please see Section X. A. 1.

-------
Table 11-18 MY 2018-2029 Lifetime Summary of Program Benefits and Cost, Discounted at 7% (relative to
Baseline la) Method Aa
Vehicle segment
Alt 2
Alt 3
Alt 4
Alt 5
Discounted pre-tax fuel savings ($billion)
HD pickups and Vans
7.1
10.9
11.9
13.0
Vocational Vehicles
7.1
13.4
12.5
18.5
Tractors/Trailers
26.6
62.7
61.8
90.7
Total
40.8
87.0
86.2
122.2
Discounted Total technology costs ($billion)
HD pickups and Vans
2.2
4.8
5.9
7.0
Vocational Vehicles
1.1
4.4
4.8
6.5
Tractors/Trailers
6.2
7.4
8.0
18.5
Total
9.5
16.6
18.7
32.0
Discounted value of emissions reductions ($billon)
HD pickups and Vans
3.1
4.8
5.2
5.7
Vocational Vehicles
4.2
7.8
7.3
10.7
Tractors/Trailers
16.9
39.5
39.3
57.1
Total
24.2
52.1
51.8
73.5
Total costs($billion)
HD pickups and Vans
3.0
5.5
6.1
7.3
Vocational Vehicles
1.5
4.8
5.8
7.5
Tractors/Trailers
8.5
9.2
10.2
20.7
Total
13.0
19.5
22.1
35.5
Total benefits($billion)
HD pickups and Vans
11.7
18.0
19.6
21.5
Vocational Vehicles
12.1
22.6
21.1
31.0
Tractors/Trailers
47.1
108.0
106.8
155.1
Total
70.9
148.6
147.5
207.6
Net benefits($billion)
HD pickups and Vans
8.7
12.5
13.5
14.2
Vocational Vehicles
10.6
17.8
15.3
23.5
Tractors/Trailers
38.6
98.8
96.6
134.4
Total
58.0
129.1
125.4
172.1
Note:
" For an explanation of analytical Methods A and B, please see Section I.D; for an explanation
of the less dynamic baseline, la, and more dynamic baseline, lb, please see Section X. A. 1.

-------
Table 11-19 MY 2018-2029 Lifetime Summary of Program Benefits and Costs, Discounted at 7% (relative to
Baseline lb), Method Aa
Vehicle segment
Alt 2
Alt 3
Alt 4
Alt 5
Discounted pre-tax fuel savings ($billion)
HD pickups and Vans
6.3
10.1
11.5
12.9
Vocational Vehicles
7.1
13.4
12.5
18.5
Tractors/Trailers
19.9
56.1
55.2
84.1
Total
33.3
79.6
79.2
115.5
Discounted Total technology costs ($billion)
HD pickups and Vans
2.0
4.4
5.3
7.0
Vocational Vehicles
1.1
4.4
4.8
6.5
Tractors/Trailers
6.1
7.3
7.8
18.4
Total
9.2
16.1
17.9
31.9
Discounted value of emissions reductions ($billon)
HD pickups and Vans
2.7
4.4
5.0
5.6
Vocational Vehicles
4.2
7.8
7.3
10.7
Tractors/Trailers
12.7
35.3
35.1
52.8
Total
19.6
47.5
47.4
68.2
Total costs($billion)
HD pickups and Vans
2.7
5.1
6.0
7.1
Vocational Vehicles
1.6
4.8
5.8
7.5
Tractors/Trailers
8.4
9.0
10.1
20.6
Total
12.7
18.9
21.9
35.2
Total benefits($billion)
HD pickups and Vans
10.4
16.7
19.0
21.3
Vocational Vehicles
12.1
22.7
21.1
31.0
Tractors/Trailers
35.9
96.8
95.6
143.9
Total
58.4
136.2
135.7
195.2
Net benefits($billion)
HD pickups and Vans
7.7
11.6
13.0
14.2
Vocational Vehicles
10.5
17.9
15.3
23.5
Tractors/Trailers
27.5
87.8
85.5
123.3
Total
45.7
117.3
113.8
161.0
Note:
" For an explanation of analytical Methods A and B, please see Section I.D; for an explanation
of the less dynamic baseline, la, and more dynamic baseline, lb, please see Section X. A. 1.
Table 11-20 and Table 11-21 show the estimated fuel savings and GHG reductions by
considered alternatives and under both baselines. Under both baselines the reductions in both
fuel and GHG's are highest under Alternative 5, higher under Alternative 3 than Alternative 4,
and lowest under Alternative 2.

-------
Table 11-20 MY 2018-2029 Lifetime Fuel Savings and GHG Emissions Reductions by Vehicle Segment,
Relative to Baseline la, Method Aa
MY 2018 -2029 TOTAL
FUEL
REDUCTIONS
UPSTREAM &
DOWNSTREAM
GHG
REDUCTIONS
(billion gallons)
(MMT)
Alternative 2
HD Pickup Trucks/Vans
6.2
77
Vocational Vehicles
6.5
86
Tractors and Trailers
23.4
323
Total
36.1
486
Alt. 3 -
^referred Alternative
HD Pickup Trucks/Vans
9.8
120
Vocational Vehicles
12.3
162
Tractors and Trailers
55.6
767
Total
77.7
1049
Alt. 4
HD Pickup Trucks/Vans
10.6
130
Vocational Vehicles
11.4
150
Tractors and Trailers
54.0
744
Total
76.0
1024
Alt. 5
HD Pickup Trucks/Vans
11.6
143
Vocational Vehicles
16.7
219
Tractors and Trailers
78.8
1087
Total
107.1
1449
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for
an explanation of the flat baseline, la, and dynamic baseline, lb, please see Preamble
Section X.A.I.

-------
Table 11-21 MY 2018-2029 Lifetime Fuel Savings and GHG Emissions Reductions by Vehicle Segment,
Relative to Baseline lb Method Aa


UPSTREAM &

FUEL
DOWNSTREAM
MY 2018 -2029 TOTAL
REDUCTIONS
GHG
REDUCTIONS

(billion gallons)
(MMT)
Alternative 2
HD Pickup Trucks/Vans
5.5
68
Vocational Vehicles
6.5
86
Tractors and Trailers
17.5
242
Total
29.5
396
Alt. 3 -
Preferred Alternative

HD Pickup Trucks/Vans
9.0
111
Vocational Vehicles
12.4
162
Tractors and Trailers
49.7
685
Total
71.1
958
Alt. 4
HD Pickup Trucks/Vans
10.1
125
Vocational Vehicles
11.4
150
Tractors and Trailers
48.1
663
Total
69.6
938
Alt. 5
HD Pickup Trucks/Vans
11.3
140
Vocational Vehicles
16.7
219
Tractors and Trailers
72.9
1006
Total
100.9
1365
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for
an explanation of the flat baseline, la, and dynamic baseline, lb, please see Preamble
Section X.A.I.
In addition to considering lifetime GHG and fuel reductions, we have also considered
calendar year level GHG and fuel reductions for calendar years 2040 and 2050 across regulatory
alternatives under both baselines. These results are present in Table 11-22 and Table 11-23.

-------
Table 11-22 Annual GHG and Fuel Reductions Relative to the Dynamic Baseline in 2040 and 2050 using
Method Aa

UPSTREAM &
FUEL REDUCTIONS

DOWNSTREAM GHG
(BILLION GALLONS)

REDUCTIONS



(MMT C02EQ)



2040
2050
2040
2050
Alt. 2 Less Stringent - Total
49.1
57.3
3.6
4.2
Tractors and Trailers
30.9
36.6
2.2
2.7
HD Pickups & Vans
6.7
7.3
0.6
0.6
Vocational Vehicles
11.5
13.4
0.8
0.9
Alt. 3 Preferred - Total
139
166
10.2
12.3
Tractors and Trailers
102
124
7.4
9.0
HD Pickups & Vans
12.6
13.8
1.0
1.2
Vocational Vehicles
24.1
28.2
1.8
2.1
Alt. 4 More Stringent - Total
116
136
8.6
10.1
Tractors and Trailers
83.1
98.7
6.0
7.2
HD Pickups & Vans
12.6
13.8
1.1
1.2
Vocational Vehicles
20.0
23.1
1.5
1.7
Alt. 5 More Stringent - Total
167
194
12.4
14.2
Tractors and Trailers
124
146
9.0
10.6
HD Pickups & Vans
14.8
16.2
1.3
1.3
Vocational Vehicles
27.8
32.0
2.1
2.3
Note:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Section X.A.I.

-------
Table 11-23 Annual GHG and Fuel Reductions Relative to the Flat Baseline in 2040 and 2050 using Method
Aa

UPSTREAM &
FUEL REDUCTIONS

DOWNSTREAM GHG
(BILLION GALLONS)

REDUCTIONS



(MMT C02EQ)



2040
2050
2040
2050
Alt. 2 Less Stringent - Total
63.7
75.2
4.7
5.5
Tractors and Trailers
44.2
53.0
3.2
3.8
HD Pickups & Vans
8.0
8.8
0.6
0.7
Vocational Vehicles
11.5
13.4
0.9
1.0
Alt. 3 Preferred - Total
153
184
11.3
13.7
Tractors and Trailers
115
141
8.4
10.2
HD Pickups & Vans
13.8
15.1
1.1
1.3
Vocational Vehicles
24.1
28.2
1.8
2.2
Alt. 4 More Stringent - Total
131
153
9.6
11.4
Tractors and Trailers
96.5
115
7.0
8.3
HD Pickups & Vans
14.0
15.3
1.1
1.3
Vocational Vehicles
20.0
23.1
1.5
1.8
Alt. 5 More Stringent - Total
181
213
13.4
15.6
Tractors and Trailers
137
163
9.9
11.8
HD Pickups & Vans
16.0
17.6
1.4
1.5
Vocational Vehicles
27.8
32.0
2.1
2.3
Note:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat
baseline, la, and dynamic baseline, lb, please see Section X.A.I.
11.2.2 Comparison of Alternatives Using Method B
EPA's Method B analyzed the impact of each alternative on both downstream and
upstream emissions, as shown in Table 11-24. The table contains the annual GHG reductions
and fuel savings in 2040 and 2050 for each alternative relative to the flat (Alternative la)

-------
baseline, presenting both the total impacts across all regulatory categories and for each individual
regulatory category.
Table 11-24 Annual GHG and Fuel Reductions in Calendar Years 2040 and 2050, Relative to the Flat
Baseline using Analysis Method B a

UPSTREAM
FUEL REDUCTIONS

+DOWN STREAM
(BILLION GALLONS)

GHG REDUCTIONS



(MMT C02eq)



2040
2050
2040
2050
Alternative la (relative to itself)
0
0
0
0
Alt. 2 Less Stringent- Total
71.8
84.0
5.4
6.3
Tractors and Trailers
44.2
53.0
3.2
3.8
HD Pickups & Vans
16.1
17.6
1.4
1.5
Vocational Vehicles
11.5
13.4
0.9
1.0
Alt. 3 Preferred - Total
166.5
198.9
12.5
14.9
Tractors and Trailers
115.5
140.7
8.4
10.2
HD Pickups & Vans
26.9
30.0
2.2
2.6
Vocational Vehicles
24.1
28.2
1.9
2.1
Alt. 4 More Stringent- Total
144.1
168.5
10.9
12.7
Tractors and Trailers
96.5
115.1
7.0
8.3
HD Pickups & Vans
27.7
30.3
2.3
2.6
Vocational Vehicles
20.0
23.1
1.5
1.8
Alt. 5 More Stringent- Total
196.8
230.0
14.8
17.2
Tractors and Trailers
136.9
162.9
9.9
11.8
HD Pickups & Vans
32.2
35.2
2.7
3.0
Vocational Vehicles
27.8
32.0
2.1
2.4
Note:
a For an explanation of analytical Methods A and B, please see Preamble Section I.D; for an explanation of
the flat baseline, la, and dynamic baseline, lb, please see Preamble Section X. A. 1.

-------
References
1	OMB Circular A-4, September 17, 2003. Available at http://www.whitehouse.gov/omb/circulars a004 a-4.
2	NEPA requires agencies to consider a "no action" alternative in their NEPA analyses and to compare the effects of
not taking action with the effects of the reasonable action alternatives to demonstrate the different environmental
effects of the action alternatives. See 40 CFR 1502.2(e), 1502.14(d).CEQ has explained that "[T]he regulations
require the analysis of the no action alternative even if the agency is under a court order or legislative command to
act. This analysis provides a benchmark, enabling decision makers to compare the magnitude of environmental
effects of the action alternatives. [See 40 CFR 1502.14(c).] * * * Inclusion of such an analysis in the EIS is
necessary to inform Congress, the public, and the President as intended by NEPA. [See 40 CFR 1500.1(a).]" Forty
Most Asked Questions Concerning CEQ's National Environmental Policy Act Regulations, 46 FR 18026 (1981)
(emphasis added).
3	http://energv.gov/eere/vehicles/vehicle-technologies-office-21st-centurv-truck.
4	http ://www3. epa. gov/smartwav/.
5	State of California Global Warming Solutions Act of 2006 (Assembly Bill 32, or AB32).
6	Committee to Assess Fuel Economy Technologies for Medium- and Heavy-Duty Vehicles; National Research
Council; Transportation Research Board (2010). "Technologies and Approaches to Reducing the Fuel Consumption
of Medium- and Heavy-Duty Vehicles," (hereafter, "NAS 2010"). Washington, D.C. The National Academies Press.
Available electronically from the National Academies Press Website at
http://www.nap.edu/catalog.php?record_id=12845 (last accessed September 10, 2010).

-------
Chapter 12: Final Regulatory Flexibility Analysis
This chapter discusses the agencies' Final Regulatory Flexibility Analysis (FRFA) that
evaluates the potential impacts of the final standards on small entities. The Regulatory
Flexibility Act, as amended by the Small Business Regulatory Enforcement Fairness Act of 1996
(SBREFA), generally requires an agency to prepare a regulatory flexibility analysis of any rule
subject to notice and comment rulemaking requirements under the Administrative Procedure Act
or any other statute unless the agency certifies that the rule will not have a significant economic
impact on a substantial number of small entities. Pursuant to this requirement, we have prepared
an FRFA for the final rule.
Throughout the process of developing the FRFA, EPA conducted outreach and held
meetings with representatives from the various small entities that could be affected by the
rulemaking to gain feedback, including recommendations, on how to reduce the impact of the
rule on these entities. The small business recommendations stated here reflect the comments of
the small entity representatives (SERs) and members of the Small Business Advocacy Review
Panel (SBAR Panel, or 'the Panel'). NHTSA maintains obligations to evaluate small business
impacts under the Regulatory Flexibility Act, but is not required to convene a SBAR Panel. As a
joint rulemaking, EPA and NHTSA have coordinated formulation of standards, including
flexibilities for small businesses.
12.1 Overview of the Regulatory Flexibility Act
In accordance with section 609(b) of the Regulatory Flexibility Act (RFA), EPA
convened an SBAR Panel before conducting the FRFA. A summary of the Panel's
recommendations is presented in the Preamble of this final rulemaking. Further detailed
discussion of the Panel's outreach, advice and recommendations is found in the Final Panel
Report contained in the docket for this final rulemaking.1
Section 609(b) of the RFA directs the Panel to report on the comments of small entity
representatives and make findings on issues related to elements of a FRFA under section 603 of
the RFA. Those elements of a FRFA are:
•	A description of, and where feasible, an estimate of the number of small entities to which
the final rule will apply
•	A description of projected reporting, record keeping, and other compliance requirements
of the final rule, including an estimate of the classes of small entities which will be
subject to the requirement and the type of professional skills necessary for preparation of
the report or record
•	An identification, to the extent practicable, of all relevant Federal rules which may
duplicate, overlap, or conflict with the final rule
•	A description of any significant alternatives to the final rule which accomplish the stated
objectives of applicable statutes and which minimize any significant economic impact of
the final rule on small entities

-------
The RFA was amended by SBREFA to ensure that concerns regarding small entities are
adequately considered during the development of new regulations that affect those entities.
Although EPA is not required by the Clean Air Act to provide special treatment to small
businesses, the RFA requires EPA to carefully consider the economic impact that our rules will
have on small entities. The recommendations made by the Panel may serve to help lessen these
economic impacts on small entities when consistent with the Clean Air Act requirements.
12.2	Need for Rulemaking and Rulemaking Objectives
Heavy-duty vehicles are classified as those with gross vehicle weight ratings (GVWR) of
greater than 8,500 lb. Section 202(a) of the Clean Air Act (CAA) requires EPA to promulgate
emission standards for pollutant emissions from new motor vehicles and engines which
emissions cause or contribute to air pollution which may reasonably be anticipated to endanger
public health or welfare. In 2009, EPA found that six greenhouse gases (GHGs) were
anticipated to endanger public health or welfare, and that new motor vehicles and new motor
vehicle engines contribute to that pollution which endangers. As explained in Section I of the
Preamble to the final rule, the D.C. Circuit upheld this endangerment finding in its entirety (a
judgment the Supreme Court declined to review), and further held that EPA had a mandatory
duty to promulgate standards for emissions of the pollutant which contributes to the
endangerment: GHGs from new motor vehicles and engines.
The Energy Independence and Security Act of 2007 (EISA) directs NHTSA to develop
regulations to increase fuel efficiency for commercial medium- and heavy-duty on-highway
vehicles and work trucks. Fundamentally, EISA seeks energy conservation. In 2010, total fuel
consumption and GHG emissions from medium- and heavy-duty vehicles accounted for 23
percent of total U.S. transportation-related GHG emissions.
EPA and NHTS A's Phase 1 Heavy-Duty Engines and Vehicles Program, which was
finalized in September 2011 (76 FR 57106), marked the first greenhouse gas emissions and fuel
efficiency standards for heavy-duty vehicles and engines. The program addressed medium- and
heavy-duty GHG emissions and fuel efficiency through the adoption of performance-based
standards that allow manufacturers to determine the optimal mix of technologies to achieve the
necessary reductions for their vehicle fleets and engines.
Building on the Phase 1 rule, this final Phase 2 rule will reduce GHG emissions and fuel
consumption associated with the transportation of goods across the United States post-2017. The
final Phase 2 rulemaking considers changes to existing engine, GHG, and fuel efficiency
standards, as well as regulatory standards and certification requirements for previously-
unregulated new trailers pulled by semi-tractors. Manufacturers of heavy-duty engines, chassis,
vehicles and trailers will be required to incorporate GHG-reducing and fuel-saving technologies
in order to comply with the agencies' performance-based standards.
12.3	Definition and Description of Small Businesses
The RFA defines small entities as including "small businesses," "small governments,"
and "small organizations" (5 U.S.C. 601) and references the Small Business Administration for
the definition of "small businesses" using size standards based on the North American Industry

-------
Classification System (NAICS) (13 CFR 121.201). The standards being considered by EPA for
this rulemaking are expected to affect a variety of small businesses. A listing of the NAICS
codes identified as relevant to the potential rulemaking, along with their respective SBA size
thresholds, is located in Table 12-1, below. In the period between the convening of the SBAR
Panel (and Initial Regulatory Flexibility Analysis) and issuing the final rule, SBA finalized new
size standards for small business classification.2 We have updated our analysis to reflect the new
size standards and noted the changes in Table 12-1.
The agencies expect that the same industries affected by the Phase 1 rulemaking will also
be affected by the final Phase 2 rulemaking. In addition, small businesses and trailer
manufacturers are also included in the final Phase 2 rule. EPA and NHTSA used the criteria for
small entities developed by SBA as a guide to identifying Small Entity Representatives (SERs)
for this rulemaking. Table 12-1 lists industries potentially directly affected by the regulation.
The NAICS code and size thresholds are shown as well.
Table 12-1 Industry Sectors Potentially Affected by the Agencies' Action
INDUSTRY
EXPECTED IN
RULEMAKING
NAICS
CODE
NAICS
DESCRIPTION
SBA SIZE THRESHOLD
(LESS THAN OR EQUAL TO)
IRFA
FRFA
Alternative Fuel
Engine Converters
333999
Misc. General Purpose
Machinery
500 employees
811198
All Other Auto Repair &
Maintenance
$7.0M
(annual receipts)
$7.5M
(annual receipts)
HD Pick-up Trucks &
Vans
336111
Automobile Manufacturing
1,000 employees
1,500 employees
Vocational Chassis,
Class 7 & 8 Tractors
336120
Heavy-Duty Truck
Manufacturing
1,000 employees
1,500 employees
Trailers
336212
Truck Trailer Manufacturing
500 employees
1,000 employees
HD Spark-Ignition
Engines
336310
Motor Vehicle Gasoline Engine
& Engine Parts
750 employees
1,000 employees
HD Compression-
Ignition Engines
333618
Other Engine Equipment
Manufacturing
1,000 employees
1,500 employees
12.4 Summary of Small Entities to which the Rulemaking will Apply
Using the information from Table 12-1, with the agencies' certification data and
employment information from the Hoover's online business information database, EPA and
NHTSA have determined that there are small business in the following affected industries:
heavy-duty truck manufacturers (vocational chassis and glider vehicle manufacturers), heavy-
duty engine manufacturers, alternative fuel engine converters, and trailer manufacturers. The
agencies believe there are about 178 trailer manufacturers of which 147 qualify as small entities
with 1,000 employees or less. EPA and NHTSA identified ten heavy-duty engine manufacturers
that are currently certifying natural gas engines. The agencies believe nine of these companies
are small businesses. About 60 companies have filed paperwork with EPA as alternative fuel
converters. Many of these service only light-duty vehicles and light-duty trucks; we estimate
that there are 20-30 companies performing aftermarket fuel conversions with heavy-duty
vehicles and heavy-duty engines, all of which are likely to qualify as small businesses under the
Phase 2 program. Currently, 20 manufacturers that make chassis for vocational vehicles certify

-------
with EPA under the Phase 1 program and the agencies have identified an additional 19 small
vocational chassis manufacturers that are not currently certifying under Phase 1.
Glider vehicles are a subset of vehicles that will be regulated under the Phase 2
rulemaking (including for regulation of criteria emissions). Glider vehicle manufacturers
traditionally manufacture or purchase new vehicle bodies (vocational vehicles or Class 7 and 8
tractors) for use with older powertrains. These engineless vehicle bodies are often referred to as
"glider kits" and to the extent glider vehicle manufacturers rely on glider kits, they can be
referred to as assemblers and well as manufacturers. The agencies were aware of four glider
vehicle manufacturers (for whom glider vehicle production was a primary business) during the
SBAR Panel process and we identified three of these manufacturers as small entities. We are not
aware of any small businesses that produce glider kits for others to assemble.1 Public comments
on the proposed rule indicated that there are more than 1,200 purchasers of glider kits, and we
presume they would all meet the Act's definition of "manufacturer," which includes anyone who
assembles motor vehicles. See Preamble Section I.E.(l)(c). This large number of businesses
that were not accounted for during the SBAR Panel is largely a result of our focus on glider
manufacturers for whom glider vehicle production is a primary business. We note that almost
every repair shop that is capable of overhauling truck engines is also capable of assembling a
glider vehicle. Perhaps most have, at some point, installed a used highway engine in a glider kit.
Producing glider vehicles is quite clearly not a major business focus for most of these additional
companies. Nevertheless, we believe that a clear majority of the companies assembling glider
vehicles, including those that do so as a side business, qualify as small businesses.
12.5	Related Federal Rules
The Phase 1 rulemaking continues to be in effect in the absence of this final rule. The
Panel noted that it was aware that the final Phase 2 rule would be a joint action by EPA and the
Department of Transportation (DOT), through NHTSA, as in the Phase 1 rulemaking. We are
also aware of other state and Federal rules related to heavy-duty vehicles and to the final Phase 2
rule under consideration. NHTSA has safety requirements for medium- and heavy-duty vehicles
located at 49 CFR part 571. California adopted its own greenhouse gas initiative, which places
aerodynamic requirements on trailers used in long-haul applications. None of these existing
regulations were found to conflict with the final rulemaking.
12.6	Projected Reporting, Recordkeeping, and Other Compliance
Requirements
For any emission control program, EPA must have assurances that the regulated products
will meet the standards. Certification and in use requirements are explicit statutory
requirements. See e.g. CAA section 203 (a). The program that EPA and NHTSA are adopting
for manufacturers subject to this rule includes testing, reporting, and recordkeeping
requirements. Testing requirements for these manufacturers includes use of EPA's Greenhouse
gas Emissions Model (GEM) vehicle simulation tool to obtain the overall CO2 emissions rate for
1 Although this discussion is written based on the assumption that no small businesses produce glider kits for others
to assemble, the conclusions would also be valid with respect to small entities that produce glider kits for sale,
should they exist.

-------
certification of vocational chassis and trailers, aerodynamic testing to obtain aerodynamic inputs
to GEM for some trailer manufacturers, and engine dynamometer testing for alternative fuel
engine converters to ensure their conversions meet the CO2, CH4 and N2O engine standards.
Reporting requirements include emissions test data or model inputs and results, technical data
related to the vehicles, and end-of-year sales information. Manufacturers will have to keep
records of this information.
12.7 Regulatory Flexibilities
The Panel developed a range of regulatory flexibilities intended to mitigate the impacts of
the final rulemaking on small businesses, and recommended that EPA propose and seek
comment on the flexibilities. The Panel's findings and discussions are based on the information
that was available during the term of the Panel and issues that were raised by the SERs during
the outreach meetings and in their written comments. It was agreed that EPA should consider
the issues raised by the SERs (and issues raised in the course of the Panel) and that EPA should
consider the comments on flexibility alternatives that would help to mitigate any negative
impacts on small businesses.
Alternatives discussed throughout the Panel process include those offered in the
development of the upcoming rule. Though some of the recommended flexibilities may be
appropriate to apply to all entities affected by the rulemaking, the Panel's discussions and
recommendations are focused mainly on the impacts, and ways to mitigate adverse impacts, on
small businesses. A summary of the Panel's recommendations, along with those provisions that
we are finalizing in this action, are detailed below. A full discussion of the regulatory
alternatives and hardship provisions discussed and recommended by the Panel, all written
comments received from SERs, and summaries of the two outreach meetings that were held with
the SERs can be found in the SBREFA Final Panel Report, located in the rulemaking docket.3 In
addition, all the flexibilities that are being adopted in the rulemaking for small businesses, as
well as those for all entities that may be affected by the rulemaking, are described in the
Preamble to the final rule.
12.7.1 Heavy-Duty Highway Engine Manufacturers and Engine Converter
Flexibilities
12.7.1.1 SBAR Panel Recommendations
Based on the comments received from SERs, the Panel recommended not having separate
standards for small business natural gas engine manufacturers. The Panel believed this would
discourage entrance into this emerging market by adding unnecessary costs to a technology that
has the potential to reduce CO2 tailpipe emissions. In addition, the Panel stated that it believes
additional leakage requirements beyond a sealed crankcase for small business natural gas-fueled
CI engines and requirements to follow industry standards for leakage could be waived for small
businesses with minimal impact on overall GHG emissions.
To reduce the compliance burden of small business engine converters who convert
engines in previously-certified complete vehicles, the Panel recommended allowing engine
compliance to be sufficient for certification. This would mean the converted vehicle would not

-------
need to be recertified as a vehicle. This flexibility would eliminate the need for these small
manufacturers to gather all the additional component-level information (e.g., transmission data,
aerodynamic performance, tire rolling resistance) in addition to the engine CO2 performance
necessary to properly certify a vehicle with GEM. In addition, the Panel recommended that
small engine converters be able to submit an engineering analysis, in lieu of measurement, to
show that their converted engines do not increase N2O emissions. Many of the small engine
converters are converting SI engines, and the catalysts in these engines are not expected to
substantially impact N2O production. Small engine converters that convert CI engines could
likely certify by ensuring that their controls require changes to the SCR dosing strategies.
The Panel recommended that small business engine manufacturers receive a one-year
delay in implementation at the beginning of Phase 2, which would allow these manufacturers to
begin certifying in MY 2019. Additionally, the Panel recommended that small engine
manufacturers producing alternative-fuel engines receive a one-year delay in implementation for
each increase in stringency throughout the program. This flexibility would provide additional
lead time to obtain the necessary equipment and perform calibration testing if needed.
12.7.1.2 What We Proposed
The agencies proposed the Panel's recommended regulatory flexibility provisions for
small businesses producing alternative-fuel engines. EPA and NHTSA proposed to offer these
entities a one-year delay in implementation at the start, and small manufacturers of alternative
fueled-engines were given an additional one-year delay for each increase in stringency
throughout the program. The agencies believed a majority of these small businesses would
manufacture their engines from standard gasoline or diesel engine architectures and additional
lead time was warranted.
The Phase 2 proposal included three new requirements for companies that manufacture
heavy-duty engines: measuring N2O emissions, reporting CO2 and CH4 emissions (which are
already measured for meeting criteria standards), and generating an engine fuel map for vehicle
manufacturers installing the subject engines. These requirements apply to all new engines,
including those fueled by gasoline and diesel alternatives such as natural gas. The agencies did
not propose separate standards for alternative fuel engines.
Alternative fuel engine converters generally modify engines that are no longer new.
Instead, they convert previously certified engines or vehicles to run on alternative fuels. In
accordance with Clean Air Act section 203, these converters are required to ensure that they are
not tampering with emissions controls. In the Phase 2 proposal, we clarified that companies
converting Phase 2-certified vehicles would be subject to CO2 and CH4 standards in the same
way that they are subject to criteria standards, but the agencies believe an engine conversion is
unlikely to increase N2O generation, and we proposed to allow engine converters to submit an
engineering analysis to demonstrate compliance with the N2O standard. See 80 FR 40551.

-------
12.7.1.3 Public Comments Received on the NPRM and What We're Finalizing
We did not receive comments on the small-business relief provisions as they apply for
engine manufacturers and alternative fuel converters. We are accordingly adopting the proposed
flexibilities.
12.7.2 Vocational Vehicle Chassis Manufacturer Flexibilities
12.7.2.1	SBAR Panel Recommendations
The Panel recommended proposing less stringent standards for emergency vehicle chassis
manufactured by small businesses. The Panel stated that it believes it is feasible for small
manufacturers to install a Phase 2-compliant engine, but recommended that the rulemaking
request comment on whether the use of LLR tires will provide enough CO2 benefits to justify
requiring small business emergency chassis manufacturers to adopt them. In addition, the Panel
recommended a simplified certification approach for small manufacturers who make chassis for
emergency vehicles that reduces the number of inputs these manufacturers would need to obtain
for GEM.
The Panel recommended proposing a low volume exemption for small business custom
chassis manufacturers based on the volume of sales. Similar to the recommendation for
emergency vehicle chassis manufacturers, the Panel stated it believes it is feasible to require
installation of a Phase 2-compliant engine and recommended that EPA request comment on the
benefits of LRR tires in this market segment. The Panel also recommended that the rulemaking
request comment on how to design a small business exemption by means of a volume exemption
and what sales volume would be an appropriate threshold.
12.7.2.2	What We Proposed
EPA and NHTSA proposed a flexibility for all emergency vehicles that included fewer
technology requirements and a simplified certification approach. Consistent with the
recommendations of the Panel, the agencies requested comments on how to design a small
business vocational vehicle program, including comments on a possible small volume threshold
below which some small business exemption may be available.
12.7.2.3	Public Comments Received on the NPRM and What We're Finalizing
Consistent with the recommendations of the Panel, the agencies are adopting less
stringent emergency vehicle standards using a simplified GEM. Innovus commented in support
of a small volume threshold for small businesses of either 200 vehicles per year or a different
threshold set based on the market share of the entity. Autocar requested further consideration of
the small business concerns of manufacturers of specialty vehicle applications, specifically
recommending a low volume threshold if the agencies are not inclined to use a manufacturer's
business size as grounds for an exemption. Examples of specialty vehicles listed by Autocar
include street sweepers, asphalt blasters, aircraft deicers, sewer cleaners, and concrete pumpers.
Innovus also requested additional flexibility for meeting OBD requirements. Capacity Trucks
commented that the terminal tractor industry is primarily comprised of small businesses who

-------
produce a total of less than 6,000 terminal tractors per year, 70 percent of which are fully off-
road vehicles.
In considering these comments, the agencies are adopting a custom chassis program for
which all manufacturers are eligible. The program includes less stringent standards and a
simplified GEM process, where the technology packages have been tailored to specific vehicle
applications, and each technology has been determined to be feasible and effective for those
vehicles. See Section V of the Preamble for more details.
12.7.3 Glider Vehicle Manufacturer Flexibilities
12.7.3.1	SBAR Panel Recommendations
The Panel stated that it believes that the number of vehicles produced by small business
glider vehicle manufacturers is too small to have a substantial impact on the total heavy-duty
GHG inventory.2 The Panel also stated that there should be an allowance to produce some
number of glider kits for legitimate purposes, such as for newer vehicles badly damaged in
crashes. The Panel therefore recommended proposing an explicit allowance for existing small
businesses to continue assembling glider vehicles without having to comply with the GHG
requirements. The Panel also recommended that any regulations for glider production be flexible
enough to allow sales levels as high as the peak levels in the 2010-2012 timeframe.
12.7.3.2	What We Proposed
The exemption that the agencies proposed for glider vehicle manufacturers was expected
to encompass small glider manufacturers. Small manufacturers who assemble 300 or fewer
gliders per year would be exempt from certification, up to each company's documented
production volumes from 2010-2014. Any additional gliders produced would have to meet the
vehicle and engine standards for their respective regulatory categories in the current model year.
For instance, tractor gliders would have to meet the tractor standards and vocational chassis
would meet the vocational standards, and for both, the engines would need to meet all applicable
GHG and criteria emission standard for the year the glider vehicle is completed.
We believed the flexibilities offered to custom chassis vocational vehicles would also
reduce the requirements of any small businesses that manufacturer vocational gliders, such as
cement mixers and emergency vehicles.
12.7.3.3	Public Comments Received on the NPRM and What We're Finalizing
Engine and vehicle manufacturers took opposing positions. Some supported the
proposed approach. Others stated that the proposed provisions exceeded EPA's authority to set
emission standards for new engines and new vehicles, in addition to objecting to the detailed
provisions as a matter of policy. See Preamble Section I.E. and Response to Comments (RTC)
Section 14.2. However, the most helpful comments were those that allowed EPA to target
2 The Panel did not have accurate data on annual glider vehicle production at the time of the report, but it believed
the production to be less than 5,000 per year, which is half of the current rate or less. The Panel also addressed only
GHG impacts, not impacts of vast increases in criteria pollutant emissions.

-------
flexibility for glider vehicles that serve an arguably legitimate purposes (such as reclaiming
relatively new powertrains from vehicles chassis that fail prematurely), without causing
substantial adverse environmental impacts.
We are finalizing the proposed glider-related provisions but have made several revisions
in recognition of the differences between gliders produced to circumvent the 2010 criteria
pollutant emission standards and those manufactured for other more legitimate purposes. The
provisions being finalized are intended to allow a transition to a long-term program in which
manufacture of glider vehicles from glider kits is permissible consistent with the original reason
OEM manufacturers began to offer glider kits - to allow the reuse of relatively new powertrains
from damaged vehicles. The long-term program as well as the transitional program are
summarized below. See Section XIII.B of the FRM for a complete description of these
provisions.
Under the provisions being finalized for the long-term program, all glider vehicles will
need to be covered by both vehicle and engine certificates. The vehicle certificate will require
compliance with the GHG vehicle standards of 40 CFR part 1037. The engine certificate will
require compliance with the GHG engine standards of 40 CFR part 1036, plus the criteria
pollutant standards of 40 CFR part 86. Used engines (including rebuilt/remanufactured engines)
may be installed in the gliders without meeting engine standards applicable for the year of glider
assembly, provided the engines are within their regulatory useful life (or meet similar criteria).
EPA is also finalizing a transitional program that will allow glider kit/vehicle
manufacturers additional flexibility. The first step allows significant production of glider
vehicles under the Phase 1 approach, but limits each manufacturer's combined production of
glider kits and glider vehicles at the manufacturer's highest annual production of glider kits and
glider vehicles for any year from 2010 to 2014. All vehicles within this cap will remain subject
to the existing Phase 1 requirements (for both engines and vehicles). Any glider kits or glider
vehicles produced beyond this cap will be subject to all requirements applicable to new engines
and new vehicles for MY 2017. Other than the 2017 production limit, EPA will continue the
Phase 1 approach until January 1, 2018. This allows small businesses to produce glider kits up
to the production limit without new constraints. Large manufacturers producing complete glider
vehicles remain subject to the 40 CFR part 1037 GHG vehicle standards, as they have been since
the start of Phase 1. However large manufacturers may provide exempted glider kits to small
businesses during this time frame, and they would not be required to obtain a vehicle certificate
for them. However, these exempted glider kits would count against the glider kU manufacturers'
production cap for 2017.
Effective January 1, 2018, the long-term program begins generally, but with certain
transitional flexibilities. In other words, except for the following allowances, glider vehicles will
need to comply with the long-term program. The exceptions are:
• Small businesses may produce a limited number of glider vehicles without
meeting either the engine or vehicle standards of the long-term program. Larger
vehicle manufacturers may provide glider kits to these small businesses without
the assembled vehicles meeting the applicable vehicle standards. This number is

-------
limited to the small vehicle manufacturer's highest annual production volume in
2010 through 2014 or 300, whichever is less.
•	Model year 2010 and later engines are not required to meet the Phase 1 GHG
engine standards.
•	Glider vehicles conforming to the previously certified vehicle configuration of the
donor vehicle do not need to be recertified to current vehicle standards.
These 2018 allowances mostly continue after 2020, but effective January 1, 2021, the
completed vehicle will need to meet the vehicle standards, even if the engine is exempt under the
small manufacturer provisions. In practice, this will likely mean that the large manufacturers
providing glider kits to small manufacturers will need to meet the vehicle standards for the
completed vehicle by obtaining a certificate and delegating final assembly to the assembler.
This transitional program combined with the additional flexibility in the long-term
program will achieve the stated goal of the Panel, which was to have any regulations for glider
production be flexible enough to allow sales levels as high as the peak levels in the 2010-2012
timeframe.
12.7.4 Trailer Manufacturer Flexibilities
12.7.4.1 SBAR Panel Recommendations
12.7.4.1.1 Box Trailers
Box trailer manufacturers have the benefit of relying on the aerodynamic technology
development initiated through EPA's voluntary SmartWay program. The Panel acknowledged
EPA's plan to adopt a simplified compliance program for all manufacturers, in which
aerodynamic device manufacturers have the opportunity to test their devices and register their
data with EPA as technologies that can be used by trailer manufacturers in their trailer
certification. This pre-approved data strategy is intended to provide all trailer manufactures a
means of complying with the standards without testing. Upon the completion of the SBREFA
Panel process, it was unclear if this strategy would be available indefinitely, or if it would be an
interim flexibility to allow manufacturers to ease into a testing-only compliance program. The
Panel recommended that, in the event that this strategy is limited to the early years of the trailer
program for all manufacturers, small manufacturers should continue to be given the option to use
pre-approved devices in lieu of testing.
The Panel stated its belief that, in the event that small trailer manufacturers adopt pre-
approved aerodynamic technologies and the appropriate tire technologies for compliance, it
would not be necessary to require the use of a vehicle emissions model, such as GEM, for
certification. Instead, the Panel stated that it could be possible for manufacturers to simply
report to EPA that all of their trailers include approved technologies.

-------
12.7.4.1.2
Non-Box Trailers
The Panel recommended that EPA not base a standard for non-box trailers on
performance of aerodynamic devices. Some of the non-box trailer manufacturer SERs have seen
prototype-level demonstrations of aerodynamic devices on non-box trailers. However, most
non-box trailer SERs identified unique operations in which their trailers are used that preclude
the use of those technologies.
Some non-box trailer manufacturers have experience with LRR tires and ATI systems.
However, the non-box trailer manufacturer SERs indicated that LRR tires are not currently
available for some of their trailer types. The SERs noted that tire manufacturers are currently
focused on box trailer applications and that there are only a few LRR tire models that meet the
needs of their customers. The Panel stated that it believes EPA should ensure appropriate
availability of these tires in order for it to be deemed a feasible means of achieving these
standards and recommended a streamlined compliance process based on the availability of
technologies. The Panel suggested that the best compliance option from a small business
perspective would be for the agencies to pre-approve tires once they are available in sufficient
quantities on the market, similar to the approach proposed for aerodynamic technologies, and to
maintain a list that could be used to exempt small businesses when no suitable tires are available.
However, the Panel stated that it recognizes the difficulties of maintaining an up-to-date list of
certified technologies. The Panel recommended that, if the rulemaking does not adopt the list-
based approach, the agency consider a simplified letter-based compliance option that allows
manufacturers to petition the agencies for an exemption if they are unable to identify tires that
meet the LRR performance requirements on a trailer family basis.
12.7.4.1.3	Trailers with Unique Use Patterns
The Panel recommended excluding all trailers that spend a significant amount of time in
off-road applications. These trailers may not spend much time at highway speeds and
aerodynamic devices may interfere with the vehicle's intended purpose. Additionally, tires with
lower rolling resistance may not provide the type of traction needed in off-road applications.
12.7.4.1.4	General Flexibilities for All Small Trailer Manufacturers
The Panel stated that it recognizes that some manufacturers, who have diverse product
lines and high sales volumes, may benefit from an emissions averaging, banking and trading
(ABT) strategy. However, due to the custom-order nature of the trailer industry, SERs have
expressed their concern that ABT may provide an opportunity for historically loyal customers or
customers with large fractions of a manufacturer's business to bargain for the portion of a
manufacturer's sales that have minimal requirements. Based on the low volume of sales and
niche market of many small business trailer manufacturers, small businesses in particular may
have little leverage in this situation and risk losing their customers to larger manufacturers who
have credits to spare. In addition, the accounting and reporting burdens of ABT may preclude
small businesses from utilizing this flexibility.
Due to the potential for reducing a small business's competitiveness compared to the
larger manufacturers, as well as the ABT recordkeeping burden, the Panel recommended EPA

-------
consider small business flexibilities to allow small entities to opt out of ABT without placing
themselves at a competitive disadvantage to larger firms that adopt ABT, such as a low volume
exemption or requiring only LRR where appropriate. The Panel recommended that EPA also
consider flexibilities for small businesses that would ease and incentivize their participation in
ABT, such as streamlined the tracking requirements for small businesses. In addition, the Panel
recommended that EPA request comment on the feasibility and consequences of ABT for the
trailer program and additional flexibilities that would promote small business participation.
In addition, for all trailer types that will be included in the Phase 2 program, the Panel
recommended a 1-year delay in implementation for small trailer manufacturers at the start of the
program to allow them additional lead time to make the proper staffing adjustments and process
changes, and possibly add new infrastructure to meet these requirements. In the event that the
agencies are unable to provide pre-approved technologies for manufacturers to choose for
compliance, the Panel recommended that the standards provide small business trailer
manufacturers an additional 1-year delay for each subsequent increase in stringency. This
additional lead time would allow these small businesses to research and market the technologies
required by the new standards.
12.7.4.2 What We Proposed
The agencies proposed many of the Panel's recommendations for small business trailer
manufacturers, and sought comment on the possibility of a small volume exemption. While
many of the smallest trailer manufactures sell significantly fewer trailers than the largest small
manufacturers, many of the smallest trailer manufacturers produce specialty trailers that would
have been candidates for exemption under the proposed off-highway or heavy-haul provisions.
Testing requirements for small businesses were largely reduced by proposed provisions
for both large and small trailer manufacturers. A majority of the small trailer manufacturers
produce non-box trailers, and we did not propose standards predicated on use of aerodynamic
controls for these trailers, which reduced the number of technologies to investigate, market, and
implement. As is seen in the Phase 1 tractor program, we expect that tire rolling resistance will
be measured by tire manufacturers and information needed for compliance would be presented to
trailer manufacturers when they purchase their tires, and no additional testing will be needed.
See 40 CFR 1037.650 of the Phase 1 regulations and 40 CFR 1037.620 of the final Phase 2
regulations.
The agencies proposed an option for pre-approved aerodynamic device data to be made
available to box trailer manufacturers for use in complying with aerodynamic requirements. We
did not set an end date for this provision and the pre-approved data would eliminate the
requirement for box trailer manufacturers to complete aerodynamic performance testing for
certification throughout the program. EPA and NHTSA expect small business box trailer
manufacturers will use the pre-approved aerodynamic devices for most of their trailers.
Additionally, the agencies proposed a simplified compliance program with options to
demonstrate trailer performance without requiring the trailer manufacturers to perform vehicle
modeling using GEM. The design-based standards proposed for non-box trailer manufacturers
would have required the use of LRR tires and ATI systems without testing of any type. The

-------
agencies developed a GEM-based equation for each box trailer subcategory that reproduces the
CO2 results of the vehicle model and box trailer manufacturers will simply insert the
performance data from any technologies installed to calculate their compliance values. As a
result, we proposed that trailer manufacturers would not use GEM for compliance in this final
rule.
For the small business trailer manufacturers that produce trailers that are regulated in this
program, EPA proposed a one-year implementation delay at the beginning of the program to
allow small business trailer manufacturers to demonstrate compliance starting in model year
2019, providing small businesses additional lead time to make the proper staffing adjustments
and process changes and possibly add new infrastructure to meet their requirements. NHTSA's
standards are voluntary until MY 2021. Since small business trailer manufacturers will already
be required to comply with EPA standards when NHTSA's fuel efficiency standards will begin,
NHTSA does not believe that an additional year of delay to comply with its fuel efficiency
standards will provide beneficial flexibility.
The agencies proposed a limited averaging program for box trailer manufacturers. The
five largest trailer manufacturers produce over 85 percent of the dry and refrigerated vans in the
market. We did not propose an option to bank or trade credits, because the volume of credits that
could be generated by large manufacturers has the potential to exceed the total sales of small
manufacturers. In such a scenario, a small manufacturer could lose all of its customers to larger
manufacturers that could sell the same number of trailers with fewer or no technologies installed.
The limited averaging program was restricted to averaging within a single model year, but the
agencies proposed to allow deficits to be carried-over for three years.
12.7.4.3 Public Comments Received on the NPRM and What We're Finalizing
EPA and NHTSA are finalizing the option for trailer manufacturers to use pre-approved
aerodynamic device data submitted by device manufacturers. We did not set an end date for this
provision and the pre-approved data would eliminate the requirement for box trailer
manufacturers to complete aerodynamic performance testing for certification throughout the
program. The agencies expect small business box trailer manufacturers will take advantage of
the pre-approved aerodynamic devices for most of their trailers, which will significantly reduce
or eliminate their testing burden.
The agencies did not receive any comments recommending an appropriate sales volume
that could qualify manufacturers for low-volume exemption. The Truck Trailer Manufacturers
Association (TTMA) and the American Trucking Associations (ATA) provided comments
suggesting that additional trailer types should be excluded from the program based on these
trailers' typical operational characteristics. We recognize that many trailers in the proposed non-
box subcategory have unique physical characteristics for specialized operations that may make
use of LRR tires and/or tire pressure systems difficult or infeasible. Instead of focusing on trailer
characteristics that indicated off-highway use, the agencies have identified three specific types of
non-box trailers that represent the majority of non-box trailers and that we believe are designed
and mostly used in on-road applications: tanks, flatbeds, and container chassis.

-------
We believe that manufacturers of tanks, flatbeds, and container chassis can relatively
easily install LRR tires and tire pressure systems, and that customers will benefit from using
these technologies. We are limiting the final non-box trailer program to tanks, flatbeds, and
container chassis. All other non-box trailers are excluded from the Phase 2 trailer program, with
no regulatory requirements. This exclusion reduces the number of small businesses in the trailer
program from 147 to 74 companies. With no regulatory requirements, these companies are
expected to have zero burden.
Additionally, the agencies are adopting provisions that would increase the number of
eligible tire pressure systems that can be installed for compliance. We proposed to only allow
automatic tire inflation (ATI) systems, but we received comments from manufacturers that were
concerned about the cost and availability of ATI systems for the trailer industry. The agencies
agree that tire pressure monitoring (TPM) systems have the potential to promote proper tire
inflation and that allowing lower cost systems will increase acceptance of the technologies. The
agencies recognize that TPM systems have the potential to promote proper tire inflation and that
allowing lower cost systems will increase acceptance of the technologies. We are finalizing
provisions to allow TPM systems to receive credit. The non-box trailers, which have design-
based tire standards, will be deemed to comply if they have a minimum of a TPM system and
lower rolling resistance tires. The increased the number of options for tire pressure systems and
inclusion of the cheaper TPM systems will improve the availability of technologies and reduce
the technology cost.
Comments from the trailer industry were strongly opposed to any averaging at any point
in the program, citing the highly competitive nature of the industry combined with a wide range
of product diversity among companies likely leading to that it would unfairly benefit the few
larger companies and be impossible to implement for many of the companies with limited
product diversity. Additionally, compared to other industry sectors, trailer manufacturers noted
that they can have little control over what kinds of trailer models their customers demand and
thus limited ability to manage the mix and volume of different products. Comments from Strick,
a small business box trailer manufacturer and a SER during the Panel process, opposed
averaging and noted the unfair advantage that larger manufacturers would have in an averaging
program.
The agencies generally agree with these concerns, and the final program limits the option
for trailer manufacturers to apply averaging to MYs 2027 and later trailers. We believe this
delay will provide the trailer manufacturers sufficient time to develop, evaluate, and market new
technologies, and become familiar with the compliance process. As the standards become more
stringent, the agencies believe the trailer manufacturers may wish for additional flexibilities in
achieving the standards. The final program limits averaging to within a given model year and
does not include banking or trading. Similar to the proposal, we are allowing deficits to be
carried-over for up to three years.
TTMA commented that all trailer manufacturers are "small businesses" relative to other
heavy-duty industries and that the one-year delay would divert sales to small businesses for that
model year. Wabash National Corporation (Wabash) argued that providing a flexibility is not
required by the RFA and not authorized by the Clean Air Act. The agencies believe that small
businesses do not have the same resources available to become familiar with the regulations,

-------
make process and staffing changings, or evaluate and market new technologies as their larger
counterparts. We believe a one-year delay will provide sufficient time for small businesses to
address these issues, without a large CO2 and fuel consumption impact. EPA is required to
consider issues of cost and lead time under section 202 (a)(2), and can reasonably differentiate
among classes of regulated entities based on these factors, and is doing so here. The cumulative
annual production of all of the small business box trailer manufacturers is less than the annual
production of the four largest manufacturers. We expect any diverted sales for this one year will
be a small fraction of the larger manufacturers' production and we are accordingly finalizing the
one-year delay for all small business trailer manufacturers.
12.8 Projected Economic Effects of the Final Rulemaking
This section summarizes the economic impact of the final Phase 2 rulemaking on small
businesses. To gauge this impact, the agencies employed a cost-to-sales ratio test to determine if
small businesses would be impacted by less than one percent, between one and three percent, and
above three percent of their sales. The costs used in this analysis for the final requirements are
based on the cost estimates developed for Chapters 2 and 7 of this RIA, and the Information
Collection Request (ICR) required by the Paperwork Reduction Act. As noted below, the
agencies believe that there will not be a significant economic impact on a substantial number of
small entities as a result of the Phase 2 rulemaking.
12.8.1 Heavy-Duty Engine Manufacturer Economic Effects
As described above, the expected incremental burden for engine manufacturers to
demonstrate compliance with the new greenhouse gas emission standards is to create a fuel map,
measure N2O emissions, and to report CO2 and CH4 emission values (which are already
measured for certification related to criteria emissions).
We expect very small engine manufacturers to rely on contract test labs to perform
emission testing, and that these labs already have N2O testing capability. As a result, there
should be no necessary capital expenditures to meet this requirement. Rather, we estimate the
incremental cost of measuring N2O for any hired certification testing to be on the order of $500
for each engine family. Manufacturers with greater resources might run their own laboratories,
in which case they would need to purchase additional analytical equipment for measuring N2O;
however, this would only be the case if the companies' revenues would support this approach as
a more cost-effective way of meeting the regulatory requirements.
The agencies believe it will cost $2,400 to generate a fuel map, which includes an
estimated eight hours of dynamometer testing time at a rate of $300 per hour.
The smallest natural gas engine manufacturer certifies two engine families with 10
employees and annual revenue of $2.4 million. Their total cost is expected to be $5,900 to meet
new requirements across their product line. These costs would be spread over several years,
especially considering the possibility of using carryover data to certify over multiple model
years. However, even applying this cost to a single year would represent only 0.2 percent of
annual revenue.

-------
The second smallest natural gas engine manufacturer certifies five engine families with
20 employees and annual revenue of $4.7 million. Their total cost is expected to be $14,700 to
meet new requirements across their product line. Concentrating these costs again to a single
model year represent only 0.3 percent of annual revenue. This worst-case assessment shows that
the new requirements will not be a substantial burden for any small engine manufacturers.
12.8.2	Alternative Fuel Engine Converter Economic Effects
Alternative fuel converters continue to be subject to criteria standards. The incremental
burden of this program is for reporting CO2 and CH4 emissions (where reporting is required),
and performing an engineering analysis to demonstrate that modified engines continue to meet
the N2O standard. CO2 and CH4 emissions are currently measured to demonstrate compliance
with CO and nonmethane hydrocarbon standards. We consider the additional burden for
manufacturers to report include these two emissions values in their reports to EPA to be minimal.
Additionally, we believe the engineering analysis required for alternative fuel engine
converters will be straightforward. Engines that do not include SCR (i.e., gasoline-fueled
engines) have no propensity for increased N2O formation and the analysis can simply state this.
There is some greater concern for engines that rely on SCR; however, the manufacturer would
only need to show that the fueling strategy and urea dosing allows for a reasonable expectation
that N2O formation across the catalyst will not increase.
Since aftermarket converters are simply verifying that their conversion did not change
previously certified emission levels and we do not require full certification testing, engine
converters are not required to generate engine fuel maps. The total estimated burden for
aftermarket converters is about 1.5 engineering hours per model (or family).
Aftermarket converters performing fuel conversion on certified vehicles have supplied
revenue and volume information as part of their reporting under 40 CFR part 85. The top five
converters cover 80 percent of the production volume from this sector. The remaining 12
companies have an average annual revenue of about $1.1 million from an average of about 200
conversions.
To assess the cost burden for these small businesses, we assume the average small-
volume engine converter must demonstrate compliance with conversions representing three
different models (or families), resulting in an annual cost of about $240, which is 0.02 percent of
average annual revenues for the 12 smallest aftermarket converters. These 12 companies also
include a range of smaller and larger companies; however, even smaller companies would clearly
not exceed 1 percent of annual revenue.
12.8.3	Vocational Vehicle Chassis Manufacturer Economic Effects
For vocational chassis manufacturers, EPA identified 19 companies that met SBA's small
business threshold of 1,500 employees or fewer. As mentioned previously, we are adopting
provisions that will allow custom chassis manufacturers (many of whom are small businesses) to
use a simplified version of our GEM vehicle compliance tool. Part of this simplification includes
use of a default driveline, which reduces the amount of data these manufacturers will have to

-------
collect and submit. Additionally, we are allowing electric vehicle manufacturers to certify
without the use of GEM.
We did not assume the same costs for every year of the program. Instead, the first year is
expected to require more capital costs and time from employees. Subsequent years include very
few capital costs and less time. We are basing our analysis on an 8-year average cost, which
includes the hourly cost of engineers, managers, attorneys, administrative and information
technology support. We project that the average cost of compliance to be $47,000 for custom
chassis manufacturers and $13,000 for electric vehicle manufacturers.
We compared these costs to the revenue information we collected from Hoovers for the
19 small business vocational chassis manufacturers. With all of the flexibilities adopted in this
rulemaking, only two small vocational chassis manufacturers (11 percent) are projected to have
an economic impact greater than one percent and no companies are projected to have an impact
greater than three percent. Table 12-2 summarizes the small business vocational chassis results.
Table 12-2 Summary of Impacts on Small Business Vocational Chassis Manufacturers

Number/Fraction of Entities with Economic Impact of...
<1 %
1% to 3%
>3%
Number of Small Businesses
17
2
0
Fraction of Small Businesses
89%
11%
0%
12.8.4 Glider Vehicle Manufacturer Economic Effects
As described in Chapter 12.4, there are large numbers of small businesses that produce
vehicles from glider kits. The large majority of these are truck-repair facilities that occasionally
find themselves in a situation where a customer wants to install an existing engine or powertrain
into a glider kit. Under the final program, such companies that qualify as small businesses and
that sold glider vehicles in 2014 may continue to produce vehicles from glider kits up to their
historical levels over the 2010-2014 time frame, or up to 300 units, whichever is less. Almost all
these companies will therefore not be constrained by the new provisions requiring additional
glider vehicles beyond the applicable threshold to meet emission standards based on the date of
the vehicle (i.e. the glider kit) into which an engine is installed . These companies will have no
change in their business practice other than the requirement to notify EPA initially, submit an
annual report with their production volumes, and add a label to their vehicles. These costs are
much less than 1 percent of revenue even if production is limited to a single new vehicle.
The remaining assessment is for companies that produced more than 300 annual units.
These companies would be subject to emission standards and would need to install newer
engines in the glider vehicles they produce beyond the 300 cap. We would expect many
customers in these circumstances to purchase a freshly manufactured vehicle instead of opting
for a glider vehicle with compliant engines, so it is possible that they may see a drop in sales.
However, any loss in sales would only be relative to recent years, and is not likely to drop below
pre-2007 levels. Thus, it is not straightforward to determine how to quantify a cost burden for
companies in this situation; however, it is apparent that any such companies should be
characterized as having a cost burden that exceeds 3 percent of annual revenue. We are aware of

-------
one small business that produces more than 300 vehicles from glider kits. Nevertheless, this
company has previously acknowledged that they could "make a profit at 300 a year."4
There are clearly fewer than 100 companies with sufficient production volumes such that
their cost burden from the rule exceeds 1 or 3 percent of annual revenue.
12.8.5 Trailer Manufacturer Economic Effects
For trailers, EPA identified 147 companies that met SBA's small business threshold of
1,000 employees or fewer. As mentioned previously, we are limiting the non-box trailer
program to tanks, flatbeds and container chassis, and exempting all other types of non-box
trailers. As a result, 73 small business trailer manufacturers have zero burden from this
rulemaking. The economic burden for the remaining 74 small business trailer manufacturers
depends on which type of trailers they manufacture. Three companies exclusively manufacture
box trailers, 69 only manufacture non-box trailers and two manufacturer both non-box and box
trailers.
Prior to the start of the regulations, we projected that trailer manufacturers would incur
some start-up costs to prepare for compliance. We assumed trailer manufacturers would
purchase new computer systems to track sales and store compliance records, and new equipment
for emissions labeling. We also assumed box trailer manufacturers would build an additional
warehouse to store aerodynamic devices. We based this analysis on the assumption that all small
box trailer manufacturers would take advantage of the pre-approved aerodynamic data option
and would not perform any testing. We do assume a small engineering cost for engineers and
managers to review the test procedures and become familiar with the requirements so they can
appropriately evaluate available technologies. We also assume continuous costs associated with
review of the regulations and guidance documents, evaluating aerodynamic and tire
technologies, creating user manuals, calculating compliance values, generating applications and
reports for compliance, and maintaining records.
We did not assume the same costs for every year of the program. Instead, the first year is
expected to require more capital costs and time from employees. Subsequent years include very
few capital costs and less time. We are basing our analysis on an 11-year average cost (the
trailer program begins three years earlier than the other heavy-duty sectors in the Phase 2 rules),
which includes the hourly cost of engineers, managers, attorneys, administrative and IT support.
We project that the average cost of compliance to be $76,000 for trailer manufacturers that are
certifying box and non-box trailers, $67,000 for manufacturers of box trailers only, and $23,000
for non-box trailer manufacturers. We compared these costs to the revenue information we
collected from Hoovers for the 147 small business trailer manufacturers. With all of the
flexibilities adopted in this rulemaking, only 18 small trailer manufacturers (12 percent) are
projected to have an economic impact greater than one percent. Table 12-3 summarizes the
small business trailer results.5

-------
Table 12-3 Summary of Impacts on Small Business Trailer Manufacturers

Number/Fraction of Entities with Economic Impact of...
<1 %
1% to 3%
>3%
Number of Small Businesses
129
15
3
Fraction of Small Businesses
88%
10%
2%
12.9 Summary of Economic Effects
The agencies identified five general heavy-duty industries that would be potentially
affected by this rulemaking: alternative fuel engine converters, heavy-duty engine
manufacturers, vocational vehicle chassis manufacturers, glider manufacturers, and trailer
manufacturers. The agencies proposed and sought comment on the recommendations from the
Panel. The flexibilities proposed for the engine manufacturers, engine converters, vocational
vehicle manufacturers, and glider manufacturers are adopted in the final rule (with increased
flexibility in some cases) and fewer than 20 percent of the small entities in those sectors are
estimated to incur a burden greater than one percent of their annual revenue. In addition to the
flexibilities proposed for the trailer program, the agencies also reduced the number of small
entities regulated by the final rules by limiting the non-box trailer program to three distinct trailer
types. As a result, more than half of the small business trailer manufacturers have zero burden
from this rulemaking. Of the remaining small business trailer manufacturers, only 12 percent are
estimated to have an economic impact greater than one percent of their annual revenue. As a
result of these findings, EPA believes it can certify that these rules will not have a significant
economic impact on a substantial number of small entities under the RFA.

-------
References
1	Final Report of the Small Business Advocacy Review Panel on EPA's Planned Proposed Rule: Greenhouse Gas
Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles: Phase 2. Signed on
January 15, 2015. Available in docket at: EPA-HQ-OAR-2014-0827.
2	Small Business Size Standards for Manufacturing. Small Business Administration. Docket ID: SBA-2014-0011.
Available online at: https://www.regulations.gov/docket?D=SBA-2014-0011.
3	Final Report of the Small Business Advocacy Review Panel on EPA's Planned Proposed Rule: Greenhouse Gas
Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles: Phase 2. Signed on
January 15, 2015. Available in docket at: EPA-HQ-OAR-2014-0827.
4	http://www.truckinginfo.com/article/storv/2013/Q4/the-return-of-the-glider.aspx.. accessed July 16, 2016.
5	Memorandum to Docket EPA-HQ-OAR-2014-0827: "Small Business Economic Burden Calculations for Trailer
SISNOSE Analysis." July 2016.

-------
Chapter 13: Natural Gas Vehicles and Engines
13.1 Detailed Lifecycle Analysis
In this section we present our assessment of the lifecycle impacts of natural gas used by
the heavy-duty truck sector. We also present the results of an analysis by the Energy
Information Administration projecting the future use of natural gas by heavy-duty trucks.
Finally, we list a number of potential technologies which could help to reduce the methane
emissions from natural gas trucks.
This section was updated and improved in a number of ways since the draft analysis in
the proposed rulemaking. First, the estimated upstream methane emissions from the natural gas
sector were updated to the 2016 Greenhouse Gas (GHG) Inventory which estimates GHG
emissions in 2014. This is important because the GHG Inventory was revised to show much
higher upstream methane emissions for natural gas. While the GHG emissions associated with
the production of petroleum was also updated in the GHG Inventory, the GHG emissions
associated with diesel fuel was not updated in this analysis because GREET, which is the source
for diesel fuel GHG emissions, has not yet been updated with the new upstream GHG emission
estimates. Using the latest GHG Inventory with higher re-estimated upstream methane
emissions from the natural gas sector responds to comments which claim that the previous year's
GHG Inventory underestimates GHG emissions from the upstream natural gas sector.
Second, methane tailpipe emissions from 2014 and later natural gas trucks, which must
meet a 0.1 g/brake horsepower-hour methane emissions standard, was estimated for this final
rule lifecycle emissions analysis based on certification data for trucks complying with the
methane emissions standard. For the proposed rule analysis we based the estimate of methane
tailpipe emissions from natural gas heavy-duty trucks either on the methane emissions standard
or on trucks prior to the methane emissions standard because data was not yet available to
estimate what actual emissions would be under the methane emissions standard.
Third, the natural gas heavy-duty truck lifecycle emissions analysis now estimates some
additional methane emissions from natural gas heavy-duty trucks. The new methane emission
points includes refueling emissions, CNG compressor emissions and methane emissions from
LNG liquefaction plants which were not includes in the lifecycle analysis in the proposed
rulemaking. Estimating and including these additional methane emissions in our lifecycle
analysis responds to comments that our analysis was missing some methane emissions from
natural gas trucks.
13.1.1 Upstream Emissions
Upstream methane emissions, occurring in the natural gas production, natural gas
processing, transmission, storage and distribution sectors, are estimated and summarized in an
annual report "Inventory of U.S. Greenhouse Gas Emissions and Sinks" (GHG Inventory)
submitted to the United Nations Framework Convention on Climate Change (UNFCCC).1 As a
basis for estimating the lifecycle impact of natural gas use by heavy-duty trucks, we used the

-------
year 2014 methane emission estimates in the most recent GHG Inventory, published in 2016.A
The GHG Inventory also includes the quantity of carbon dioxide which is co-produced with
methane throughout the natural gas system and emitted to the atmosphere through venting,
flaring, and as fugitive emissions.
The GHG Inventory is updated annually to account for new emission sources (e.g., new
natural gas wells), updated data, emission factors and/or methodologies, and to account for
changes in emissions due to changes in policy, regulations and industry practices. The GHG
Inventory reflects emission reductions due to existing state regulations, National Emission
Standards for Hazardous Air Pollutants (NESHAP) promulgated by EPA in 1999,2 the New
Source Performance Standards (NSPS OOOO) promulgated by EPA in 2012,3 and Natural Gas
Star (a flexible, voluntary partnership that encourages oil and natural gas companies to adopt
proven, cost-effective technologies and practices that improve operational efficiency and reduce
methane emissions).4
Emission estimates in the GHG Inventory are generally bottom-up estimates which are
per-unit (compressor, pneumatic valve, etc.) emission estimates based on measured or calculated
emission rates from such emission sources.
In addition to the national-level data available through the GHG Inventory, facility-level
petroleum and natural gas systems data is also available through EPA's Greenhouse Gas
Reporting Program (GHGRP). These data represent a significant step forward in understanding
GHG emissions from this sector and EPA expects that this data will be an important tool for the
Agency and the public to analyze emissions, and understand emission trends. EPA is already
using GHGRP data to update emission estimates in the GHG inventory, and EPA plans to
continue to leverage GHGRP data to update future GHG Inventories.
Before discussing the lifecycle emissions of CNG and LNG, it is important to understand
the logistics of providing natural gas for CNG and LNG. The natural gas which comprises CNG
is expected to be off-loaded from the natural gas system where the vehicles using CNG are
refueled. This is because the natural gas used as CNG is compressed at the retail stations and
fleet facilities which fuel the CNG vehicles. To get the natural gas to the CNG retail facilities,
the natural gas must be shipped through the distribution system downstream of the natural gas
transmission system. When the natural gas is transmitted through the distribution system, the
methane emissions are higher because the methane emissions from the distribution system are
added to the rest of the upstream methane emissions.
Because LNG plants are located separate from the retail facilities, they can be located to
access the lowest cost feedstock. This means the natural gas for LNG can be sourced from the
larger natural gas transmission pipelines which are upstream of the distribution pipelines. This
provides two advantages for LNG: 1) by avoiding the natural gas distribution system, the natural
gas is priced lower, and 2) avoiding the natural gas distribution system avoids the methane
emissions which occur from the distribution system. Table 13-1 contains the 2014 methane
emissions estimate based on the GHG Inventory. About 10 percent of the natural gas consumed
A Compared to the 2015 U.S. GHG Inventory, the 2016 U.S. GHG Inventory natural gas methane emission estimates
are much higher for natural gas production, about the same level of methane emissions for natural gas processing,
and much lower from natural gas transmission, storage and distribution.

-------
in the US is sourced from Canada, and the GHG Inventory does not include the methane
emissions from the Canadian natural gas. Table 13-1 contains a second column of values which
adjusts the field production and the natural gas processing upwards by 10 percent to estimate and
account for those methane emissions.
Table 13-1 Methane Emissions from the Natural Gas System in 2014
EMISSION POINT FROM NG
FACILITIES
METHANE EMISSIONS
(GIGAGRAMS)
METHANE EMISSIONS
ADIUSTED FOR
CANADIAN NATURAL
GAS (GIGAGRAMS)
Field Production
4359
4843
NG Processing
960
1067
Transmission and Storage
1282
1282
Subtotal without Distribution
6601
7192
Distribution
444
444
Total with Distribution
7045
7636
The methane emissions attributed to the production of natural gas does not account for
the methane emissions caused when producing the natural gas associated with the production of
crude oil. According to the Energy Information Administration, natural gas produced along with
(associated with) crude oil production comprises 18 percent of the total quantity of natural gas
produced in the U.S. To estimate the methane emissions from associated natural gas wells, we
accessed the estimated methane emissions from the petroleum sector in the GHG Inventory,
which is 2694 gigagrams (kilotons) of methane in 2014. To estimate what fraction of these
methane emissions is being emitted by associated wells versus petroleum only wells, we applied
a fraction of associated wells to total crude oil wells. There are 503,873 associated wells out of a
total of 898,268 crude oil wells, or 56 percent.
EPA is taking additional steps to reduce the emissions of methane from the natural gas
and oil production facilities. On May 12, 2016, EPA finalized regulations (2016 NSPS OOOOa)
which, among other things, include methane standards for new, modified, and reconstructed oil
and gas equipment used across the oil and gas source category (before this amendment, these
rules only covered VOC, not methane directly), and require the use of reduced emissions
completions (RECs) at hydraulically fractured oil wells.B 5 In March of 2016, the Obama
Administration and the Environmental Protection Agency announced plans to regulate emissions
from existing oil and gas sources.6 7 The goal of these various actions is to achieve an
aggregated 40 to 45 percent reduction in methane emissions relative to methane emissions in
2012. The lifecycle analysis in this Chapter 13 does not take into account the 2016 NSPS, or any
future action that would address existing sources of methane emissions. As such, this analysis
B Reduced emission completions is a technology for capturing natural gas emissions during the time that the well is
being completed and the production from the well is inconsistent and includes a lot of water.

-------
likely overestimates future methane emissions from natural gas facilities for this lifecycle
analysis which attempts to model emissions in the year 2025.
The GHG Inventory also includes the quantity of carbon dioxide which is co-produced
with methane throughout the natural gas system and emitted to the atmosphere through venting,
flaring, and as fugitive emissions. Similar to how we adjusted the methane emissions to account
for Canadian-produced natural gas, we made a similar adjustment here to estimate the quantity of
carbon dioxide being emitted from Canadian natural gas wells. The quantity of carbon dioxide
being emitted from natural gas wells is summarized in Table 13-2.
Table 13-2 Carbon Dioxide Emissions from the Natural Gas System in 2014
EMISSION POINT FROM NG
FACILITIES
CARBON DIOXIDE
EMISSIONS (GIGAGRAMS)
CARBON DIOXIDE
EMISSIONS ADIUSTED
FOR CANADIAN
NATURAL GAS
(GIGAGRAMS)
Production
18,585
20,650
NG Processing
23,713
26,348
Transportation and Storage
39
39
Distribution
14
14
Total
42,351
47,050
In the GHG Inventory, EPA assessed the amount of uncertainty with its emission
estimates and provided a lower and upper bound estimate for its emission estimates. The lower
bound emission estimate is 19 percent lower than the best case estimate in Table 13-1 and the
upper bound estimate is 30 percent higher than the best case estimate.
In the 2016 Second Biannual Report of the United States of America, EPA projects that
total methane emissions will increase in the future due to increases in natural gas production.8
Table 13-3 summarizes the projected increase in US methane emissions from the Climate Action
Plan and the projected increase in natural gas production referenced from Energy Information
Administration's 2015 Annual Energy Outlook.
Table 13-3 Projected Natural Gas Production Volume and Methane Emissions (g/million BTU)
YEAR
2014
2025
Methane Emissions
Teregram C02eq.
641
674
Natural Gas Production (dry)
trillion cubic feet
25.57
30.51
As Table 13-3 shows, methane emissions from natural gas facilities are expected to
increase from 641 teregram CCheq in 2014 to 674 teregram CO2 eq. in 2025, about a 5 percent
increase. At the same time, natural gas production of dry natural gas is expected to increase

-------
from 25.6 trillion cubic feet in 2014 to 30.5 trillion cubic feet in 2025, about a 19 percent
increase. When estimating the methane emissions on the same natural gas production basis, the
methane emissions are projected to be 12 percent lower in 2025 than 2014.c
In the GHG Inventory, emissions associated with powering the units or equipment (i.e.,
compressors, pumps) used in natural gas production, processing, transmission and distribution
are aggregated with all the other fossil fuel combustion activities. Rather than attempt to
disaggregate those specific GHG emissions from the rest of the process emissions in the GHG
Inventory, we instead used the estimated emissions for these sources provided by GREET.9
Table 13-4 summarizes the process energy consumed to produce and process natural gas.
Table 13-4 Process Energy Demand by the Natural Gas System (BTU/million BTU)
FUEL
TYPE
PRODUCTION
NATURAL
GAS
PROCESSING
TRANSMISSION/
DISTRIBUTION
TOTAL
TOTAL -
INCLUDES
PROCESS
ENERGY
FOR
CANADIAN
GAS
Conv
Wells
Shale
Wells
Weighted
Average
Natural
Gas
22,016
20,955
21,307
26,123
0
47,687
52,986
Diesel
2816
2680
2725
272
0
3030
3367
Electricity
256
244
248
816
0
1067
1185
Gasoline
256
244
248
0
0
251
279
Residual
Fuel
256
244
248
0
0
251
279
Totals
25,600
24,367
24,777
27,211
0
52,286
58,096
Table 13-5 contains the factors we used to convert the GREET process energy demands
used to operate the equipment used to produce, process and distribute natural gas to carbon
dioxide emissions for those process fuels.10
c The 12% reduction figure is calculated by multiplying the methane emissions estimate in 2025 by the ratio of 2014
natural gas production over the 2025 natural gas production (674x25.6/30.5) and the resulting value is 115, which is
88% of 641, or 18% less.

-------
Table 13-5 Carbon Dioxide Emission Factors for Process Fuel Consumption
PROCESS FUEL
GC02/BTU
Natural Gas
0.0398
Diesel
0.0555
Electricity
0.1549
Gasoline
0.0535
Residual Fuel
0.0563
Table 13-6 summarizes the total estimated methane and carbon dioxide emissions emitted
by the upstream natural gas system. Two estimates are provided, one of which includes the
emissions from the distribution system representing the upstream emissions for CNG. The
second estimate summarizes the emissions excluding the emissions from the distribution system
representing the upstream emissions for LNG, since it is expected to access the natural gas from
the transmission portion of the natural gas system.
Table 13-6 Projection of Year 2025 Emissions from the Natural Gas System (grams/million BTU)

METHANE EMISSIONS
CARBON DIOXIDE
CNG Analysis (includes CH4
emissions from the distribution
system)
320
3885
LNG Analysis (does not include
CH4 emissions from the
distribution system)
305
3885
13.1.2 Downstream Emissions
The GHG Inventory does not estimate the methane emissions for natural gas once the
natural gas is diverted for use by the transportation sector, thus, we obtained information from
other sources. Natural gas can be used by vehicles either as a compressed gas (CNG) or as
liquefied natural gas (LNG). We discuss the emissions of both.
13.1.2.1 Compressed Natural Gas (CNG)
To make CNG available to trucks, the natural gas must be compressed from the pressure
that it is available from the distribution pipelines to a pressure over 3600 psi to enable filling the
truck CNG storage tanks which are either 3,000 pound per square inch gauge (psig) or 3,600
psig. We used the GHG emissions from GREET for compression for this step which reflects
national-average emissions for electricity generation for the electricity required to compress
CNG.11 We also estimated that fugitive emissions from compressors are 34 grams of methane
per million BTU of natural gas compressed. The estimate is based on an EPA report which
estimated methane emissions from reciprocal compressors at a storage facility to be 300,000
standard cubic feet of methane per year.12 This value is supported by a more recent review of
compressor emissions.13 The packing seal emissions are assumed to be emitted from a typical
sized reciprocating compressor that are used in retail CNG stations which compresses 20,000
standard cubic feet of natural gas per hour.14 We assumed that these compressors operate 24
hours per day. The GHG emissions associated with electricity generation for compressing

-------
natural gas and the fugitive emissions associated with natural gas compression for CNG are
summarized in Table 13-7.
Table 13-7 Estimated Emissions for Electricity Generated to Power CNG Compressors (g/million BTU)
FUGITIVE EMISSIONS
E
LECTRICITY GENERATION
Methane
Methane
Carbon Dioxide
Nitrous Oxide
34
6.9
3988
0.06
An important advantage that CNG has over LNG is that only a single facility, the retail
outlet, is required for distributing CNG, while LNG requires both a liquefaction plant and a retail
outlet. The simplified logistics of providing CNG also provides fewer opportunities for
emissions and leakage to the environment.
The act of refueling CNG trucks can contribute to methane emissions. When the truck
driver connects the refueling hose to the refueling port on the truck and begins refueling, there
may be refueling emissions from the station equipment and also the truck's refueling nozzle. We
estimate a quantity of refueling emissions based on the emission limits of several pieces of
equipment involved in the refueling process. As summarized in Table 13-8, United Nations
regulation number 110 Revision 3 specifies emissions limits for flexible piping, refueling fittings
and pressure relief valves.15 In deriving an emissions estimate, we assume that these various
refueling hardware devices emit half of these emissions limits over the respective hardware's
lifetime. For example, flexible fuel lines are limited to emitting 95 cubic centimeters per day per
meter of flexible fuel line of methane or natural gas per day. We assumed that 3.5 meters of
flexible piping would be required and that the emissions levels would be 95/2 or 47.5 cubic
centimeter per day per meter of piping. To estimate the emissions associated with decoupling
the refueling fittings we used emissions data from an emissions study.16 The methane emissions
associated with decoupling the CNG refueling nozzles is based on actual measurements of
decoupling CNG refueling nozzles. Table 13-8 summarizes the emissions standard from which
we estimate the quantity of methane emissions, and summarizes the resulting emission value per
million BTU of natural gas consumed.

-------
Table 13-8 Summary of CNG Refueling Emissions

Emissions
Standard
Other
Assumptions
g/MMbtu
Flexible Piping
95 cm3/meter-day
(assume 3.5
meters)
Refueling 60
gallons equivalent
over 12 minutes
0.0001
Refueling Fittings
15 cm3/hour
Refueling 60
gallons equivalent
over 12 minutes
0.0001
Pressure Relief
disk
15 cm3/hour
Refueling 60
gallons equivalent
over 12 minutes
0.0001
CNG Decoupling
Fueling Hose
Emissions
5 cm3/refiieling
event
Refueling 60
gallons
0.449
Another potential source of fugitive emissions is from small leaks in the CNG fuel
storage system. While CNG has an advantage over LNG because it is contained in a sealed
system, the very high pressure at which CNG is stored dramatically increases fugitive emissions
if a fitting pipe were to develop a leak. The level of fugitive emissions for a certain sized hole is
directly proportional to the pressure. We do not have any data on the fugitive emissions from
CNG trucks, therefore, in our lifecycle analysis, we assume that CNG fugitive emissions are zero
which likely underestimates the methane emissions from CNG trucks.
13.1.2.2 Liquefied Natural Gas (LNG)
The first step in making LNG available to trucks is the liquefaction step. As discussed
above, the liquefaction plant is likely to be constructed near natural gas transmission pipelines to
access the natural gas at the lowest price point. The liquefaction step involves the removal of
heat from the natural gas until it undergoes a phase change from a gas to a liquid at a low
pressure. Once the natural gas is liquefied, it is stored in an insulated storage tank to keep the
LNG liquefied.
LNG plants are configured depending on their ultimate capacity. Large LNG export
facilities produce 5 million metric tons, or more, per year of LNG and the economy of scale of
these large plants support the significant addition of capital to reduce their operating costs. An
LNG plant solely producing LNG for truck fuel is expected to be significantly smaller than the

-------
large LNG export facilities and so the capital invested is expected to be much lower, thus, their
operating costs would be expected to be much higher, and their energy efficiency much lower on
a percentage basis. The California Air Resources Board estimated that the liquefaction plants
used for producing truck LNG fuel are 80 percent efficient, compared to 90 percent efficient for
large LNG export facility.17 Recently, CARB estimated the lifecycle impacts of LNG using both
90 and 80 percent efficient LNG liquefaction plants (this assessment by CARB is solely for
illustrative purposes - to qualify for credit under the Low Carbon Fuel Standard (LCFS), the
actual LNG plant performance would need to be the basis for requesting credit under the Low
Carbon Fuel Standard). In our lifecycle analysis of LNG as a truck fuel, we assumed that LNG
plants are 80 percent efficient based on the earlier CARB paper along with additional review of
LNG plant types most likely to be used for providing LNG fuel for truck stops.18 19 Methane
emissions from LNG plants are estimated to be 14 grams per million BTU based on a National
Energy Technology Laboratory report.20 For our GHG analysis, we estimate the carbon dioxide
emitted when 20 percent of the natural gas is combusted to provide the energy required to liquefy
the natural gas to LNG. The upstream emissions associated with the natural gas consumed in the
liquefaction process must be accounted for and added onto the LNG produced by the plant.
These emissions are included as indirect emissions. Table 13-9 summarizes the GHG emissions
attributed to the liquefaction plant.
Table 13-9 LNG Liquefaction Plant Emissions (g/million BTU)

METHANE
CARBON DIOXIDE
Direct Emissions
14
15,175
Indirect Emissions
61
971
Total Emissions
76
16,146
To transport the LNG to the retail station, the LNG is loaded into an insulated horizontal
trailer designed specifically for transporting LNG. If the LNG in the trailer were to warm
sufficiently to cause the LNG to reach the pressure relief valve venting pressure, there would be
boil-off emissions from the trailer. However, since the LNG is super cooled, boil off events are
likely to be rare. We used a CARB estimate of boil-off emissions for LNG transportation
between the LNG plant and retail outlets.21 Table 13-10 contains the estimate of boil off
emissions and the emissions from the vehicle transporting the LNG to retail.
Table 13-10 Boil-Off Emissions Estimate for LNG Transportation to Retail (g/million BTU)

METHANE
CARBON DIOXIDE
NITROUS OXIDE
Fuel Use (Diesel Fuel)
0.45
378
0.009
Methane Boil Off
0.43
0
0
Emissions



Total
0.88
378
0.009
LNG is stored in the insulated storage tank at the retail facility. Heat gain in the storage
tank could eventually lead to boil-off emissions. Service stations with little LNG demand are at
a higher risk of boil-off emissions compared to service stations which have a significant
throughput volume. LNG stations could be configured to avoid boil-off events to the
atmosphere, such as venting to a co-located CNG facility, or venting to a nearby natural gas
pipeline re-liquefying the boil off vapor and flaring the boil off gas.22 We used a GREET

-------
emission estimate to provide an estimate of the boil-off emissions from LNG retail facilities.23
Table 13-11 summarizes the estimated boil off emissions for LNG retail facilities.
Table 13-11 Boil-Off Emissions Estimate for LNG Retail Facilities (g/million BTU)
LNG RETAIL BOIL-
35
OFF EMISSIONS

The act of refueling LNG trucks can contribute to methane emissions. We used the same
method described above for estimating the emissions from refueling CNG trucks to estimate
refueling emissions from LNG trucks. Table 13-12 below summarizes our estimate for the
emissions from LNG trucks.
13-12 Summary of LNG Refueling Emissions

Emissions
Standard
Other
Assumptions
g/MMbtu
Flexible Piping
95 cm3/meter-day
(assume 3.5
meters)
Refueling 60
gallons equivalent
over 12 minutes
0.0001
Refueling Fittings
15 cm3/hour
Refueling 60
gallons equivalent
over 12 minutes
0.0001
Pressure Relief
Valves
15 cm3/hour
Refueling 60
gallons equivalent
over 12 minutes
0.0001
LNG Decoupling
Fueling Hose
2.4 cm3/refiieling
event
Refueling 60
gallons
0.435
The total well to tank emissions for CNG and LNG are summarized in Table 13-14.
These emissions represent the total of upstream and downstream emissions which includes
delivering the fuel to the truck fuel storage tank.
Table 13-13 Total Well to Tank Emissions Estimate for CNG and LNG (g/million BTU)

METHANE
CARBON DIOXIDE
NITROUS OXIDE
CNG
361
7598
0.06
LNG
432
20,409
0.009

-------
13.1.3 Vehicle Emissions
13.1.3.1 Vehicle Configurations
There are several different ways that diesel heavy duty engines can be configured to use
natural gas as a fuel. The first is a spark ignition natural gas (SING), Otto cycle SING heavy
duty engine burns the fuel stoichiometrically and uses a three-way catalyst. Some also add an
oxidation catalyst to provide the greatest emissions reduction. Stoichiometric combustion is
used in most light-duty SING engines and is used in heavy-duty service as well, but is
particularly popular for natural gas trucks. Problems with thermal stress and low power density
have favored the use of the lean-burn combustion system in some heavy duty engine
applications. The use of cooled EGR provides further potential to increase the engine output
and, at the same time, decreases NOx emissions. In this case the engine compression ratio is
reduced similar to that of a gasoline engine, about 12 to 1 or more, and thus its thermal efficiency
is lower than a diesel-like engine by about 10-15 percent, depending on the driver.
The second is a direct injection natural gas (DING), diesel cycle. The DING engine uses
a small quantity of diesel fuel (pilot injection) or a glow plug as ignition sources. As the
injection system for the diesel fuel does not have the capability of greater injection quantities,
this option has no dual-fuel properties, but allows for about 95 percent of the fuel to be provided
by natural gas. Additionally, the optimization of the pilot injection can be made to achieve lower
emissions. An advanced high pressure direct injection (HPDI) fuel system combining the
injection of both diesel fuel and natural gas can be used for lean burn combustion. This enables
the engine to maintain the efficiency advantage of a compression ignition engine while running
mainly CNG/LNG.
The third is a mixed-fuel natural gas (MFNG), diesel cycle. In a mixed-fuel engine,
natural gas is mixed with intake air before induction to the cylinder and diesel fuel is used as
ignition source. Mixed-fuel vehicle/engine means any vehicle/engine engineered and designed
to be operated on the original fuel(s), or a mixture of two or more fuels that are combusted
together. Mixed-fuel system means that a diesel engine works with two types of fuels together.
In fact the engine is a diesel thermodynamic cycle and the energy is given by the diesel and the
natural gas fuel. In mixed-fuel conversion the original engine is not modified in any way, a
conversion system is installed in order to permit the engine to run on both fuels. The conversion
of the engine is totally reversible, in fact it is possible to choose the mode how to run the engine
(diesel / mixed-fuel). When the engine runs in diesel mode, the engine runs in the same way as
per the original configuration. Engine results showed that the efficiency of the engine could
decrease by about 2-5 percent in mixed-fuel mode compared to diesel mode and that the diesel
replacement was approximately 40-60 percent efficient.
Each of these natural gas engine types has its merits. The SING engine is less costly, but
is less fuel efficient and because of the lower compression ratio it has less torque than the two
diesel cycle engines. The DING engine is likely the most expensive because of the special
natural gas/diesel fuel injection system and large required amount of natural gas (LNG or CNG)
storage since the truck must run on natural gas. However, because the truck can run almost
completely on natural gas, the DING engine has the potential to more quickly pay down the
higher investment cost of the natural gas truck. The MFNG engine provides the truck owner the

-------
flexibility to operate on natural gas or diesel fuel, but at the expense of a slower natural gas
investment pay down rate because it can operate at most 50 percent of the time on natural gas.
An important advantage of LNG is the increased energy density compared to CNG. At
present, CNG stored at its maximum storage pressure is only 25 percent of the energy density of
diesel fuel, while LNG contains about 60 percent of the energy density of diesel fuel. Because of
its higher energy density, LNG is favored over CNG for long-haul trucking.
An adsorbent for natural gas (ANG) material technology0 called metal organic
framework (MOF) for storing CNG has been invented and is being tested for large scale use.
The technology involves filling the CNG tank with a specially designed substance which looks
similar to a pelletized catalyst. The substance establishes a matrix which causes the methane
molecules in natural gas to become better organized and store the same quantity of natural gas in
a smaller volume at the same pressure, or store the same density of natural gas at a lower
pressure. This MOF could improve the energy density of CNG which would make it a better
candidate for natural gas storage for long range combination trucks, while avoiding the boil-off
events that are a risk with using LNG.
13.1.3.2 Tailpipe Emissions
When assessing the methane emissions from both CNG and LNG trucks, it is important
to separate those trucks built or converted before 2014 to those built or converted in 2014 and
later. The trucks built before 2014 are only required to meet a nonmethane hydrocarbon
(NMHC) standard, which means that the methane emissions from these trucks are unregulated.
Our certification data shows that the methane tailpipe emissions from these trucks/buses ranges
from 2-5 g/bhp-hr for both spark ignition (gasoline type) and compression ignition (diesel type)
engines.
For 2014 and later OEM compression ignition natural gas trucks or natural gas
conversions of 2014 and later diesel trucks, the trucks must meet a 0.1 g/bhp-hr methane
emission standard in the case of a larger truck engine tested with an engine dynamometer, and a
0.05 g/mile methane emission standard in the case of smaller trucks tested on a chassis
dynamometer.E For spark ignition (gasoline style) engines, the standards take effect in 2016.24
The natural gas truck manufacturers are allowed to offset methane emissions over the standard
by converting the methane emission exceedances into CO2 equivalent emissions and using CO2
credits. For the initial natural gas engine certifications that EPA has received for 2014, 2015 and
2016 model years, the truck manufactures chose to continue to emit high levels of methane
(ranges from 0.7 to 2 g/bhp-hr) and use carbon dioxide credits to offset those emissions. We
don't know if this practice of using CO2 credits to offset high methane emissions will continue in
the future; however, for evaluating the lifecycle impacts of natural gas heavy-duty vehicles, the
2014 and later natural gas heavy-duty trucks may in fact have an emissions profile in-between
the pre-2014 trucks and the 2014 and later trucks. Our emissions analysis assumes that these
D Menon, V.C., Komarneni, S. 1998 "Porous Adsorbents for Vehicular Natural Gas Storage: A Review," Journal of
Porous Materials 5, 43-58 (1998); Burchell, T "Carbon Fiber Composite Adsorbent Media for Low Pressure Natural
Gas Storage" Oak Ridge National Laboratory.
E An exception is that small volume, heavy-duty natural gas truck manufacturers are exempt from EPA's GHG
regulations.

-------
trucks are emitting 1 gram per brake horsepower-hour methane emissions. In cases when these
trucks experience an increase in emissions due to deterioration or malfunction of the engines,
fuel supplies or associated emission control devices on these trucks, the methane emissions could
be higher than estimated. Table 13-14 summarizes the emission standards and the estimated
methane emissions from heavy-duty trucks assumed in the analysis.
Table 13-14 Methane Emission Standards and Estimated Emissions from Heavy-Duty Trucks


PRE-2014
2014 AND LATER
Methane Standard

None
0.1 g/bhp-hr
Estimated Emissions
g/bhp-hr
2-5
1

g/million BTU
214-534
107
13.1.3.3 Boil-off, Venting and other Fugitive Emissions
Truck drivers requiring LNG fuel drive up to an LNG retail outlet or fleet refueling
facility and fill up with LNG fuel. Prior to refueling it may be advantageous or necessary, due to
high pressure in the truck's LNG tank, to reduce the pressure in the truck's LNG tank to speed
up the refueling process. In some cases the retail station is equipped with another hose and
associated piping to vent the excess gas to the retail stations' storage tank, or perhaps to a natural
gas pipeline. However, for those retail outlets without such vent lines to the storage tank, the
truck driver may simply vent the truck's storage tank to atmosphere. As part of a sensitivity
analysis for our lifecycle analysis, we estimate the emissions for venting an LNG tank prior to
refueling. A major issue with respect to GHG emissions associated with LNG trucks is boil-off
emissions from the trucks themselves. When the liquefied natural gas is pumped into the truck
LNG tanks, it is "supercooled," meaning that the temperature of the LNG is well below the boil-
off pressure and temperature. A typical refueling temperature of LNG is -190F, which
corresponds to 164 pounds per square inch absolute, or 149 pounds per square inch gauge. If the
truck is driven extensively the drawdown of liquid level will cause some of the fuel to boil off
and thus cool the rest of the liquid in the LNG storage tank. It is possible that the fuel would
maintain its supercooled temperature, or possibly even cool further below its supercooled
temperature until the LNG is completely consumed.
If the truck is not driven or is driven very little, the very low temperature LNG warms
through ambient temperature gradient through the tank wall causing the temperature and pressure
of the LNG to rise. When the pressure reaches a maximum of 230 psi ta safety release valve on
the LNG storage tank releases methane gas directly to the atmosphere until the pressure drops to
the reset pressure of the safety release valve. There are two industry standards used to design
tanks to reduce the temperature increase, one for a 3 day hold timeF and one for a 5 day hold
time.G Hold time is the minimum time elapsed between when the truck's LNG tank is refueled
and when it begins to vent.
F National Fire Protection Association 52, Compressed Natural Gas (CNG) Vehicular Fuel System Code, 2002
Edition.
G SAE International (2008) SAE J2343: Recommended Practice for LNG Medium and Heavy-Duty Powered
Vehicles. Warrendale, Pennsylvania.

-------
If there is a boil-off event, a large amount of methane would be released. If aware of the
impending boil-off such as when the truck is being maintained, the truck driver could hook up
the LNG tank to a hose which would vent the natural gas emissions to a CNG system which
would reuse the boil-off natural gas as CNG, or combust the LNG to carbon dioxide, or vent the
natural gas emission to a natural gas pipeline. Otherwise the boil-off emission would simply
vent to the atmosphere.
When an LNG fuel tank venting (refueling venting or boil-off) incident occurs, there are
two separate processes which occur that contribute to methane emissions during the venting.
The most obvious process is the pressure drop, from 230 to 170 psi, in the gaseous space above
the liquid. The volume of gas vented is proportional to the reduction in absolute pressure in the
tank. Since the drop in absolute pressure is 244 to 184 psi (14.7 psi is added to the 230 and 170
psi gauge pressure), about 25 percent of the gas in the tank is vented (184 psi is 25 percent of the
way from 244 psi to zero pressure). The second process is the vaporization of liquid during the
pressure reduction in the LNG tank. The boiling point of any liquid decreases as the pressure
decreases. Thus, when the LNG undergoes the pressure reduction during a venting/boil-off, the
boiling point of the methane decreases and to balance the system, some of the liquid methane
must boil off to cause the liquid to be cooled. The quantity of liquid methane which must boil
off from the liquid is calculated from methane's heat of vaporization over the boiling point
temperature change, which drops from -178 F to -189 F as the pressure drops from 230 to 170
psi.
The amount of natural gas which boils off during a venting event varies based on the
quantity of liquid in the LNG storage tank. The greatest amount of natural gas which is lost
during a venting/boil off event occurs when the tank is closest to being full. For a 200 gallon
tank system, each boil off event has the potential to release on the order of 3-9 gallons or 5,300 -
15,800 grams of CH4 which translates to 132 - 400K grams of CO2-equivalent emissions,
assuming methane has global warming potential (GWP) of 25 over a 100 year lifetime.25 If the
vehicle continues to sit after boil-off events begin to occur with boil-off events each day and up
to several boil-offs per day, as much as million grams of CO2-equivalent emissions may be
emitted over the twenty or so days at which point the vehicle LNG tank would be completely
empty.
Table 13-15 summarizes the starting and ending conditions and the loss from the tank for
venting incidents (200 gallon LNG tank decreases in pressure from 230 to 170 psi) when the
LNG tank is 90 percent, 50 percent and 10 percent full. A refueling venting event is more likely
to occur when the tank is mostly empty, so the 50 and 10 percent cases are the most likely cases
to consider.

-------
Table 13-15 Estimated Quantity of Boil-Off from a 200 Gallon LNG Fuel Tank for a Single Boil-Off Event

PERCENT FULL
PERCENT FULL
LIQUID LOSS
TOTAL MASS

(INITIAL)
(FINAL)
(GALS)
LOSS (LBS)

90
83.2
13.6
38.7
Boil-off Scenarios
50
46.2
7.6
24.8

10
9.3
1.5
11.0
Table 13-15 shows that if a truck had 200 gallon of LNG storage capacity, the estimated
quantity of liquid boil-off volume would range from 2 to 14 liquid gallons of LNG depending on
the fill level of the LNG tank. When the quantity of LNG gas loss is included, the total loss
ranges from 11 to 39 lbs.
The quantity of LNG tank boil-off or venting per distance driven by the truck depends on
the frequency of boil-off or venting incidents. As described above, a truck's driving profile
plays a key role in determining the boil-off risk from LNG trucks. Fleets which purchase LNG
trucks do so with the intent of driving the LNG truck extensively to pay off the much higher
purchase price of the LNG truck. For this reason, there are likely to be few boil-off incidents,
except for cases when the truck is forced out of its routine. Examples of when the truck might be
sidelined include times when the truck is being maintained, the immediate period after the truck
is involved in an accident, or perhaps when the owning company experiences a loss of workload
or files for bankruptcy. We have no data which would allow us to estimate the frequency when
these sorts of incidents would occur, and even if we did, we still could not estimate the frequency
of boil-offs that occur in these cases.
As the truck ages, it likely would be sold by the company which originally purchased it to
avoid having to deal with the increased maintenance that occurs with older trucks. Figure 13-1
shows the estimated vehicle miles traveled by class 8 trucks as they age (the data is from the
MOVES Model.
140,000
120,000
100,000
CD
CD
^ 80,000

Q 40,000
l/>	'
_0J
1 20,000
0
0	5	10	15	20
Truck Age (Years)
Figure 13-1 Vehicle Miles Traveled by Combination Trucks in 2014

-------
Figure 13-1 shows that the mileage driven by combination trucks decreases as they age.
By the time that a combination truck is about 17 years old, it is driven about half the number of
miles per year as a new truck. It would seem that the risk of boil-off incidents increases with
these older trucks.
Venting incidents during refueling can occur at any time, and there is an incentive to do
so when it is time to refuel. The decision to vent an LNG tank in most cases is solely up to the
truck driver who is often under pressure to complete his work in less time to maximize profits.
There is a lot of uncertainty in estimating the quantity of boil-off and venting from an
LNG truck. To reflect this uncertainty, we assume two different boil-off/venting emission
estimates. The low estimate assumes that 35 grams per million BTU of fuel consumed is
emitted, which is from GREET.26 The high estimate assumes a boil-off event and a venting
event each time the truck is refueled and before that tank full of LNG is used up, and this
quantity is estimated to be 835 g per million BTU of fuel consumed.
The crankcase of these engines receives leakage from the combustion chamber across the
piston rings, which can contain methane. The crankcase of the spark ignition engines is
normally vented into the intake of the engines; thus, any methane emissions from the crankcase
which is not combusted in the engine would be accounted for in both the engine-out and tailpipe
emissions. For compression ignition engines, however, the crankcase emissions are typically
vented into the exhaust pipe downstream of the aftertreatment devices, although they are
accounted for in addition to the engine-out emissions during certification. Engine-out emissions
are subjected to deterioration factors based on well-established procedure, which may make
estimating engine-out emissions more robust than deterioration factors for vented crankcase
emissions. Moreover, deterioration of crankcase emissions may be more variable as the engines
accumulate more miles. Thus, sealed crankcases would achieve more robust control of methane
emissions.
Another potential source of methane emissions from CNG and LNG trucks is fugitive
emissions in the form of leaks from the fuel piping to the engine. Thus, either while parked or
operated, the vehicle fuel and engine systems could leak methane to the environment. We do not
have, nor did we attempt to estimate this type of methane fugitive emissions from CNG or LNG
trucks.
Table 13-16 summarize the estimated tailpipe emissions for CNG trucks, and Table
13-17 summarizes the estimated tailpipe and boil-off and venting emissions for LNG trucks.
Table 13-16 Estimated Tailpipe Emissions for CNG Trucks (g/MMbtu)


METHANE
CARBON
DIOXIDE
NITROUS
OXIDE
2014 and Later
Direct
107
60,702
2
Indirect
2

0
Total
109
60,702
2

-------
Table 13-17 Estimated Tailpipe and Boil-Off Emissions for LNG Trucks (g/MMbtu)


METHANE
CARBON
DIOXIDE
NITROUS
OXIDE
2014 and Later
assuming low
Venting and Boil-
Off Emissions
Direct
141.8
60,702
2
Indirect
2.7

0
Total
144.5
60,702
2
2014 and Later
Assuming High
Venting and Boil-
Off Emissions
Direct
942
60,702
2
Indirect
18.2

0
Total
960
60,702
2
13.1.3.4 Thermal Efficiency
While not an emission source per se, the thermal efficiency of the natural gas engine also
plays a role in the lifecycle emissions of the truck. Thermal efficiency is defined by the amount
of energy that is obtained to propel the truck compared to the energy consumed by the engine. If
a fuel-engine is less thermally efficient, then it consumes more fuel, or more BTUs, to travel the
same distance, thus emitting more carbon dioxide per distance traveled, or work performed.
We estimate that SING engines can be as much as 15 percent less efficient than
compressed ignition engines which operate on diesel fuel. Conversely, DING and MFNG
engines which operate at a higher compression ratio, are estimated to be 5 percent less energy
efficient compared to a diesel engine. In our lifecycle analysis, we provide two different
sensitivities for natural gas vehicles assuming that they are 5 percent (thermal high) and 15
percent (thermal low) less efficient.
13.1.4 Results of Lifecycle Analysis
To estimate the lifecycle impact of natural gas used by heavy-duty trucks, we totaled the
carbon dioxide, methane and the nitrous oxide emissions for the upstream and downstream
portions of the natural gas system. The methane and nitrous oxide emissions are converted to
carbon dioxide-equivalent emissions using global warming potentials ((GWPs); these are a
measure of the relative contribution of global warming of emissions of a given gas in comparison
to that of carbon dioxide over a given time period). The GWPs EPA is currently using is from
the AR4 (2007) IPCC report for 100 year timeframe, which is 25 and 298 for methane and N2O,
respectively.
To establish the impacts of natural gas use in the heavy-duty fleet, it was necessary to
compare the lifecycle impacts of natural gas against its replacement, which is a diesel fueled
heavy-duty truck. The lifecycle greenhouse gas emissions of a diesel fuel truck is from the 2015
version of the GREET lifecycle model for the current production and use of diesel fuel.27 We
used this GREET diesel fuel lifecycle estimate for the baseline for comparison with the natural
gas lifecycle assessment. The GHG Inventory that was updated in 2016 shows much higher
methane emissions from crude oil production wells in the U.S. However, the recently finalized
methane emission regulations requires that oil wells utilize reduced emission completion

-------
technology to reduce methane emissions from oil wells. Thus, while methane emissions are
likely higher than shown by the GREET model in 2015, it is unclear what the methane emissions
will be in 2025 which is the analysis year for this lifecycle analysis. We use the 2015 GREET
lifecycle values for our lifecycle analysis in 2025. Table 13-18 summarizes the lifecycle
emissions for diesel fuel estimated by GREET.
Table 13-18 Estimated Diesel Fuel Lifecycle Greenhouse Gas Emissions (g/million BTU)

CARBON
DIOXIDE
METHANE
NITROUS
OXIDE
TOTALS
C02EQ
Well to Tank
13,792
81
0.27
15,896a
Tank to Wheels
78,993
29
0.18
79,772
Well to Wheels
92,785
110
0.45
95,668
Note:
a The totals are calculated using 25 and 298 for the GWPs for methane and nitrous oxide, respectively.
The National Energy Technology Laboratory (NETL) has also estimated the lifecycle
impact of diesel trucks and recently updated its previous analysis that was conducted for a diesel
fuel truck in 2005 to the year 2014.28 The NETL lifecycle analysis shows much higher well to
tank emissions than GREET, much lower tank to wheels emissions than GREET, but overall
somewhat lower GHG emission than GREET. In the discussion below about the relative
lifecycle analysis of natural gas versus diesel, we discuss the impact if the NETL diesel fuel
truck lifecycle analysis was used instead of GREET.
To illustrate the relative full lifecycle impact of natural gas-fueled heavy-duty vehicles
versus diesel fueled heavy-duty vehicles, we assessed two different scenarios. The first is a
conversion of a diesel engine to use CNG. Of the tens of thousands of heavy-duty natural gas
trucks currently in use, over 90 percent are of this type. While these can be conversions of older
trucks, we assume that they would still be subjected to the 0.1 gram per brake horsepower- hour
methane standard. However, based on certification data, these trucks emit much higher methane
emissions than the methane standard allows and we assume that this truck emits 1.0 gram of
methane per brake horsepower-hour. We provide two estimates for the lower thermal
efficiencies of CNG and LNG trucks. One assumes that the truck is 5 percent less thermally
efficient (thermal high) and the second assumes that the truck is 15 percent less thermally
efficient (thermal low - 10 percent less efficient than the 5 percent less thermally efficient case).
The second scenario is a combination truck fueled on LNG which is assumed to be in
compliance with the 2014 methane standard. Because it is high mileage truck, the most realistic
assumption is that the truck must use LNG as a fuel to provide the necessary range for the
dedicated natural gas engine. We make two different assumptions with respect to refueling and
boil off emissions. In the natural gas average case, we assume a modest quantity of refueling
and boil-off methane emissions estimated by GREET. The second boil-off emission estimate is
based on venting the LNG storage tank to the atmosphere each time the driver refills his tank, or
one LNG boil-off event between each time the driver must refuel his tank. As we discussed in
the discussion about refueling and truck boil-off emissions, we don't expect this to be a common
practice for newer trucks that are operated regularly. However, as the use of these trucks
decreases as they age and are sold into the secondary market, the risk for refueling and boil-off
emission events increases - this estimate provides a simple sensitivity emission estimate.

-------
The estimated lifecycle emissions of CNG and LNG trucks, assuming projected upstream
emissions in 2025, is summarized in Table 13-19.
Table 13-19 Full Lifecycle Analysis of a Natural Gas Truck (g/million BTU)
TRUCK
EMISSION
CARBON
METHANE
NITROUS
TOTAL
THERMAL
TOTALS
TYPE
CATEGORY
DIOXIDE

OXIDE
CO2 EQ.a
EFFICIENCY
5% AND 15%
C02EQ.a
INCLUDING
THERMAL
EFFICIENCY
IMPACT
CO2EQ.a
2014 or
Well to Tank
7598
361
0.06
16,643
832
17,475
later





2496
19,139
CNG
Tank to
60,702
109
2.
64,022
3035
67,057
Truck
Wheels




9105
73,127

Well to
68,299
470
2.06
80,664
3867
84,531

Wheels




11,602
92,266
2014 or
Well to Tank
20,409
432
0.009
31,214
1561
32,775
Later





4682
35,890
LNG
Tank to
60,702
145
2
64,911
3035
67,946
Truck
Wheels




9105
74,749
Avg.
Well to
81,111
574
2.01
96,057
4596
100,652
Boil-Off
Wheels




13,787
109,844
2014 or
Well to Tank
20,409
432
0.009
31,214
1561
32,775
Later





4682
35,890
LNG
Tank to
60,702
960
2
85,303
3035
88,388
Truck
Wheels




9105
94,408
High
Well to
81,111
1392
2.01
116,517
4913
121,430
Boil-Off
Wheels




14,739
131,256
Note:
a The CCheq totals are calculated using 25 and 298 for the GWPs for methane and nitrous oxide, respectively.
The CNG and LNG lifecycle assessment relative to a diesel truck lifecycle analysis is
shown in Figurel3-2. Another comparison made in Figure 13.2 is the relative tailpipe-only
emissions for diesel and natural gas trucks. The quantity of carbon dioxide, methane and nitrous
oxide emissions from a diesel truck is from GREET. The carbon dioxide emissions from natural
gas-fueled truck is calculated and is based on the carbon-hydrogen content of methane. The
methane emissions from a natural gas-fueled truck is based on natural gas truck certification data
(does not include any methane emissions from the natural gas storage tanks onboard the truck
nor other fugitive emissions).

-------
CO
E
„E
cr
ai
fM
O
u
m
140,000
120,000
100,000
80,000
60,000
40,000
20,000
0
mm
M
fTTTT
I I
Diesel CNG Diesel
Tailpipe Tailpipe
I	
CNG 2014+ 2014+
2014+ LNG Avg LNG High
BO BO
~ Thermal Low
¦	Thermal High
¦	CH4:&N20
EC02
y
Full Lifecycle Analyses
Figure 13-2 Tailpipe Emissions Comparison and Full Lifecycle Analysis of Diesel, CNG and LNG Trucks
(Projected Upstream Methane Emissions in 2025, Methane GWP of 25)
In the first two bars of Figure 13-2, it shows that based solely on tailpipe emissions (with
thermal efficiency adjustments and assuming 1 gram per brake horsepower-hour methane
emissions at the truck), natural gas trucks are estimated to emit about 10 percent less GHG
emissions than diesel engines if the engine is only 5 percent less efficient than the diesel engine,
and about the same GHG emissions if the engine is 15 percent less efficient than the diesel
engine. The three full lifecycle analyses represented by the right three bars in the figure shows
that post-2014 CNG trucks are estimated to emit about 12 percent less GHG emissions as diesel
trucks if the CNG trucks are 5 percent less efficient, although if their thermal efficiency is 15
percent less efficient, their GHG advantage would decrease to about 5 percent.
Figure 13-2 shows that LNG trucks which are only 5 percent less efficient and which
emit an average amount of boil-off emissions, emit about 3 percent more GHG emissions than
diesel trucks when we assume an average of refueling and boil-off emissions. Conversely, if the
LNG trucks are 15 percent less efficient, then LNG trucks emit about 13 percent more GHG
emissions than diesel trucks. In the case of the LNG trucks which emit a high amount of boil-off
emissions, the LNG trucks emit 25 percent and 34 percent more GHG emissions than diesels for
the 5 percent and 15 percent less efficient natural gas engines, respectively. In comparing CNG
to LNG, the LNG trucks appear higher emitting than CNG trucks mostly because of the low
thermal efficiency of small liquefaction facilities. If the LNG plant were to be 95 percent
efficient instead of the 80 percent efficiency we assume, the difference between the average boil-
off emitting LNG trucks and CNG trucks disappears. The 2014 lifecycle analysis of diesel
trucks by NETL shows diesel trucks emitting about 3 percent lower GHG emissions (CO2 eq.)
than the diesel GHG emissions we used from GREET, thus, our analysis would show that natural

-------
gas fueled trucks would be about 3 percent higher emitting in GHG emissions if we used the
NETL as the basis for diesel fueled trucks
It is important to point out the uncertainties associated with the lifecycle estimates
provided in Figure 13-2. As discussed above, there is uncertainty in both the upstream and
downstream methane emission estimates for natural gas facilities and equipment, and the trucks
that consume natural gas. In the GHG Inventory, EPA estimates a range of natural gas emissions
from the upstream natural gas production sector. The range varies from -19 percent to +30
percent relative to the principal estimate. To illustrate the impact the range has on the relative
life cycle impacts of natural gas versus diesel trucks, Figure 13-3 shows the impact on the
relative lifecycle emissions for CNG trucks when the low and high methane emissions are
compared to the best estimate case we used in the above analyses for a CNG truck emitting 1
gram of methane per brake horsepower-hour.
CO
E
„E
cr
a>
fM
O
u
m
120,000
100,000
80,000
60,000
40,000
20,000
0
m mm
[TTTTl
~ Thermal Low
\ ¦ Thermal High
¦ CH4:&N20
E3C02
Diesel Nat Gas Diesel Nat Gas Nat Gas Nat Gas
Tailpipe Tailpipe ^	Low Avg High
y
Full Lifecycle Analyses
Figure 13-3 Tailpipe Emissions Comparison and Full Lifecycle Analysis of a Diesel and CNG Truck - Low,
Average and High Upstream Natural Gas Methane Emissions
(Projected Upstream Methane Emissions in 2025, Methane GWP of 25)
Figure 13-4 shows that higher and lower upstream emissions, based on the uncertainty
factors provided in the GHG Inventory, does impact the relative GHG lifecycle impact of CNG
trucks, but the effect is quite modest.
The GWPs used to assess the relative climate impacts of methane and nitrous oxide can
also effect the relative lifecycle impacts natural gas trucks compared to diesel trucks. The GWPs
of methane and nitrous oxide vary based on the timescale assumed. To illustrate this point, we
added two more sets of figures as sensitivities for comparing the lifecycle impacts of CNG and
LNG natural gas trucks to diesel trucks if the greenhouse gas emissions are evaluated over a
different lifetime. The GWPs that we use are the two alterative GWPs reported by IPCC in its
4th Assessment Report evaluated at 20 year and 500 year GHG lifetimes. Table 13-20

-------
summarizes the GWPs at the different lifetimes along with the GWPs used in the primary
analysis summarized above.
Table 13-20 Summary of GWPs

PRIMARY
ANALYSIS
SENSITIVITY ANALYSES

100 Year
20 Year
500 Year
Methane (CH4)
25
72
7.6
Nitrous Oxide
(N2O)
298
289
153
It is important to point out that while there are fairly significant differences in methane
emissions between the various natural gas cases being studied and compared to diesel trucks, the
nitrous oxide emissions vary very little across all the cases. Therefore, when comparing the
relative lifecycle impacts using different GWPs, the impact on relative lifecycle emissions is
almost exclusively due to changes in the methane GWP. Figures 13-4 and 13-5 show the relative
lifecycle effects of natural gas trucks compared to diesel trucks when the GWPs used are based
on 20 year and 500 year lifetimes.
CO
E
„E
cr
a>
fM
O
u
m
200,000
180,000
160,000
140,000
120,000
100,000
80,000
60,000
40,000
20,000
0
mm
ill
+<
+<
+<
+<
+<
~ Thermal Low
¦	Thermal High
¦	CH4:&N20
BC02
Diesel CNG
Tailpipe Tailpipe
Diesel CNG 2014+ 2014+
2014+ LNG Avg LNG High
i	BO BO
y
Full Lifecycle Analyses
Figure 13-4 Comparison of Tailpipe Emissions and Full Lifecycle Analyses
of Diesel, CNG and LNG Trucks (Projected Upstream Methane Emissions in 2025, Methane GWP of
72)

-------
CO
E
„E
cr
ai
fM
O
u
m
100,000
80,000
60,000
40,000
20,000
0
m H
~ Thermal Low
¦	Thermal High
¦	CH4:&N20
EC02
	
Diesel CNG Diesel CNG	2014+ 2014+
Tailpipe Tailpipe 2014+	LNG Avg LNG High
|		BO BO
y
Full Lifecycle Analyses
Figure 13-5 Comparison of Tailpipe Emissions and Full Lifecycle Analyses of Diesel, CNG and LNG Trucks
(Projected Upstream Methane Emissions in 2025, Methane GWP of 7.6)
Figures 13-4 and 13-5 show that when evaluated over a shorter timescale, the higher
GWP for methane increases the relative lifecycle impact of natural gas trucks compared to diesel
trucks. Conversely, when evaluated over a longer timescale, the lower GWP for methane
decreases the relative lifecycle impact of natural gas trucks compared to diesel trucks.
We compared our lifecycle emission estimates for natural gas, relative to diesel fuel, with
the estimates provided by the California Air Resources Board (CARB) for its Low Carbon Fuel
Standard (LCFS). For our emissions estimate used in the comparison we used the carbon
dioxide-equivalent (CCheq) emissions estimated for 2014 and later engines, which must comply
with a methane tailpipe emissions standard, and assumed that the engine was 5 percent less
thermally efficient than a comparable diesel engine. Both analyses used GWPs based on 100
year timescale (i.e., a GWP of 25 for methane and 298 for nitrous oxide). For the CARB
emissions estimates, we used the estimates made for what CARB terms "illustrative purposes"
using the values printed in the April 3, 2015 workshop handouts.29 CARB estimates that CNG
engines emit 86 percent of the CCheq emissions as a diesel truck using the EER-adjusted values
which reflect all percent lower energy efficiency than a diesel truck. When we adjust our
analysis to reflect a truck which is 11 percent less efficient than a diesel truck, our analysis
estimates that CNG engines emit 89 percent of the CCheq emissions as a diesel truck. An
important reason why CARB estimates lower CNG truck GHG emissions than our analysis is
that a much larger portion of the electricity used to compress natural gas is renewable in
California than the rest of the country. Also, our analysis accounts for the recent more accurate
GHG Inventory estimates which show higher natural gas upstream emissions. Using the same
assumption that natural gas trucks are 11 percent less efficient CARB estimates LNG engines
emit about 94 percent of the CCheq emissions. After adjusting our analysis to also assume that
trucks are 11 percent less efficient, our natural gas lifecycle analysis estimates that LNG trucks

-------
emit 106 percent of the CCheq emissions as a diesel truck. The reasons why LNG truck
emission estimates are so much higher than CARB's is because we assume that LNG
liquefaction plants are only 80 percent efficient as opposed to CARB' assumption that LNG
liquefaction plants are 90 percent efficient. Also, CARB assumes no boil-off or venting
emissions from LNG trucks and for this comparison, we used our more modest boil-off and
venting assumption, as described above. Overall, our estimates seem to be consistent to those
estimated by CARB when we account for the different assumptions used in the respective
analyses. Both our heavy-duty truck lifecycle analyses are expected to improve for natural gas
compared to diesel fuel as we consider the effects of the 2016 NSPS and later methane emissions
standards.
13.2 Projecting Natural Gas use in HD Trucks
We reviewed several information sources and projections to estimate how much natural
gas is currently being used and is projected to be used by heavy-duty trucks.
In the Energy Information Administration's Annual Energy Outlook (AEO) 2015, EIA
shows natural gas use comprising about only 0.35 percent of total heavy duty fuel consumption
in 2013, and natural gas use by Class 8 trucks is about 0.2 percent.30 In 2014, AEO 2014 shows
natural gas comprising about 0.4 percent of total heavy-duty fuel demand and about the same for
Class 8 heavy-duty truck demand.
An estimate by the Natural Gas Vehicle for America (NGVA) of the number of natural
gas trucks operating today supports this level of fuel demand made by EIA. In a meeting with
NGVA, NGVA presented their estimate that 62,000 heavy-duty trucks are fueled by natural gas
in 2014. The MOVES database estimates that there are 12.4 million heavy-duty trucks in 2014.
Combined, the NGVA and MOVES numbers estimate that natural gas heavy-duty trucks
comprise 0.5 percent of the heavy-duty truck population.
Most projections show increasing natural gas consumption by the heavy duty truck fleet.
An obvious set of projections to review was the set of projections provided in the National
Academy of Sciences (NAS) report.31 The NAS report attached a figure, sourced from Citi
Research, which provided projections by ACT, PACCAR, Frost and Sullivan and the National
Petroleum Council.32 This figure is reproduced below as Figure 13-6.

-------
Figure 33. Near-Term Class 8 Natural Gas Penetration Forecasts
16% -
14% -
12% -
10% -
8% -
6% -
4% -
2% -
0% -I	1	1	1	1	1	
2012	2013	2014	2015	2016	2017
Source: Citi Research
Figure 13-6 Near-Term Class 8 Natural Gas Penetration Forecasts
All these studies referenced by the NAS start out on the basis that sales of natural gas
trucks comprise 2 percent of the Class 8 heavy duty truck fleet sales in 2012, and then project
different growth rates from that point forward. However, it is unlikely that any of these
projections considered the possibility that crude oil prices would collapse during the timeframe
of the projections. Starting during the summer of 2014, crude oil prices started to decline until
they reached a low price of under 30 dollars per barrel. Recently (early 2016), crude oil has been
selling in the range of 30 to 45 dollars per barrel. Natural gas vehicle truck sales declined in
2014 due to the lower diesel fuel prices, although most of the decline is attributed to lower light-
duty natural gas vehicle sales.33 There appear to be other drivers of natural gas truck sales, such
as air quality concerns and state subsidies that can offset the underlying economic factors. While
natural gas truck sales have been higher than 1 percent of total heavy-duty truck sales in recent
years, it is likely that they will fall below 1 percent if crude oil prices stay low for some time.
We tried to gain some insight in how each study referenced in the NAS report was
conducted. The ACT Research study shows the most aggressive growth rate for natural gas
heavy-duty trucks. The ACT Research projection did not seem to consider the economics of
natural gas versus diesel fuel. Instead, the ACT projection seemed to be based on a consumer
acceptance profile of a new technology, presumably assuming that the technology is already
ACT Research
—PACCAR-High
Frost St Sullivan - High
Frost Si Sullivan - Low
NPC - Reference
^—PACCAR - Low
Frost & Sullivan- Fcst

-------
economically competitive. In a recent ACT press release for a more recent report, it was
acknowledged that the growth rates ACT projected earlier were too aggressive and a more
modest growth rate is more likely.34 The NPC projection shows a similar growth rate as that
estimated by ACT Research, but NPC's projection for increased uptake of the natural gas
technology begins in 2015 instead of 2012. In its study, NPC assumed that the increased capital
cost for a natural gas truck compared to a diesel truck study decreases from $60,000 to $20,000
by 2040.35 This cost decrease seems excessive, and it is likely an important part of the
explanation of why the NPC study shows such a large increase in natural gas use by heavy-duty
trucks. We did not have access to core assumptions used in the PACCAR and Frost and Sullivan
projections to assess their viability.
We searched for the Citigroup report on the Web and in addition to the figure provided in
the NAS report, we found Citi Group's projection shown in the context of the other projections
referenced by NAS from the Citi bank report in Figure 13-7.36 Citi Group's projection is less
optimistic than the ACT projection, but is more optimistic than the NPC reference case
projection.
Figure 34. Long-Term Class 8 Natural Gas Penetration Forecasts
60% -|
Citi Penetration
E stimate
for 2020
ACT Research
NPC-High
NPC - Reference
PACCAR-High
PACCAR - Low
¦Frost &. Sullivan- High
Frost & Sullivan- Fest
F rost & Sullivan - Low
0%
2012 2016 2020 2024 2028 2032 2036 2040 2044 2043
Source: Citi Research
Figure 13-7 Long-Term Class 8 Natural Gas Penetration Forecasts

-------
In its Annual Energy Outlook, EIA projects the use of different fuels by the
transportation sector.37 This projection was not referenced in the NAS report, but our review
found it to be especially credible. We routinely use EIA projections for much of our analysis
work and thus, using it here would be consistent with other analyses we conducted for this
rulemaking. However, we also specifically reviewed the methodology EIA used to project use of
natural gas by trucks to assess its viability.
First, EIA estimates that natural gas fueled 0.4 percent of the energy use of heavy-duty
trucks in 2014 and this estimate is consistent with the fraction of the heavy-duty fleet which is
fueled by natural gas.
Second, the EIA projection is based on an economic analysis which considers the
increased cost of manufacturing a natural gas truck over a diesel truck, the fuel savings for using
natural gas instead of diesel fuel, and whether the payback time of the fuel savings against the
increased truck cost would trigger purchases of natural gas trucks. As part of this analysis, EIA
assumes that lighter heavy-duty trucks would use CNG which is a lower cost technology suited
for the shorter driving distances for these trucks. The long haul trucks, however, require larger
stores of fuel to extend the driving range which is satisfied by storing the natural gas as a liquid.
LNG has about 60 percent of the energy density of diesel fuel, compared to CNG which has only
25 percent of the energy density of diesel fuel. To satisfy the long driving range of the long haul
trucks, EIA assumed that they would use LNG as a fuel. All the assumptions used by EIA for
conducting its economic analysis seem reasonable.
Third, EIA is one of the several well-respected organizations in the world for collecting
and analyzing today's fuel prices and projecting future fuel prices. According to the Alternative
Fuels Data Center, one of the most important assumptions in projecting the future use of natural
gas in the transportation sector is the relative price of natural gas to the price of diesel fuel.
Figure 13-8 summarizes the total retail prices and the cost components that make up the final
average year 2014 retail prices of diesel fuel, CNG and LNG, whose prices are expressed on a
diesel gallon-equivalent basis.

-------
CNG
LNG
Diesel
B Raw Fuel Cost/Crude Oil ,\ Liquifaction/Refining o Distribution and Marketing BTaxes
Figure 13-8 Relative Retail Cost of CNG and LNG to Diesel Fuel ($/gal DGE 2014 crude oil prices)
In 2014, the natural gas price purchased by industrial users was about $6 per million
BTU, which corresponds to $0.60 per diesel gallon equivalent. The price of crude oil has been
volatile during 2014 as the Brent crude oil price started at about $110 per barrel ($2.38/gallon),
but decreased to under $50 per barrel towards the end of 2014. Of course, in 2015, the price of
crude oil dropped to under $40 per barrel, and even dropped to under $30 per barrel in 2016.
From EIA's website, the average retail diesel fuel price in the first part of 2014 was about $3.80
cents per gallon. When comparing the natural gas spot market price on a diesel equivalent basis
to the diesel fuel price, it appears that natural gas is priced about one quarter of the diesel fuel
price. However, if used as compressed natural gas, the natural gas must be distributed through
smaller distribution pipeline system that exists in cities, which dramatically increases the price of
the natural gas to $ 1.15 per DGE. Then the natural gas must be compressed and stored at a retail
outlet which adds another $1.30 per DGE. The estimated retail price of CNG is $2.35 on a diesel
gallon equivalent basis (DGE), or about $1.45 DGE less than diesel fuel.
Similarly, if natural gas is converted to LNG, the resulting retail LNG price is much
higher than the raw natural gas price. LNG liquefaction plants are assumed to be located close to
large transmission pipelines away from cities, thus, they would likely pay the same low price as
industrial users. However, for producing LNG, the natural gas must be liquefied which adds
about $0.75 DGE. When the LNG is transported to retail outlets and marked up, the LNG is
priced $0.60 per DGE higher. The tax applied to LNG is on a per gallon basis, thus, is much
more than on a DGE basis because of LNG's lower energy density. All these steps add
substantially to the price of the LNG and the estimated retail price of LNG is $2.65 DGE, or
$1.15 DGE less than diesel fuel.
The relative prices of CNG and LNG versus diesel fuel are quite different in 2015.
Figure 13-9 shows what impact the much lower crude oil price in 2015 has had on the relative
economics of using CNG and LNG.

-------
CNG	LNG	Diesel
>1 Raw Fuel Cost/Crude Oil	t" Liquifaction/Refining	1 Distribution and Marketing
Figure 13-9 Relative Retail Cost of CNG and LNG to Diesel Fuel ($/gal DGE 2015 crude oil prices)
As shown in Figure 13-9, LNG is priced nearly the same as diesel fuel, and CNG is
priced about $0.50 less than diesel fuel. If these relative fuel prices are accurate, there would be
no opportunity for paying down the higher capital costs of LNG trucks. While CNG tends to be
priced somewhat lower than diesel fuel, the breakeven point is certainly much longer than
desired by fleet owners. Also, given the volatility in crude oil prices, potential fleet owners may
be reluctant to move towards CNG and LNG because fleet owners will want a predictable
payback of their higher priced natural gas truck (and perhaps natural gas refueling facilities)
purchases.
In its projections, EIA estimates that crude oil prices will range from $70 to $80 per
barrel until 2021 and then increase to $100 per barrel in 2030 and increase to $140/BBL in 2040.
Natural gas prices are expected to only slightly increase over this period.
The fifth reason why the EIA projections seem reasonable is because the payback hurdle
assumptions assumed for truck fleet owners seem reasonable. EIA projects that natural gas
trucks begin to be purchased when the payback times are 4 years or less based on a survey
conducted by the American Trucking Associations. The ATA survey found that 24 percent of
respondents would choose natural gas trucks over diesel trucks if the payoff is 4 years, another
57 percent would choose natural gas if the payoff is 3 years, the next 15 percent would choose
natural gas if the payoff is 2 years and the last 5 percent would choose natural gas if the payoff is
1 year or less.38 This is consistent with some conversations we have had with some fleet owners.
The NAS cites the pay back for the extra cost of natural gas trucks as 2 years, but other sources
report a longer return closer to 4 years.H
H Early LNG Adopters Experience Mixed Results; Truck News, October 1, 2013.

-------
The results of EIA's economic analysis and projected natural gas use in heavy duty trucks
presented in the 2015 Annual Energy Outlook is presented in Figure 13-10.39
EIA Projection of Fuel Prices and Natural Gas Use by HD
8	Engines
4.5
3.5
2.5
NG percent of
1.5
0.5
2010
2015
2020
2025
2030
2035
2040
Year
Figure 13-10 EIA Projection of Fuel Prices and Natural Gas Use by HD Engines
Figure 13-10 shows, as we discussed above, that natural gas currently supplies only about
0.5 percent of total heavy-duty truck fuel demand and is expected to continue to do so until about
2030. Starting in 2033, EIA estimates that the price of diesel fuel will increase above $4 per
gallon which will create the economic incentive to purchase natural gas trucks. As expected, the
EIA projection that the price differential between natural gas and diesel fuel continues to
increase results in the effect that the uptake of natural gas use in the heavy-duty truck fleet
accelerates as the price differential increases.
A very interesting conclusion of the EIA projection is the natural gas penetration
differences between the different heavy-duty truck classes. Figure 13-11 summarizes the
projected use of natural gas by the AEO for different truck classes.

-------
20
18
Medium
2010	2015	2020	2025	2030	2035	2040	2045
Year
Figure 13-11 EIA Projection of NG use by Truck Weight Class
Figure 13-11 shows that the only heavy-duty sector which is projected to see a large
penetration of natural gas is the heavy, heavy-duty sector which increases to 17 percent by 2040.
The light and medium classes of the heavy-duty truck fleet do not show increases in natural gas
use. The likely reason EIA's analysis shows little CNG or LNG use by light and medium heavy-
duty trucks is because they are driven far less and their use does not justify the higher purchase
price. According to the Vehicle Inventory and Use Survey, light and medium heavy duty trucks
average less than l/3rd the annual mileage of the heaviest trucks.40 EIA is using a distribution of
VMT for new class 7 and 8 trucks as shown in Figure 13-12.

-------
Annual Miles Traveled (thousand miles)
Figure 13-12 Percent of Class 7 and 8 Truck Fleet by Annual Miles Traveled
Figure 13-12 shows that although about half of class 7 and 8 trucks are driven less than
60 thousand miles per year, the other half is driven from 60 thousand to over 200 thousand miles
per year. It is these high mileage long haul trucks which are prime candidates for using LNG
because of their ability to pay down the high marginal natural gas truck cost.
Since EIA does not report the payback times as an output of its projections, we conducted
our own analysis of sample payback times solely for illustrative purposes. We assessed the time
required for the lower fuel cost of LNG to payback the incremental truck cost of using LNG
assuming that a truck averages 120,000 miles per year. There were several important aspects of
the payoff analysis that we conducted. First, based on the EIA analysis which found that the
heavy, heavy-duty trucks sector is the only one which will see natural gas use increase
dramatically, we studied the payback of natural gas use with combination trucks. Second, as
concluded by EIA, we also assume that the higher energy density of LNG will make it the most
likely natural gas fuel type used by the heavy, heavy-duty trucks. Third, the higher natural gas
truck cost was approximated from the analysis EIA conducted for its Annual Energy Outlook.
Fourth, the analysis presents a simple payback as well as a discounted payback using a 7 percent
discount factor. Fifth, we evaluated the payback in 2015, and we also assessed what the payback
might be in 2020, 2030 and 2040 and assume some changes in the future years as discussed in
some example evaluation cases below. Table 13-21 presents the results of our payback analysis
for natural gas combination trucks.

-------
Table 13-21 Combination Truck Payback Analysis

CASE 1
CASE 2
CASE 3
CASE 4

2015 DUEL
2015 HPDI
2030 HPDI
2040 HPDI

FUELED



Miles per Year
120,000
120,000
120,000
120,000
Miles per Gallon
6.0
6.0
8.2
9.0
Incremental NG Truck Cost
55,000
70,000
60,000
55,000
($)




Incremental NG
970
1613
1935
1935
Maintenance Cost per year
($)




Diesel Fuel Price ($/gal)
2.70
2.70
3.84
4.38
Natural Gas Price ($/gal
2.90
2.90
2.69
3.36
DGE)




Diesel Fuel Cost per Year
54,000
54,000
56,200
58,400
Natural Gas Fuel Cost Per
58,400
62,500
44,080
49,710
Year (LNG)




Lower NG Efficiency (%)
5%
5%
5%
5%
Vehicle NG Use (%)
50%
95%
95%
95%
Simple Payback (years)
Negative
Negative
4.96
6.30
Discounted Payback (years)
Negative
Negative
6.1
8.3
We evaluated two different cases for 2015. Case 1 assumes a mixed fuel (MFNG) LNG
fueled, heavy-duty combination truck which exceeds 26,000 gross vehicle weight rating and
averages 120,000 miles per year. But because this truck can operate on diesel only, the truck can
manage with a more modest storage quantity of LNG, thus reducing the cost of LNG storage.
When this truck is fueled by LNG, it is estimated to be 5 percent less thermally efficient than a
similar diesel truck. The fuel costs are the average prices during 2015. The second case we
evaluated for 2015 is a direct injection natural gas (DING) truck. Because this truck must fuel
on LNG or be parked (the diesel fuel is simply used to enhance the combustion process), there
must be more LNG storage capacity and the truck purchase price is estimated to be $70,000
more than a diesel truck. This case also assumes 120,000 miles accumulated per year. In both
2015 cases, neither of the trucks achieve any degree of payback as the LNG fuel price is higher
than the diesel fuel price. This illustrates the difficulty fleet owners' face when considering the
purchase of natural gas trucks in this low crude oil price environment.
For Case 3, we assessed a 2030 case using EIA fuel price projections. Like the second
case, the truck is a DING truck, but because it is fifteen years later, we assumed a modest cost
reduction due to a learning curve. Due to the large price spread between diesel and natural gas,
this truck's discounted payback time is 6.1 years.
For Case 4, we assessed a 2040 case using EIA fuel price projections. Like the previous
two cases, the truck is a DING truck, and we assumed a further modest cost reduction due to a
learning curve. This truck is also assumed to accumulate 120,000 annual miles fueled on LNG.
While diesel fuel prices are expected to increase between 2030 and 2040, EIA projects that

-------
natural gas will increase faster than diesel fuel prices during this time and this truck's discounted
payback time actually increases to 8.3 years.
Since the EIA analysis was pessimistic with respect to truck types other than Class 8
combination trucks, we assessed the payback of refuse trucks. This is particularly interesting
because refuse trucks have been one of the bright spots for use of natural gas as a transportation
fuel source (from the standpoint of the natural gas industry). Refuse trucks have a couple of
advantages over combination trucks. One is that the constant stop-start characteristics of its
driving cycle result in very poor fuel economy, which increases natural gas use and shortens the
payback times for the higher engine and fuel system costs. Another advantage of refuse trucks is
that because they drive relatively short distances between refueling, they can rely on a smaller
quantity of CNG storage which reduces the CNG fuel system cost. Refuse trucks usually use
compression ignition-based engines which are converted to SING engines by adding a spark plug
and lowering the compression ratio. SING engines solely use natural gas and do not require (and
cannot operate on) any other fuel, which increases natural gas use and enhances payback
compared to duel fueled natural gas engines (DING).
Refuse trucks also have some disadvantages for using natural gas use. Since the refuse
trucks stop and start up so frequently, they cannot achieve a high annual mileage compared to
over-the-road trucks, and this offsets the advantages of the lower fuel economy. Since the
engines are usually converted to SING engine types (spark ignited), the engines are retrofitted to
use a lower compression ratio, which reduces its fuel economy by about 15 percent compared to
diesel fueled compression engines.
Figure 13-22 summarizes the payback analysis that we conducted for refuse trucks.

-------
Table 13-22 Refuse Truck Payback Analysis

CASE 1
CASE 2
CASE 3
CASE 4

2015 CNG
2020 CNG
2030 CNG
2040 CNG

REFUSE
REFUSE
REFUSE
REFUSE

TRUCK
TRUCK
TRUCK
TRUCK
Miles per Year
25,000
25,000
25,000
25,000
Miles per Gallon
2.8
2.9
3.8
4.2
Incremental NG Truck Cost
35,000
35,000
30,000
27,500
($)




Incremental NG
403
403
403
403
Maintenance Cost per year




Diesel Fuel Price ($/gal)
2.90
3.17
3.84
4.38
Natural Gas Price ($/gal
2.60
2.75
2.39
3.06
DGE)




Diesel Fuel Cost per Year
25,850
27,395
25,080
26,086
Natural Gas Fuel Cost Per
27,664
28,310
18,760
21,860
Year (CNG)




Lower NG Efficiency (%)
15
15
15
15
Vehicle NG Use (%)
100
100
100
100
Simple Payback (years)
Negative
Negative
4.7
6.5
Discounted Payback (years)
Negative
Negative
5.8
8.7
Figure 13-22 shows, as for the combination trucks, that the collapse in crude oil prices
does not allow for a payback of the refuse truck's higher purchase price in 2015. However, even
when crude oil prices increase in 2030 and 2040, refuse trucks still experience a fairly long
payback period of 6 years or more on a discounted basis. Refuse trucks may actually show a
reasonable or even favorable payback if the waste management company which operates the
trucks is able to use waste methane gas from landfills at a much lower natural gas price point.
Given the negative payback for natural gas vehicles in 2015, it suggests that existing
subsidies for natural gas likely play an important role in encouraging its use. According to EIA,
half the natural gas consumption by cars and trucks is in California and this may be partially due
to subsidies and other incentives California offers. California subsidies the purchase price of
natural gas vehicles, and also offsets the cost of natural gas dispensing stations. The Low
Carbon Fuel Standard (LCFS) in place in California also incentivizes natural gas use because
natural gas is considered to cause less of an impact on the climate than petroleum-based gasoline
and diesel fuel. The majority of the other half of the NG fleet is also in states which subsidize
the natural gas truck purchase price or service station construction costs.
Based on the EIA projections for crude oil and natural gas prices, the payback time of
LNG trucks is expected remain long (more than 4 years) until sometime after 2030 when diesel
fuel prices are projected to increase above 4 dollars per gallon. Thus, natural gas use by heavy-
duty trucks is not projected to increase above 1 percent of the heavy-duty fuel demand until after
2030. Even if the economics improve for using CNG and LNG in the heavy-duty fleet, another
hurdle is fuel availability since these fuels are not already widely available. Figure 13-13 shows
the number of CNG and LNG public and private service stations relative to the number of
gasoline and diesel fuel service stations and truck stops, respectively.

-------
50000
to
Q.
2 45000
to
" 40000
¦D 35000
C
(0
|2 30000
o
fB 25000
to
0)
u
aj
to
20000
15000
a>
-Q
E
10000
5000
c
o
so +J
CO
^ CD
= ¦5
2 £
-9 
< CO
F
~150,000 total
Service Stations
1
I Planned
~ # Stations
o
sP
°j	
^	Q.
C	O
+j
00	u
M	=5
(D	>
Number of Total Number Number of Total Number
Service Stations of Service Trucks Stops of Truck Stops
Selling CNG Stations Selling LNG
Figure 13-13 CNG and LNG Availability at Service Stations
As Figure 13-13 shows, CNG and LNG fuel availability at service stations is 1 percent or
less of the availability of gasoline and diesel fuel. Even if a business owner finds the purchase of
one or more new natural gas trucks an attractive investment, if the fuel is not available in the
area, the business owner may have to forgo purchasing the natural gas trucks. A fleet owner
might be in the position to also install a natural gas service station or establish a contract with a
third party fuel provider to provide the fuel, but that may require making a large purchase of
trucks to justify the installation of the service station or the establishment of the contract. If the
fleet owner would need to build a CNG or LNG refueling station to enable purchasing the natural
gas trucks, then the combined cost of the service station installation and the natural gas truck
purchase could make the prospect uneconomic even if the natural gas truck purchase by itself
would be justified. LNG availability is particularly challenging because in addition to an LNG
service station, a LNG liquefaction plant would be needed as well. If the economics turn
favorable for using natural gas in the truck fleet, the conversion to natural gas is likely to be slow
due to the need to build out the fuel availability.

-------
13.2.1 Dimethyl Ether
Although NAS focused its recommendations on natural gas, it also discussed dimethyl
ether (DME), which is a potential heavy-duty truck fuel sourced from natural gas. Dimethyl
ether has a high cetane number (more than 55), although its energy density is about 60 percent of
that of diesel fuel. Dimethyl ether is a volatile fuel, like liquid petroleum gas, that can be stored
as a liquid at normal ambient temperatures under moderate pressure. Typical DME fuel tanks
would be designed to prevent any significant evaporative emissions.
A DME fueled truck is only modestly more expensive than a diesel fuel truck. The fuel
tank is more expensive than a diesel fuel tank, but much less expensive than an LNG tank since
it does not need to be heavily insulated. The engine modifications to enable using DME are also
modest. Because DME does not have carbon-carbon bonds that form particulate particles during
combustion, the particulate filter, which is normally installed on recent year diesel trucks, can be
eliminated. This offsets some of the increased DME engine and fuel tank costs.
Although DME is sourced from cheap natural gas, the conversion of natural gas to DME
and moving the fuel to retail outlets greatly increases the cost of the fuel. As Figure 13-14
shows, DME is more expensive than LNG, but still lower in cost than diesel fuel. Similar to
Figure 13-8, the diesel fuel price used in Figure 13-14 is based on crude oil prices in early 2014.
a nn
en

J. JU
3.°° HE


2.50 	
So 2.00 H
\fy 85?




~Z 1.50 KK
to
°
U 1'°° n
°'50 S
n nn 881











CNG LNG Diesel DME
¦ Raw Fuel Cost/Crude Oil \ Liquifaction/Refining 5 Distribution and Marketing ¦ Taxes
Figure 13-14 Relative Retail Cost of DME to CNG, LNG and Diesel Fuel ($/gal DGE)
DME is estimated to cost $3.50/ DGE, or $0.30 DGE less than diesel fuel. Using the
much lower crude oil prices in early 2016 would show that DME is more expensive than diesel
fuel.
Because there is very little DME use in the US (there is only a very small fleet of DME
trucks being contemplated in California), we did not conduct a lifecycle assessment of DME.

-------
We will, however, discuss a few aspects of a lifecycle analysis for DME. First, since DME is
sourced from natural gas, the upstream methane emissions from the natural gas industry would
still be allocated to DME. Second, there are not venting issues associated with DME as with
LNG or CNG refueling. Third, because DME has a lifetime of less than one week, it is not a
long-lived well-mixed gas, and therefore has little direct climate impacts.41

-------
References
1	US Greenhouse Gas Emissions and Sinks: 1990 -2014, US Environmental Protection Agency, April 15, 2016
(EPA 430-R-16-002).
2	National Emission Standards for Hazardous Air Pollutants (NESHAP): for the Oil and Natural Gas Production and
Natural Gas Transmission and Storage, Final Rule, 40CFR part 63, subpart HH; June 17, 1999.
3	Oil and Natural Gas Sector: New Source Performance Standards and National Emission Standards for Hazardous
Air Pollutants Reviews; Final Rule, 40 CFR parts 60 and 63, Environmental Protection Agency, August 16, 2012.
4	www3.epa.gov/gasstar/.
5	Oil and Natural Gas Sector: Emission Standards for New and Modified Sources; 40CFR 60, May 12, 2016.
6	https://www.whitehouse.gOv/the-press-office/2016/03/10/us-canada-joint-statement-climate-energy-and-arctic-
leadership.
7	https://blog.epa.gov/blog/2016/03/epa-taking-steps-to-cut-methane-emissions-from-existing-oil-and-gas-sources
8	2016 Second Biennial Report of the United States of America, Under the United Nations Framework Convention
on Climate Change. Table 4.
9	GREET 1 Model, Argonne National Laboratory, 2015.
10	Inventory of US Greenhouse Gas Emissions and Sinks: 1990 - 2014, Annex 2, Tables A-38 and A-39, April 15,
2016.
11	GREET 1 Model, Argonne National Laboratory, 2015.
12	Methane Emissions from the Natural Gas Industry - Volume 2: Technical Report; Prepared by the Natural Risk
Management Research Laboratory, EPA - 600/R-96-080b, June 1996.
13	Harrison, Matthew R., Allen, David, et al; Natural Gas Industry Methane Emission Factor Improvement Study
Final Report, Cooperative Agreement XA-83376101, Prepared for the Environmental Protection Agency, December
2011.
14	Dan Hanson of Ariel Compressors, November 2015.
15	E/ECE/324/Rev.2/Add.l09/Rev.3 - E/ECE/TRANS/505/Rev.2/Add.l09/Rev.3; United Nations; June 25, 2014.
16	Jenks, Christopher; Technology Assessment of Refueling-Connection Devices for CNG, LNG and Propane;
Transit Cooperative Research Program, Conducted by the Science Applications International Corporation (SAIC)
for the Federal Transit Administration, April 1998, Number 25.
17	Detailed California-Modified GREET Pathway for Liquefied Natural Gas (LNG) from North American and
Remote Natural Gas Sources, Version 1.0, July 20, 2009.
18	Bronfenbrenner, James C. et al, Selecting a Suitable Process; LNGIndustry.com; summer 2009.
19	Kunert, Sebastian et al, Small is Beautiful - Mini LNG Concept; Hamworthy Gas Systems; June 2008.
20	National Energy Technology Laboratory, Lifecycle Inventory Data - Process Documentation File;
DF_Stagel_0_2010. l.doc.
21	Detailed California-Modified GREET Pathway for Liquefied Natural Gas (LNG) from North American and
Remote Natural Gas Sources, Version 1.0, July 20, 2009.
22	Powers, Charles A., Best Practices to Avoid LNG Fueling Station Venting Losses; Prepared for the Brookhaven
National Laboratory, Contract 157688, June 25, 2010.
23	GREET 1 Model, Argonne National Laboratory, 2015.
24	The emission standards were required by the rulemaking named "Greenhouse Gas Emissions Standards and Fuel
Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles" published September 15, 2011.
25	LNG ANG Venting Calculations Spreadsheet, LNG ANG Venting Calcs.xlsx.
26	GREET 1 Model, Argonne National Laboratory, 2015.
27	GREET 1 Model, Argonne National Laboratory, 2015.
28	Cooney, Greg of Booze Allen Hamilton; Approaches to Developing a Cradle-to-Grave Lifecycle Analysis of
Conventional Petroleum Fuels Produced in the U.S. with an Outlook to 2040; for the National Energy Technology
Laboratory (NETL), October 6, 2015.
29	CA-GREET 1.8b versus 2.0 CI Comparison Table, LCFS Workshop Handout, California Air Resources Board,
April 3, 2015.

-------
30	Annual Energy Outlook 2015 with projections to 2040, US Energy Information Administration, April 2015.
(www.eia.gov/forecasts/aeo).
31	Reducing the Fuel Consumption and Greenhouse Gas Emissions of Medium- and Heavy-Duty Vehicles, Phase
Two: First Report, National Research Council, National Academy of Sciences, 2014.
32	Energy 2020: Trucks, Trains and Automobiles, Citi GPS: Global Perspectives and Solutions, June 2013.
33	FC Gas Intelligence, http://analysis.fc-gi.com/natural-gas -vehicles/us-heavy-duty-ngv-sales-robust-despite-oil-
price-impact-payback; July 8, 2015.
34	Natural Gas Adoption Increases Slowly, ACT Reports, HDT Truckinginfo, www.truckinginfo.com. September
17, 2014.
35	Advancing Technology for America's Transportation Future, Summary Report, National Petroleum Council,
August 1, 2012.
36	Energy 2020: Trucks, Trains and Automobiles, Citi GPS: Global Perspectives and Solutions, June 2013.
37	Issues in Focus: Section 6" Heavy-Duty Natural Gas Vehicles, Energy Information Administration, Annual
Energy Outlook 2012, June 25, 2012.
38	Conversation with Nicholas Chase, Energy Information Administration, October 2014.
39	Annual Energy Outlook 2015 with projections to 2040, US Energy Information Administration, April 2015.
(www.eia.gov/forecasts/aeo).
40	U.S. Department of Commerce, Bureau of the Census, 2002 Vehicle Inventory and Use Survey, Microdata, CD,
2005. (Additional resources: www.census.gov/svsd/www/tiusview.html).
41	Good, D.A et al, Lifetimes and global warming potentials for dimethyl ether and for fluorinated ethers:
CH30CF3, CHF20CHF2, CHF20CF3, Journal of Geophysical Res, Vol 103, Pages 28,181 - 28,186, November
20, 1998.

-------