Control of Air Pollution from New Motor
Vehicles: Heavy-Duty Engine and
Vehicle Standards
Draft Regulatory Impact Analysis
gPk United States
Environmental Protection
^1 Agency
-------
Control of Air Pollution from New Motor
Vehicles: Heavy-Duty Engine and
Vehicle Standards
Draft Regulatory Impact Analysis
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
£% United States
Environmental Protection
^1 Agency
EPA-420-D-22-001
March 2022
-------
Table of Contents
Table of Contents i
List of Acronyms v
List of Tables xiv
List of Figures xxvii
Executive Summary 1
Chapter 1 Technology to Control Emissions from Heavy-Duty Engines 6
1.1 Compression-Ignition Engine Technologies 6
1.1.1 Current Catalyst Technologies 6
1.1.2 Catalyst Durability 12
1.1.3 Improving SCRNOx Reduction at Low Exhaust Temperatures 15
1.1.4 Closed Crankcases 19
1.1.5 Opposed-Piston Diesel Engines 22
1.2 Spark-Ignition Engine Technologies 23
1.2.1 Technology Description for NMHC, CO, and NOx Control 23
1.2.2 Technology Description for PM Control 32
1.2.3 Technologies to Address Evaporative Emissions 33
1.3 Fuels Considerations 36
1.3.1 Natural gas 36
1.3.2 Biodiesel 37
1.4 Advanced Powertrain Technologies 50
1.4.1 Hybrid 50
1.4.2 Battery-Electric and Fuel Cell 53
Chapter 2 Compliance Provisions 63
2.1 Compression-Ignition Engine Dynamometer Test Procedures 63
2.1.1 Current CI Test procedures 63
2.1.2 Potential updates to CI Test procedures 65
2.2 Manufacturer-Run In-Use Testing Program for Compression-Ignition Engines 72
2.2.1 Current In-Use Program and Standards 72
2.2.2 Information evaluated for proposed updates 75
2.2.3 Proposed Updates to CI Engine In-Use Test Program and Off-Cycle Standards 92
2.3 Spark-Ignition Test Procedures and Standards 98
2.3.1 Current SI Test procedures 98
2.3.2 Proposed updates to SI Test procedures and Standards 100
2.3.3 Idle Test Procedures Considered 105
2.4 Compliance Assurance 105
2.4.1 Improved Engine Control Module Security as a Deterrent to Tampering 105
Chapter 3 Feasibility Analysis for the Proposed Standards 108
3.1 Compression-Ignition Technology Feasibility 108
3.1.1 Diesel Technology Demonstration Programs 108
3.1.2 Baseline Technology Effectiveness 126
3.1.3 Projected Heavy-Duty Diesel Technology Effectiveness 128
3.1.4 GHGImpacts 130
3.1.5 Technology Cost 131
3.2 Spark-Ignition Technology Feasibility 140
l
-------
3.2.1 Baseline Technology Effectiveness 140
3.2.2 Projected Technology Effectiveness 150
3.2.3 Technology Cost 163
Chapter 4 Health and Environmental Impacts 171
4.1 Health Effects Associated with Exposure to Pollutants 171
4.1.1 Ozone 171
4.1.2 Particulate Matter 173
4.1.3 Nitrogen Oxides 177
4.1.4 Carbon Monoxide 179
4.1.5 Diesel Exhaust 180
4.1.6 Air Toxics 182
4.1.7 Exposure and Health Effects Associated with Traffic 185
4.2 Environmental Effects Associated with Exposure to Pollutants 190
4.2.1 Visibility Degradation 190
4.2.2 Plant and Ecosystem Effects of Ozone 193
4.2.3 Deposition 194
4.2.4 Environmental Effects of Air Toxics 201
4.3. Climate-Related Effects from GHG Emissions 201
Chapter 5 Emissions Inventory 204
5.1 Introduction 204
5.2 Model and Data Updates 204
5.2.1 Methodology Overview 205
5.2.2 MOVES Emission Rates for Control Scenarios 207
5.3 National Emissions Inventory Results 244
5.3.1 Proposed Options 245
5.3.2 Alternative 246
5.3.3 Impacts of Heavy-Duty Gasoline Refueling Controls 246
5.4 Emissions Inventories for Air Quality Modeling 247
5.4.1 Control Scenario Evaluated for the Air Quality Modeling Analysis 248
5.4.2 Estimated Differences in the Emission Reductions between the SMOKE-MOVES and
National-Scale Emission Inventories 251
5.4.3 Estimated Differences in the Emission Reductions between Option 1 and the Control
Scenario Analyzed for Air Quality Modeling 252
5.5 Chapter 5 Appendix 253
5.5.1 Estimation of Engine-Certified Fraction among Model Year 2027 Diesel-Fueled
LHD2b3 Vehicles 253
5.5.2 Zero-Mile Emission Rates for the Control Scenarios 255
5.5.3 Details of the Emission Impacts of the Proposed Option 1 Program 256
5.5.4 Onroad Heavy-Duty NOx Emissions by Engine Operational Process for the Baseline,
Proposed Option 1, and Proposed Option 2 Standards 260
5.5.5 Year-Over-Year Criteria Pollutant Emissions for Calendar Years Between 2027 and
2045 261
5.5.6 Sensitivity Analysis for Battery-Electric Vehicles and Fuel Cell Electric Vehicles . 266
5.5.7 National Heavy-duty Vehicle Emissions Inventory Comparison of the California
Heavy-duty Omnibus Regulation and the Proposed Option 1 270
Chapter 6 Air Quality Impacts 279
ii
-------
6.1 Current Air Quality 279
6.1.1 Ozone 279
6.1.2 PM2.5 281
6.1.3 NO; 282
6.1.4 CO 283
6.1.5 Air Toxics 283
6.1.6 Visibility 284
6.1.7 Deposition 284
6.2 Air Quality Modeling Methodology 284
6.2.1 Air Quality Model 284
6.2.2 Model Domain and Configuration 284
6.2.3 Model Inputs 285
6.2.4 CMAQ Evaluation 286
6.2.5 Model Simulation Scenarios 286
6.3 Air Quality Modeling Results 287
6.3.1 Ozone Design Value Impacts of Proposed Rulemaking 288
6.3.2 Annual PM2.5 Design Value Impacts of Proposed Rulemaking 291
6.3.3 24-hour PM2.5 Design Value Impacts of Proposed Rulemaking 294
6.3.4 Nitrogen Dioxide Concentration Impacts of Proposed Rulemaking 298
6.3.5 Carbon Monoxide Concentration Impacts of Proposed Rulemaking 299
6.3.6 Air Toxics Impacts of Proposed Rulemaking 301
6.3.7 Visibility Impacts of Proposed Rulemaking 303
6.3.8 Deposition Impacts of Proposed Rulemaking 304
6.3.9 Demographic Analysis of Air Quality 305
Chapter 7 Program Costs 313
7.1 Technology Package Costs 313
7.1.1 Direct Manufacturing Costs 314
7.1.2 Indirect Costs 322
7.1.3 Technology Costs per Vehicle 327
7.2 Operating Costs 331
7.2.1 Costs Associated with Increased Diesel Exhaust Fluid (DEF) Consumption in Diesel
Engines 332
7.2.2 Costs Associated with ORVR and the Estimated Reduction in Fuel Costs for Gasoline
Engines 339
7.2.3 Repair Cost Impacts Associated with Longer Warranty and Useful Life Periods 341
7.3 Program Costs 356
7.3.1 Total Technology Costs 357
7.3.2 Total Operating Costs 365
7.3.3 Total Program Costs 375
Chapter 8 Estimated Benefits 377
8.1 Overview 377
8.2 Updates to EPA's Human Health Benefits Methods 377
8.3 Health Impact Assessment for PM2.5 and Ozone 380
8.3.1 Preparing Air Quality Modeling Data for Health Impacts Analysis 380
8.3.2 Selecting Air Pollution Health Endpoints to Quantify 382
8.3.3 Calculating Counts of Air Pollution Effects Using the Health Impact Function 385
iii
-------
8.3.4 Quantifying Ozone-Attributable Premature Mortality 385
8.3.5 Quantifying PIVfo.s-Attributable Premature Mortality 386
8.4 Economic Valuation Methodology for Health Benefits 387
8.5 Characterizing Uncertainty in the Estimated Benefits 389
8.6 Estimated Number and Economic Value of Health Benefits 391
8.7 Present Value of Total Benefits of Proposed Option 1 and 2 396
8.8 Unquantified Benefits 399
Chapter 9 Comparison of Benefits and Costs 402
9.1 Methods 402
9.2 Results 403
Chapter 10 Economic Impact Analysis 405
10.1 Impact on Sales, Fleet Turnover and Mode Shift 405
10.1.1 Sales 406
10.1.2 EPA's Research to Estimate Sales Effects 408
10.1.3 Fleet Turnover and Emissions Impacts 415
10.1.4 Potential for Mode Shift 416
10.1.5 Effects on Domestic and International Shares of Production 417
10.1.6 Summary of Sales, Turnover, and Mode Shift Impacts 418
10.2 Employment Impacts 418
10.2.1 Economic Framework for Employment Impact Assessment 418
10.2.2 Employment Impacts in the Motor Vehicle and Parts Manufacturing Sectors 420
10.2.3 Employment Impacts on Related Sectors 425
10.2.4 Summary of Employment Impacts 426
Chapter 11 Small Business Analysis 428
11.1 Definition and Description of Small Businesses 428
11.2 Overview of the Heavy-Duty Program and Type of Entities Covered 428
11.3 Impacts on Small Entities: Heavy-Duty Secondary Vehicle Manufacturers 429
11.4 Impacts on Small Entities: Heavy-Duty Vehicle Manufacturers 432
11.4.1 Heavy-Duty Electric Vehicle Manufacturers 433
11.4.2 Heavy-Duty Conventional Vehicle Manufacturers 433
11.5 Impacts on Small Entities: Heavy-Duty Alternative Fuel Engine Converters 434
11.6 Summary Table of Impacts on Small Businesses Subject to the Rule 435
References 436
iv
-------
List of Acronyms
Acronym
Definition
°C
Degrees Celsius
Microgram
nm
Micrometers
20xx$
U.S. Dollars in calendar year 20xx
A/C
Air Conditioning
ABT
Averaging, Banking and Trading
AC
Alternating Current
ACEA
European Automobile Manufacturers Association
ACES
Advanced Collaborative Emission Study
AECD
Auxiliary Emissions Control Device
AEO
Annual Energy Outlook
AES
Automatic Engine Shutdown
AFDC
Alternative Fuels Data Center
AHS
American Housing Survey
A1
Aluminum
A12Ti05
Aluminum Titanate
AMOC
Atlantic Meridional Overturning Circulation
AMT
Automated Manual Transmission
ANL
Argonne National Laboratory
ANPR(M)
Advanced Notice of Proposed Rulemaking
APU
Auxiliary Power Unit
AQ
Air Quality
AQCD
Air Quality Criteria Document
AR4
Fourth Assessment Report
ARB
California Air Resources Board
ARB HHDDT
California Air Resources Board Heavy Heavy-Duty Diesel Test
ASC
Ammonia Slip Catalyst
ASL
Aggressive Shift Logic
ASM
Annual Survey of Manufacturers
ASTM
ASTM International, formerly American Society for Testing and Materials
ASTM
ASTM International, formerly known as American Society for Testing and Materials
AT
Automatic Transmissions
ATA
American Trucking Association
ATIS
Automated Tire Inflation System
ATRI
Alliance for Transportation Research Institute
ATSDR
Agency for Toxic Substances and Disease Registry
ATUS
American Time Use Survey
Avg
Average
B100
1 methyl-ester biodiesel fuel
B20
0.2 biodiesel blended with 0.8 petroleum distilate diesel fuel
BenMAP
Benefits Mapping and Analysis Program
BETP
Bleed Emissions Test Procedure
BEV
Battery Electric Vehicle
bhp
Brake Horsepower
bhp-hr
Brake Horsepower Hour
BLS
Bureau of Labor Statistics
BMEP
Brake Mean Effective Pressure
BSFC
Brake Specific Fuel Consumption
BTS
Bureau of Transportation Statistics
BTS
Bureau of Labor Statistics
v
-------
BTU
British Thermal Unit
Ca
Calcium
CAA
Clean Air Act
CAAA
Clean Air Act Amendments
CaC03
Calcium Carbonate
CAD/CAE
Computer Aided Design And Engineering
CAE
Computer Aided Engineering
CAFE
Corporate Average Fuel Economy
CAN
Controller Area Network
CARB
California Air Resources Board
CaS04
Calcium Sulfate
CBI
Confidential Business Information
CCP
Coupled Cam Phasing
CCSP
Climate Change Science Program
CDA
Cylinder Deactivation
CDC
Centers for Disease Control
CDPF
Catalyzed Diesel Particulate Filter
CFD
Computational Fluid Dynamics
CFR
Code of Federal Regulations
ch4
Methane
CI
Compression-ignition
CILCC
Combined International Local and Commuter Cycle
CIPM
International Committee for Weights and Measures (Bureau International des Poids et
Mesures)
CITT
Chemical Industry Institute of Toxicology
CMAO
Community Multiscale Air Quality
CNG
Compressed Natural Gas
CO
Carbon Monoxide
C02
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
CPSI
Cells per Square Inch
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
Cu
Copper
CuO
Copper(II) oxide or cupric oxide
CVD
Cardiovascular Disease
CVT
Continuously-Variable Transmission
CW
Curb Weight
D/UAF
Downward and Upward Adjustment Factor
DAAAC
Diesel Aftertreatment Accelerated Aging Cycle
DARAP
Diesel Aftertreatment Rapid Aging Protocol
DCP
Dual Cam Phasing
DCT
Dual Clutch Transmission
vi
-------
DE
Diesel Exhaust
DEAC
Cylinder Deactivation
DEER
Diesel Engine-Efficiency and Emissions Research
DEF
Diesel Exhaust Fluid
deSOx
Removal of sulfur oxide compounds
DF
Deterioration Factor
DHHS
U.S. Department of Health and Human Services
Diesel HAD
Diesel Health Assessment Document
DMC
Direct Manufacturing Costs
DO
Dissolved Oxygen
DOC
Diesel Oxidation Catalyst
DOD
Department of Defense
DOE
Department of Energy
DOHC
Dual Overhead Camshaft Engines
DOT
Department of Transportation
DPF
Diesel Particulate Filter
DPM
Diesel Particulate Matter
DR
Discount Rate
DRIA
Draft Regulatory Impact Analysis
DWL
Discrete Variable Valve Lift
EAS
Exhaust Aftertreatment System
EC
European Commission
EC
Elemental Carbon
EC
Economic Census
ECM
Electronic Control Module
ED
Emergency Department
EERA
Energy and Environmental Research Associates
EFR
Engine Friction Reduction
EGR
Exhaust Gas Recirculation
EHPS
Electrohydraulic Power Steering
EIA
Energy Information Administration (part of the U.S. Department of Energy)
EISA
Energy Independence and Security Act
EIVC
Early Intake Valve Closing
EMS-HAP
Emissions Modeling System for Hazardous Air Pollution
EO
Executive Order
EPA
Environmental Protection Agency
EPMA
electron probe microanalysis
EPS
Electric Power Steering
ERG
Eastern Research Group
ERM
Employment Requirements Matrix
ESC
Electronic Stability Control
ETC
Electronic Throttle Control
ETW
Estimated Test Weight
EV
Electric Vehicle
F
Frequency
FCEV
Fuel Cell Electric Vehicle
Fe
Iron
FEL
Family Emission Limit
FET
Federal Excise Tax
FEV1
Functional Expiratory Volume
FHWA
Federal Highway Administration
FIA
Forest Inventory and Analysis
FMCSA
Federal Motor Carrier Safety Administration
vii
-------
FOH
Fuel Operated Heater
FR
Federal Register
FRM
Final Rulemaking
FTP
Federal Test Procedure
FVC
Forced Vital Capacity
s
Gram
g/s
Gram-per-second
g/ton-mile
Grams emitted to move one ton (2000 pounds) of freight over one mile
gal
Gallon
gal/1000 ton-
mile
Gallons of fuel used to move one ton of payload (2,000 pounds) over 1000 miles
GCAM
Global Change Assessment Model
GCW
Gross Combined Weight
GCWR
Rated Gross-combined Weight (vehicle + trailer)
GDI
Gasoline Direct Injection
GDI
Gasoline Direct Injection
GDP
Gross Domestic Product
GEM
Greenhouse gas Emissions Model
GEOS
Goddard Earth Observing System
GHG
Greenhouse Gas
GIFT
Geospatial Intennodal Freight Transportation
GPF
Gasoline Particulate Filter
GREET
Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation
GSF1
Generic Speed Form one
GUI
Graphical User Interface
GVWR
Gross Vehicle Weight Rating
GVWR
Rated Gross Vehicle Weight
GWP
Global Wanning Potential
H20
Water
HABs
Harmful Algal Blooms
HAD
Diesel Health Assessment Document
HC
Hydrocarbon
HD
Heavy-Duty
HDDEFTP
Heavy-Duty Diesel Engine Federal Test Procedure
HDE
Heavy-Duty Engine
HDOE FTP
Heavy-Duty Otto-Cycle Engine Federal Test Procedure
HDT
Heavy-Duty Truck
HDUDDS
Heavy Duty Urban Dynamometer Driving Cycle
HDV
Heavy-Duty Vehicle
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
HNCO
Iso-cyanic Acid
hp
Horsepower
lirs
Hours
HRV
Heart Rate Variability
HSC
High Speed Cruise Duty Cycle
HTUF
Hybrid Truck User Forum
viii
-------
HWFE
Highway Fuel Economy Drive Cycle
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
ICP-MS
nductively coupled plasma mass spectrometry
IMAC
Improved Mobile Air Conditioning
IMPROVE
Interagency Monitoring of Protected Visual Environments
IPCC
Intergovernmental Panel on Climate Change
IRAF
infrequent regeneration factor
IRFA
Initial Regulatory Flexibility Analysis
IRIS
Integrated Risk Information System
ISA
Integrated Science Assessment
JAMA
Journal of the American Medical Association
k
Thousand
K
Potassium
kg
Kilogram
KI
kinetic intensity
km
Kilometer
km/h
Kilometers per Hour
kW
Kilowatt
L
Liter
LA92
Inventory development dynamometer driving schedule
lb
Pound
LD
Light-Duty
LDT
Light-Duty Truck
LHD
Light Heavy-Duty
LIVC
Late Intake Valve Closing
LLC
Low Load Cycle
LLNL
Lawrence Livermore National Laboratory's
LPG
Liquified Petroleum Gas
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
MAGICC
Model for the Assessment of Greenhouse-gas Induced Climate Change
MCF
Mixed Conifer Forest
MD
Medium-Duty
MDPV
Medium-Duty Passenger Vehicle
MECA
Manufacturers of Emissions Control Association
mg
Milligram
Mg
Magnesium
Mg(OH)2
Magnesium Hydroxide
mg/hp-hr
Milligrams per horsepower-hour
ix
-------
MHD
Medium Heavy-Duty
MHEV
Mild Hybrid
mi
mile
min
Minute
MM
Million
MMBD
Million Barrels per Day
MMT
Million Metric Tons
Mn
Manganese
MOVES
MOtor Vehicle Emissions Simulator
MP-AES
Microwave Plasma Atomic Emission Spectroscopy
mpg
Miles per Gallon
mph
Miles per Hour
MRL
Minimal Risk Level
MSAT
Mobile Source Air Toxic
MT
Manual Transmission
MTS
Maximum Test Speed
MW
Megawatt
MY
Model Year
N2
Molecular Nitrogen
N20
Nitrous Oxide
Na
Sodium
NA
Not Applicable
NAAQS
National Ambient Air Quality Standards
NACFE
North American Council for Clean Freight Efficiency
NAFA
National Association of Fleet Administrators
NAICS
North American Industry Classification System
NAS
National Academy of Sciences
NASTC
National Association of Small Trucking Companies
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
NH3
Ammonia
NHS
National Highway System
NHTSA
National Highway Traffic Safety Administration
NiMH
Nickel Metal-Hydride
NIOSH
National Institute of Occupational Safety and Health
NIST
National Institute for Standards and Technology
Nm
Newton-meters
NMHC
Nonmethane Hydrocarbons
NMMAPS
National Morbidity, Mortality, and Air Pollution Study
NO
Nitric Oxide
no2
Nitrogen Dioxide
NOAA
National Oceanic and Atmospheric Administration
NOx
Oxides of Nitrogen
NPRM
Notice of Proposed Rulemaking
X
-------
NPV
Net Present Value
NRC
National Research Council
NRC-CAN
National Research Council of Canada
NREL
National Renewable Energy Laboratory
NTE
Not-to-exceed
NTEA
National Truck and Equipment Association
NTP
National Toxicology Program
NVH
Noise Vibration and Harshness
O&M
Operating and maintenance
03
Ozone
OAQPS
Office of Air Quality Planning and Standards
OBD
Onboard diagnostics
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
OOIDA
Owner-Operator Independent Drivers Association
OPEC
Organization of Petroleum Exporting Countries
ORD
EPA's Office of Research and Development
ORNL
Oak Ridge National Laboratory
ORVR
On-Board Refueling Vapor Recovery
ORVR
Onboard refueling vapor recovery
OTAO
Office of Transportation and Air Quality
P
Phosphorus
Pa
Pascal
PAH
Polycyclic Aromatic Hydrocarbons
PCV
Positive Crankcase Ventilation
PEF
Peak Expiratory Flow
PEMFC
Proton-Exchange Membrane Fuel Cell
PEMS
Portable Emissions Monitoring System
PFI
Port Fuel Injection
PFI
Port Fuel Injection
PGM
Platinum Group Metal
PHEV
Plug-in Hybrid Electric Vehicles
PLT
Production-line testing
PM
Particulate Matter
PM10
Coarse Particulate Matter (diameter of 10 um or less)
PM2.5
Fine Particulate Matter (diameter of 2.5 um 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
xi
-------
RMC-SET
Ramped Modal Cycle Supplementary Emissions Test
RPE
Retail Price Equivalent
RPM
Revolutions per Minute
RSWT
Reduced-Scale Wind Tunnel
S
Second
S
Sulfur
SAB
Science Advisory Board
SAB-HES
Science Advisory Board - Health Effects Subcommittee
SAE
SAE International, formerly Society of Automotive Engineers
SAR
Second Assessment Report
SAV
Submerged Aquatic Vegetation
SBA
Small Business Administration
SBREFA
Small Business Regulatory Enforcement Fairness Act
SCR
Selective Catalyst Reduction
SEA
Selective enforcement audit
SER
Small Entity Representation
SET
Supplemental Emission Test
SGDI
Stoichiometric Gasoline Direct Injection
SHED
Sealed Housing Evaporative Determination
SHEV
Strong Hybrid Vehicles
SI
Spark-Ignition
SiC
Silicon Carbide
SIDI
Spark Ignition Direct Injection
S02
Sulfur Dioxide
SOA
Secondary Organic Aerosol
SOC
State of Charge
SOFC
Solid Oxide Fuel Cells
SOHC
Single Overhead Cam
SOx
Sulfur Oxides
SOx
Oxides of Sulfur
SPR
Strategic Petroleum Reserve
SSZ-13
A Chabazite-type Aluminosilicate ABC-6 Zeolite
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
TCO
Total Cost of Ownership
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
TRBDS
Turbocharging and Downsizing
TRU
Trailer Refrigeration Unit
TSD
Technical Support Document
TSS
Thermal Storage
TW
Test Weight
xii
-------
TWC
Three-Way Catalyst
U.S.
United States
U/DAF
Upward and Downward Adjustment Factor
UBE
Useable battery energy
UCT
Urban Creep and Transient Duty Cycle
UFP
Ultra Fine Particles
ULSD
Ultra-low sulfur diesel
URE
Unit Risk Estimate
USD A
United States Department of Agriculture
USGCRP
United States Global Change Research Program
uv
Ultraviolet
UV-b
Ultraviolet-b
VGT
Variable-geometry Turbine
VIN
Vehicle Identification Number
VIUS
Vehicle Inventory Use Survey
VMT
Vehicle Miles Traveled
voc
Volatile Organic Compound
VSL
Vehicle Speed Limiter
VTEC-E
Variable Valve Timing & Lift Electronic Control-Economy
VTRIS
Vehicle Travel Information System
WL
Variable Valve Lift
WT
Variable Valve Timing
WACAP
Western Airborne Contaminants Assessment Project
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
Zn
Zinc
ZSM-5
Zeolite Socony Mobil-5, an Aluminosilicate Pentasil Zeolite within the family of zeolites
Xlll
-------
List of Tables
Table ES-1: Projected Heavy-Duty Emission Reductions in 2045 from the Proposed Options 1
and 2 Standards 2
Table ES-2: Total Program Costs: Undiscounted Annual Costs in 2045 and Annualized Costs
through 2045 at 3% and 7% Discount Rates (Billions of 2017 dollars) 3
Table ES-3: 2045 Annual Value, Present Value and Equivalent Annualized Value of Benefits of
the Proposed Options 1 and 2 (billions, 2017$)a'b 3
Table ES-4: 2045 Annual Value, Present Value and Equivalent Annualized Value of Costs,
Benefits and Net Benefits of the Proposed Options 1 and 2 (billions, 2017$)a'b 4
Table 1-1: Engine-displacement specific catalyst substrate volume for MY2019 MHDDE and
HHDDE 11
Table 1-2: Average Crankcase Emission Rates (gram/hour) 20
Table 1-3: ORVR results with modeled values for test procedure limitations 33
Table 1-4: NREL Fuel Samples off Specification for ASTM D6751 (or equivalent B20) limit for
Na + K and Ca + Mg 43
Table 1-5: Test procedure LODs for NREL studies 43
Table 1-6: NREL 2016 Metals results for UOP-389 and ICP-MS70 45
Table 1-7: EPA 2017 Metals results for ICP-AES analysis of Na, K, Ca, and Mg performed by
CDF A 48
Table 1-8: EPA 2017 Metals results for ICP-AES analysis of Mo, P, B, Ba, Cu, Mn, Si, Ti, V,
and Zn performed by CDF A 49
Table 1-9: BEV & FCEV Sales Percentages 57
Table 1-10: BEV Truck Offerings or Planned Offerings in the US (as November 2019) 58
Table 1-11: BEV Bus Offerings or Planned Offerings in the US (as November 2019) 59
Table 2-1: SET Mode Weighting Factors for the 2010 NOx and Phase 1 GHG Standards 64
Table 2-2: New SET Mode Weighting Factors in Phase 2 66
Table 2-3: Breakdown of vehicles from combined NREL Fleet DNA and CARB datasets 68
Table 2-4: Representative Summary of GEM Generated Profiles for the Engine Duty-cycle 71
Table 2-5: NOx Emission Levels for a 2010 Compliant Engine on Three Candidate LLCs 72
Table 2-6: Engine Standards and In-use Measurement Allowance 73
Table 2-7: Percentage of Tests Meeting or Exceeding 0.90 Pass Ratio Threshold 74
Table 2-8: Average Pass Ratios for Data Submitted Since 2005 75
Table 2-9: Number of Diesel Vehicles with MY 2010+ Engines by NOx FEL Group from the
heavy-duty in-use testing Program 77
Table 3-1: Major engine specifications for the MY2017 Cummins X15 engine used for the
CARB Low NOx Stage 3 Research Program 108
Table 3-2: Summary of catalyst specifications for developmental EAS with light-off SCR 112
Table 3-3: Baseline (degreened) emissions results for the OE Cummins EAS 113
xiv
-------
Table 3-4: 0-hour (degreened) emissions results for the developmental EAS system with light-off
SCR 113
Table 3-5: Emissions results for the developmental EAS system with light-off SCR after 334
hours of accelerated thermal and chemical aging using the DAAAC (equivalent to
approximately 145,000 miles of operation). The SET (2021) results represent updated 40 CFR
§1036.505 SET procedures 114
Table 3-6: Emissions results for the developmental EAS system with light-off SCR after 667
hours of accelerated thermal and chemical aging using the DAAAC (equivalent to
approximately 290,000 miles of operation). The SET (2021) results represent updated 40 CFR
§1036.505 SET procedures 114
Table 3-7: Emissions results for the developmental EAS system with light-off SCR after 1000
hours of accelerated thermal and chemical aging using the DAAAC (equivalent to
approximately 435,000 miles of operation). The SET (2021) results represent updated 40 CFR
§1036.505 SET procedures 115
Table 3-8: Emissions results for the developmental EPA Stage 3 EAS system with light-off SCR
and separate DOC and DPF after 1000 hours of accelerated thermal and chemical aging using
the DAAAC (equivalent to approximately 435,000 miles of operation). The SET (2021)
results represent updated 40 CFR §1036.505 SET procedures 116
Table 3-9: Emissions results for the developmental EPA Stage 3 EAS system with light-off SCR
and separate DOC and DPF after 1379 hours of accelerated thermal and chemical aging using
the DAAAC (equivalent to approximately 600,000 miles of operation). The SET (2021)
results represent updated 40 CFR §1036.505 SET procedures 117
Table 3-10: Off-cycle NOx emissions results for the developmental EAS system with light-off
SCR and separate DOC and DPF after 1000 hours of accelerated thermal and chemical aging
using the DAAAC (equivalent to approximately 435,000 miles of operation) 120
Table 3-11: Off-cycle NMHC emissions results for the developmental EAS system with light-off
SCR and separate DOC and DPF after 1000 hours of accelerated thermal and chemical aging
using the DAAAC (equivalent to approximately 435,000 miles of operation) 121
Table 3-12: Off-cycle CO emissions results for the developmental EAS system with light-off
SCR and separate DOC and DPF after 1000 hours of accelerated thermal and chemical aging
using the DAAAC (equivalent to approximately 435,000 miles of operation) 121
Table 3-13: Summary of catalyst specifications for developmental EAS with close-coupled light-
off SCR 123
Table 3-14: Summary of catalyst specifications for developmental EAS with light-off SCR
mounted under-cab 124
Table 3-15: Major engine specifications for the MY2018 Cummins X15 engine used for EAS
and CD A development by EPA 125
Table 3-16: Summary of certification data for FTP cycle 127
Table 3-17: Summary of certification data for SET cycle 127
Table 3-18: 2018 Detroit DDI5 engine emissions in g/hp-hr 128
Table 3-19: 2018 Cummins B6.7 engine emissions in g/hp-hr 128
Table 3-20: 2018 Navistar A26 engine emissions in g/hp-hr 128
xv
-------
Table 3-21: MY2019 average engine displacement and diesel EAS specifications for Light HDE,
Medium HDE, and Heavy HDE applications based on projected-sales-weighted-averages
from EPA heavy-duty diesel engine certification data 132
Table 3-22: MY2019 DOC System Costs for Light HDE, Medium HDE, & Heavy HDE
Applications 133
Table 3-23: MY2019 CDPF System Costs for Engine-dynamometer Certified Light HDE,
Medium HDE, Heavy HDE, and Urban Bus Applications 133
Table 3-24: MY2019 SCR System Costs for Engine-dynamometer Certified Light HDE,
Medium HDE, Heavy HDE, and Urban Bus Applications 134
Table 3-25: Summary of MY2019 EAS Costs for Engine-dynamometer Certified Light HDE,
Medium HDE, Heavy HDE, and Urban Bus Applications 134
Table 3-26: Proposed Option 1 and 2 SCR System Costs for Engine-dynamometer Certified
Light HDE, Medium HDE, Heavy HDE, and Urban Bus Applications 136
Table 3-27: Summary of proposed Option 1 and 2 EAS Costs* for Engine-dynamometer
Certified Light HDE, Medium HDE, Heavy HDE, and Urban Bus Applications 137
Table 3-28: Summary of CD A Costs from Teardown Study 138
Table 3-29: Summary of Costs from the FEV Exhaust Aftertreatment System Study 139
Table 3-30: Heavy-Duty Gasoline Vehicle Emissions Investigation Vehicle Specifications.... 141
Table 3-31: Average distance from exhaust manifold to catalyst 141
Table 3-32: Description of Real-world PEMS Testing Routes 142
Table 3-33: Test weights and dynamometer coefficients used for NVFEL HD gasoline testing.
143
Table 3-34: Family Emission Limits Reported for the Six Certified HD SI Engines in MY 2019;
NOx and NMHC values are converted from g/hp-hr to mg/hp-hr to match the units of our
proposed standards 151
Table 3-35: Average emission performance for Certified HD SI Engines in MY 2019 151
Table 3-36: Comparison of Simulated 6.8L V10 SET Composite Emissions to Proposed
Standards 155
Table 3-37: Proposed Option 1 Spark-Ignition Exhaust Emission Standards for SET Duty Cycle
156
Table 3-38: SET Operation Mode Power Comparison 158
Table 3-39: Major engine specifications of the MY2019 HD SI gasoline engine used for the EPA
demonstration program 159
Table 3-40: Spark-Ignition Demonstration Program Preliminary FTP Results 160
Table 3-41: Spark-Ignition Demonstration Program Preliminary SET Results 161
Table 3-42: 2019 MY Sales-Weighted Baseline SI Engine Technology Costs (2019$) 164
Table 3-43: 2019 MY HHD and Urban Bus Gaseous-fueled Technology Baseline Costs (2019$)
164
Table 3-44: Projected Gaseous Fueled Engine Technology Cost based on Proposed Standards
(2019$) 165
xvi
-------
Table 3-45: Projected Liquid Fueled SI Engine Technology Cost to meet Proposed Standards
($2019$) 165
Table 3-46: Summary of HD SI Engine Technology Cost Comparison 166
Table 3-47: Assumptions for gasoline-fueled heavy-duty spark-ignition vehicles for conventional
carbon requirements to meet the proposed ORVR 169
Table 3-48: Estimated Costs for ORVR Over Tier 3 as Baseline 170
Table 5-1: Updates to MOVES CTINPRM from MOVES2014b 205
Table 5-2: MOVES Heavy-duty Regulatory Classes and Relevant MOVES Fuel Types 208
Table 5-3: MOVES Running Operating Mode Definitions 210
Table 5-4: Heavy-duty Compression Ignition Duty-Cycle Cycle NOx Standards for the Proposed
Options and Alternative Scenarios 211
Table 5-5: Proposed Option 1 Weighted Average Heavy Heavy-duty Compression Ignition
Duty-Cycle Test NOx Standards 212
Table 5-6: Rduty Ratios Calculated for Each Scenario 213
Table 5-7: Calculation of Rin use by MOVES Operation Mode 216
Table 5-8. Calculated Off-Cycle NOx Standards used for the Control Scenarios 219
Table 5-9: Calculation of Voluntary IdleNOx/C02 Standard (g/kg) 220
Table 5-10: Calculation of the Off-cycle NOx Standard Compliant Emission Rate for HHD
Diesel Vehicles for the Proposed Option 1 221
Table 5-11: Useful Life and Warranty Periods for Heavy-duty DieselA Engines and
Aftertreatment Systems in the Control Scenarios 226
Table 5-12: Estimated Vehicle Age at the End of the Warranty Period and the Useful Life for
Each Heavy-duty Diesel Regulatory Class for the Proposed Option 1 Model Year 2027
Scenario 228
Table 5-13: Estimated Vehicle Age at the End of the Warranty Period and the Useful Life for
Each Heavy-duty Diesel Regulatory Class in the Baseline and Control Scenarios 228
Table 5-14: NOx Tampering & Mal-maintenance (T&M) Emission Effects 231
Table 5-15: Calculation of sa 232
Table 5-16: MOVES ageGroupID Which Are Used to Define Running and Start Emission Rates
232
Table 5-17: Calculation of NOx 12-hour Cold Starts from the CARB Stage 1 HHD Engine from
the Cold and Hot FTP Cycle 235
Table 5-18: HHD Cold Start Emissions for Proposed and Alternative Scenarios 236
Table 5-19: Calculation of HHD and MHD Extended Idle NOx g/hr Emission Rates 238
Table 5-20: Running Emission Rate Reductions From Heavy-duty Gasoline Vehicles Due to
Proposed and Alternative Standards, Rgasoiine, Across All Heavy-duty Gasoline Regulatory
Classes and Operating Modes 241
Table 5-21: Phase-In of Onboard Refueling Vapor Recovery (ORVR) for Heavy-duty Trucks 244
xvii
-------
Table 5-22: National Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,
2040, and 2045 — Proposed Option 1 Program Emissions Relative to Heavy-Duty Vehicle
Emissions Baseline 245
Table 5-23: National Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,
2040, and 2045 — Proposed Option 2 Program Emissions Relative to Heavy-Duty Vehicle
Emissions Baseline 246
Table 5-24: National Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,
2040, and 2045 — Alternative Program Emissions Relative to Heavy-Duty Vehicle Emissions
Baseline 246
Table 5-25: Emission Reductions Due to Adoption of ORVR for Heavy-duty Vehicles Relative
to Heavy-Duty Vehicle Emissions Baseline 247
Table 5-26. Summary of Differences between Emissions in Proposed Option 1 and the Control
Scenario Analyzed for Air Quality Modeling 249
Table 5-27: Duty-Cycle NOx Standards for Proposed Option 1 and the Control Scenario
Analyzed for Air Quality ModelingA 249
Table 5-28. Rduty Ratios Calculated for the Control Scenario Analyzed for Air Quality Modeling
250
Table 5-29: Warranty Mileages and Years in Option 1 and the Control Scenario Analyzed for Air
Quality Modeling 250
Table 5-30:Useful Life Mileages and Years in Option 1 and the Control Scenario Analyzed for
Air Quality Modeling 251
Table 5-31: Onroad Vehicle Emission Reductions from the Air Quality Modeling Control
Scenario Using SMOKE-MOVES Inventories and National Inventories 252
Table 5-32 Comparison of the Onroad Vehicle Emission Reductions from the Air Quality
Modeling Control Scenario vs. Option 1 253
Table 5-33: Sales Volumes of Model Year 2027 LHD2b3 Diesel-Fueled Vehicles Estimated by
MOVES CTINPRM 253
Table 5-34: National Heavy-duty Vehicle NOx Emissions (Annual US Tons) For Calendar Years
Between 2027 and 2045 262
Table 5-35: National Heavy-Duty Vehicle VOC Emissions (Annual US Tons) For Calendar
Years Between 2027 and 2045 263
Table 5-36: National Heavy-duty Vehicle PM2.5 (Exhaust Only) Emissions (Annual US Tons)
For Calendar Years Between 2027 and 2045 264
Table 5-37: National Heavy-Duty Vehicle CO Emissions (Annual US Tons) For Calendar Years
Between 2027 and 2045 265
Table 5-38: Fractions of the BEVs and FCEVs Based on AEO2018 Compared with the Fractions
of the BEVs Based on NREL Study in Each of the Heavy-Duty Vehicle Categories by Model
Year 267
Table 5-39: Percent Reduction from the Baseline for Proposed Option 1 and Sensitivity Cases
Based on AEO2018 and NREL in Calendar Years 2030, 2040, and 2045 268
Table 5-40: Percent Reduction from the Baseline for Proposed Option 2 and Sensitivity Cases
Based on AEO2018 and NREL in Calendar Years 2030, 2040, and 2045 268
xviii
-------
Table 5-41: Percent Reduction from the Baseline for the Alternative and Sensitivity Cases Based
on AEO2018 and NREL in Calendar Years 2030, 2040, and 2045 269
Table 5-42: Duty-Cycle NOx Standards for the CARB Omnibus and EPA Proposed Option 1A
271
Table 5-43: Weighted Average Heavy Heavy-duty Compression Ignition Duty-Cycle Test NOx
Standards used to estimate NOx emission rates for the Omnibus Nationwide Scenarios 272
Table 5-44: Calculated Average Off-Cycle Standards for the Omnibus from the Average Idling
and Engine Cycles Standards and Off-Cycle Scaling Factors 273
Table 5-45: Warranty Mileages and Years in Omnibus and EPA Proposed Option 1 274
Table 5-46: Useful Life Mileages and Years in Omnibus Program and EPA Proposed Option 1
275
Table 5-47: Extended Idle NOx emission rates for the Omnibus Program 275
Table 5-48: Running Emission Rate Reductions From Heavy-duty Gasoline Vehicles For the
EPA Proposed Option 1 and Omnibus Nationwide Scenarios 276
Table 5-49 National Heavy-duty Vehicle NOx Emission Reductions Relative to the Baseline
Case For Omnibus Nationwide Scenarios (Reductions Relative to EPA Proposed Option 1
Shown for Comparison) 277
Table 6-1: Average Change in Projected 8-hour Ozone Design Values in 2045 due to Proposed
Rule 290
Table 6-2: Change in 8-hour Ozone Design Values for Counties Projected to be Above the Level
of the 2015 8-hour Ozone NAAQS in 2045 291
Table 6-3: Average Change in Projected Annual PM2.5 Design Values in 2045 due to Proposed
Rule 293
Table 6-4: Change in Annual PM2.5 Design Values for Counties Projected to be Above the Level
of the Annual PM2.5 NAAQS in 2045 294
Table 6-5: Average Change in Projected 24-hour PM2.5 Design Values in 2045 due to Proposed
Rule 296
Table 6-6: Change in 24-hour PM2.5 Design Values for Counties Projected to be Above the 24-
hour PM2.5 NAAQS in 2045 297
Table 6-7: Demographic Analysis of Projected 2045 Ozone Reductions (ppb) from the Proposed
Rule, by Race/Ethnicity 308
Table 6-8: Demographic Analysis of Projected 2045 PM2.5 Reductions (|ig/m3) from the
Proposed Rule, by Race/Ethnicity 309
Table 6-9 Demographic Analysis of Projected 2045 Ozone Reductions (ppb) from the Proposed
Rule, by Poverty Status 310
Table 6-10 Demographic Analysis of Projected 2045 PM2.5 Reductions (|ig/m3) from the HD
2027 Proposed Rule, by Poverty Status 311
Table 7-1: ICCT Cost Estimates of 12L Diesel Engine-Related Emission Control Costs
Associated with Past US Emission Standards (2015 dollars) 315
Table 7-2: Diesel Engine-Related Emission Control System Costs in the "No Action" Baseline*
316
xix
-------
Table 7-3: Gasoline Engine-Related Emission Control System Costs in the "No Action"
Baseline* 316
Table 7-4: CNG Engine-Related Emission Control System Costs in the "No Action" Baseline*
317
Table 7-5 Diesel Technology and Package Direct Manufacturing Costs per Engine by Regulatory
Class, 2017 dollars* 318
Table 7-6 Gasoline Technology and Package Direct Manufacturing Costs per Engine by
Regulatory Class, 2017 dollars* 319
Table 7-7: CNG Technology and Package Direct Manufacturing Costs per Engine by Regulatory
Class, 2017 dollars* 319
Table 7-8: GDP Price Deflators* Used to Adjust Costs to 2017 Dollars 319
Table 7-9: Seed volume factors used in this analysis for each Option 322
Table 7-10: Retail Price Equivalent Factors in the Heavy-Duty and Light-Duty Industries 324
Table 7-11: Scaling Factors Applied to RPE Contribution Factors to Reflect Changes in their
Contributions, Diesel & CNG Regulatory Classes* 326
Table 7-12: Scaling Factors Applied to RPE Contribution Factors to Reflect Changes in their
Contributions, Gasoline Regulatory Classes 326
Table 7-13: Simplified Example of Indirect Warranty Costs Calculated on an Incremental vs.
Absolute Technology Package Cost (values are not from the analysis and are for presentation
only) 327
Table 7-14: MY2027 Technology Costs for LHD2b3 Diesel, Average per Vehicle, 2017 Dollars
328
Table 7-15: MY2031 Technology Costs for LHD2b3 Diesel, Average per Vehicle, 2017 Dollars
328
Table 7-16: MY2027 Technology Costs for LHD45 Diesel, Average per Vehicle, 2017 Dollars
328
Table 7-17: MY2031 Technology Costs for LHD45 Diesel, Average per Vehicle, 2017 Dollars
329
Table 7-18: MY2027 Technology Costs for MHD67 Diesel, Average per Vehicle, 2017 Dollars*
329
Table 7-19: MY2031 Technology Costs for MHD67 Diesel, Average per Vehicle, 2017 Dollars*
329
Table 7-20 MY2027 Technology Costs for HHD8 Diesel, Average per Vehicle, 2017 Dollars 329
Table 7-21: MY2031 Technology Costs forHHD8 Diesel, Average per Vehicle, 2017 Dollars
330
Table 7-22: MY2027 Technology Costs for Urban bus Diesel, Average per Vehicle, 2017
Dollars 330
Table 7-23: MY2031 Technology Costs for Urban bus Diesel, Average per Vehicle, 2017
Dollars 330
Table 7-24: MY2027 Technology Costs for LHD45, MHD67 & HHD8 Gasoline, Average per
Vehicle, 2017 Dollars 330
xx
-------
Table 7-25: MY2031 Technology Costs for LHD45, MHD67 & HHD8 Gasoline, Average per
Vehicle, 2017 Dollars 331
Table 7-26: MY2027 Technology Costs for HHD8 CNG, Average per Vehicle, 2017 Dollars 331
Table 7-27: MY2031 Technology Costs forHHD8 CNG, Average per Vehicle, 2017 Dollars 331
Table 7-28: MY2027 Technology Costs for Urban bus CNG, Average per Vehicle, 2017 Dollars
331
Table 7-29: MY2031 Technology Costs for Urban bus CNG Average per Vehicle, 2017 Dollars
331
Table 7-30: Diesel Exhaust Fluid Consumption Rates for Diesel Vehicles in the Baseline Case
332
Table 7-31: Derivation of DEF Consumption per Ton of NOx Reduced 333
Table 7-32: Diesel Exhaust Fluid Price per Gallon (2017 dollars) 334
Table 7-33: Diesel Exhaust Fluid Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with LHD Engines {cents/mile in 2017 dollars,
Undiscounted) * 335
Table 7-34: Diesel Exhaust Fluid Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with LHD45 Engines {cents/mile in 2017 dollars,
Undiscounted) 336
Table 7-35: Diesel Exhaust Fluid Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with MHD67 Engines {cents/mile in 2017 dollars,
Undiscounted) 337
Table 7-36: Diesel Exhaust Fluid Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with HHD8 Engines {cents/mile in 2017 dollars,
Undiscounted) 338
Table 7-37: Diesel Exhaust Fluid Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with Urban Bus Engines {cents/mile in 2017
dollars, Undiscounted) 339
Table 7-38: Gasoline Vehicle Average Retail Fuel Cost per Mile during the first 10 Years of
each Model Year Lifetime by MOVES Sourcetype equipped with LHD45 Engines*
{cents/mile in 2017 dollars, Undiscounted) 340
Table 7-39: Gasoline Vehicle Average Retail Fuel Cost per Mile during the first 10 Years of
each Model Year Lifetime by MOVES Sourcetype equipped with MHD67 Engines*
{cents/mile in 2017 dollars, Undiscounted) 341
Table 7-40: Gasoline Vehicle Average Retail Fuel Cost per Mile during the first 10 Years of
each Model Year Lifetime by MOVES Sourcetype equipped with HHD8 Engines* {cents/mile
in 2017 dollars, Undiscounted) 341
Table 7-41: Class-8 Diesel Repair & Maintenance Costs per Mile (100,000 miles per year, 2018
dollars*) 343
Table 7-42: Estimated Vehicle Age at the End of the Warranty Period and the Useful Life for
each Heavy Heavy-duty Diesel Combination Truck for the Baseline Case 344
Table 7-43: Class-8 Diesel Short-haul Combination Truck Repair & Maintenance Costs per Mile
(see Table 7-41) with MOVES HHD8 Mileage Accumulation 344
xxi
-------
Table 7-44: Class-8 Diesel Long-haul Combination Truck Repair & Maintenance Costs per Mile
(see Table 7-41) with MOVES HHD8 Mileage Accumulation 344
Table 7-45: Percentage of Total Repair & Maintenance Costs Attributable to Different Vehicle
Systems 346
Table 7-46: Diesel Emission Repair Costs, Average Cost per Mile during the first 10 Years of
each MY Lifetime by MOVES Sourcetypes equipped with LHD2b3 Engines {cents/mile in
2017 dollars, Undiscounted)* 350
Table 7-47: Diesel Emission Repair Costs, Average Cost per Mile during the first 10 Years of
each MY Lifetime by MOVES Sourcetypes equipped with LHD45 Engines {cents/mile in
2017 dollars, Undiscounted) * 351
Table 7-48: Diesel Emission Repair Costs, Average Cost per Mile during the first 10 Years of
each MY Lifetime by MOVES Sourcetypes equipped with MHD67 Engines {cents/mile in
2017 dollars, Undiscounted) * 352
Table 7-49: Diesel Emission Repair Costs, Average Cost per Mile during the first 10 Years of
each MY Lifetime by MOVES Sourcetypes equipped with HHD8 Engines {cents/mile in 2017
dollars, Undiscounted) * 353
Table 7-50: Diesel Emission Repair Costs, Average Cost per Mile during the first 10 Years of
each MY Lifetime by MOVES Sourcetypes equipped with Urban Bus Engines {cents/mile in
2017 dollars, Undiscounted) * 354
Table 7-51: Gasoline Emission Repair Costs, Average Cost per Mile during the first 10 Years of
each MY Lifetime by MOVES Sourcetypes equipped with LHD45 Engines {cents/mile in
2017 dollars, Undiscounted)* 354
Table 7-52: Gasoline Emission Repair Costs, Average Cost per Mile during the first 10 Years of
each MY Lifetime by MOVES Sourcetypes equipped with MHD67 Engines {cents/mile in
2017 dollars, Undiscounted)* 355
Table 7-53: Gasoline Emission Repair Costs, Average Cost per Mile during the first 10 Years of
each MY Lifetime by MOVES Sourcetypes equipped with HHD8 Engines {cents/mile in 2017
dollars, Undiscounted)* 355
Table 7-54: CNG Emission Repair Costs, Average Cost per Mile during the first 10 Years of
each MY Lifetime by MOVES Sourcetypes equipped with HHD8 Engines {cents/mile in 2017
dollars, Undiscounted)* 356
Table 7-55: CNG Emission Repair Costs, Average Cost per Mile during the first 10 Years of
each MY Lifetime by MOVES Sourcetypes equipped with Urban Bus Engines {cents/mile in
2017 dollars, Undiscounted) * 356
Table 7-56: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
Diesel, Millions of 2017 dollars * 358
Table 7-57: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
Diesel, Millions of 2017 dollars * 359
Table 7-58: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
Diesel by Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting * 359
Table 7-59: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
Diesel by Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting * 360
XXll
-------
Table 7-60: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
Diesel by Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting * 360
Table 7-61: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
Diesel by Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting * 360
Table 7-62: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
Gasoline, Millions of 2017 dollars * 361
Table 7-63: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
Gasoline, Millions of 2017 dollars * 362
Table 7-64: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
Gasoline by Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
362
Table 7-65: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
Gasoline by Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
362
Table 7-66: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
Gasoline by Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
363
Table 7-67: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
Gasoline by Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
363
Table 7-68: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
CNG, Millions of 2017 dollars * 363
Table 7-69: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
CNG, Millions of 2017 dollars * 364
Table 7-70: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
CNG by Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting * . 364
Table 7-71: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
CNG by Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting * . 364
Table 7-72: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
CNG by Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting * . 365
Table 7-73: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
CNG by Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting * . 365
Table 7-74: Technology Cost Impacts Relative to the Baseline Case, All Fuels by Proposed
Option, Millions of 2017 dollars, Present Values at 3% Discounting * 365
Table 7-75: Technology Cost Impacts Relative to the Baseline Case, All Fuels by Proposed
Option, Millions of 2017 dollars, Present Values at 7% Discounting * 365
Table 7-76: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
Diesel, Millions of 2017 dollars * 367
Table 7-77: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
Diesel, Millions of 2017 dollars * 368
Table 7-78: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
Diesel by Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting * 368
xxm
-------
Table 7-79: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
Diesel by Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting * 368
Table 7-80: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
Diesel by Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting * 369
Table 7-81: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
Diesel by Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting * 369
Table 7-82: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
Gasoline, Millions of 2017 dollars 369
Table 7-83: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
Gasoline, Millions of 2017 dollars 370
Table 7-84: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
Gasoline by Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
370
Table 7-85: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
Gasoline by Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
370
Table 7-86: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
Gasoline by Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
371
Table 7-87: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
Gasoline by Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
371
Table 7-88: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
CNG, Millions of 2017 dollars 371
Table 7-89: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
CNG, Millions of 2017 dollars 372
Table 7-90: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
CNG by Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting * . 372
Table 7-91: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
CNG by Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting * . 372
Table 7-92: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case,
CNG by Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting * . 373
Table 7-93: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case,
CNG by Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting * . 373
Table 7-94: Operating Cost Impacts Relative to the Baseline Case, All Fuels by Proposed
Option, Millions of 2017 dollars, Present Values at 3% Discounting * 373
Table 7-95: Operating Cost Impacts Relative to the Baseline Case, All Fuels by Proposed
Option, Millions of 2017 dollars, Present Values at 7% Discounting * 373
Table 7-96: Fuel Cost and Transfer Impacts of the Proposed Option 1 Relative to the Baseline
Case, Gasoline, Millions of 2017 dollars 374
Table 7-97: Fuel Cost and Transfer Impacts of Proposed Option 2 Relative to the Baseline Case,
Gasoline, Millions of 2017 dollars 374
xxiv
-------
Table 7-98: Total Technology & Operating Cost Impacts of the Proposed Option 1 Relative to
the Baseline Case, All Regulatory Classes and All Fuels, Millions of 2017 dollars 375
Table 7-99: Total Technology & Operating Cost Impacts of Proposed Option 2 Relative to the
Baseline Case, All Regulatory Classes and All Fuels, Millions of 2017 dollars 376
Table 8-1: Summary of CMAQ-Derived Population-Weighted Ozone and PM2.5 Air Quality
Metrics for Health Benefits Endpoints Associated with Proposed Option 1 382
Table 8-2: Health Effects of Ambient Ozone and PM2.5 384
Table 8-3: Estimated Avoided PM2.5 Mortality and Illnesses in 2045 for Proposed Option 1 (95%
Confidence Interval) a'b 393
Table 8-4: Estimated Avoided Ozone Mortality and Illnesses in 2045 for the Proposed Option 1
(95% Confidence Interval)21 394
Table 8-5: Estimated Economic Value of PM2.5- and Ozone-Attributable Premature Mortality
and Illnesses in 2045 for Proposed Option 1 (95% Confidence Interval; millions of 2017$)a
395
Table 8-6: Total Ozone and PM2.5-Attributable Benefits in 2045 for Proposed Option 1 (95%
Confidence Interval; billions of 2017$)a'b 396
Table 8-7: Benefits Inputs that Change Over Time used to Calculate Year-over-Year Estimates
397
Table 8-8: Undiscounted Stream and Present Value of Human Health Benefits from 2027
through 2045: Monetized Benefits Quantified as Sum of Short-Term Ozone Respiratory
Mortality Ages 0-99, and Long-Term PM2.5 All-Cause Mortality Ages 30+ (Discounted at 3%
and 7%; billions of 2017$)a b 398
Table 8-9: Undiscounted Stream and Present Value of Human Health Benefits from 2027
through 2045: Monetized Benefits Quantified as Sum of Long-Term Ozone Respiratory
Mortality Ages 30+, and Long-Term PM2.5 All-Cause Mortality Ages 65+ (Discounted at 3%
and 7%; billions of 2017$fb 399
Table 8-10: Unquantified Criteria Pollutant Health and Welfare Benefits Categories 401
Table 9-1: 2045 Annual Value, Present Value and Equivalent Annualized Value of Costs,
Benefits and Net Benefits of the Proposed Options 1 and 2 (billions, 2017$)a>b 404
Table 10-1: Pre and Low-Buy Sales Effects Coefficients 410
Table 10-2: Regulatory Costs and HD Vehicle Prices Used to Estimate Elasticities 411
Table 10-3: Elasticity Estimates 412
Table 10-4: Illustrative Pre-Buy Results from the 2031 Implementation Date for Proposed
Option 1 414
Table 10-5: Illustrative Low-Buy Results from the 2031 Implementation Date for Proposed
Option 1 415
Table 10-6: Sectors Used in this Analysis 422
Table 10-7: Employment per $1 Million Expenditures (2017$) in the Motor Vehicle
Manufacturing Sectora 423
Table 10-8: Estimated Employment Effects Due to Increased Costs of Vehicles and Parts (Cost
Effect), in Job-Years 425
xxv
-------
Table 11-1: Summary of Impacts on Small Businesses Subject to the Rule
xxvi
-------
List of Figures
Figure 1-1: Functional schematic showing relative positioning of exhaust emission control
components arranged within an in-line exhaust system (top) and integrated into a box-style
system (bottom) 7
Figure 1-2: Integrated series heavy-duty truck exhaust emission control systems from Cummins
Emission Solutions (top) and box-style system from Eberspacher (bottom), with cut-away
showing some of the internal components (bottom right) 8
Figure 1-3: NOx reduction efficiency of an early, developmental Cu-zeolite SCR formulation
relative to DOC and SCR inlet temperatures and SCR space velocity (adapted from
McDonald et al. 2011) 12
Figure 1-4: Potential layout of a 2027+ dual-SCR system in an in-line configuration (top) and
comparable components integrated to improve passive thermal management (bottom) 19
Figure 1-5: Tailpipe Exhaust and Crankcase Emission Rates from Two Heavy-Duty Diesel
Trucks 21
Figure 1-6: Schematic of an ORVR system 34
Figure 1-7: SCR performance over the hot start HDDE FTP61' 40
Figure 1-8: SCR NOx conversion for the first inch of aged SCR catalysts.54 42
Figure 1-9: SCR NOx conversion for the seventh inch of aged SCR catalysts.54 42
Figure 2-1: Window size comparison and load distribution profile 69
Figure 2-2: Example of a LLC 7 Candidate Cycle 71
Figure 2-3: Sample of valid and invalid NTE events, separated by exclusion zones 74
Figure 2-4: NOx sensor time to on after engine cold-start 76
Figure 2-5: MOVES Operating Modes (OpModes) by scaled tractive power and vehicle speed 78
Figure 2-6: MOVES OpMode Emission Rates from HHD Engine Broken Down by Engine
Family 79
Figure 2-7: NOx Emission Rates from 81 Vehicles with 0.20 g/bhp-hr FEL HHD Engines by
MOVES OpMode and Aftertreatment Temperature 80
Figure 2-8: NOx Emission Rates from 20 Vehicles with 0.20 g/bhp-hr FEL MHD Engines by
MOVES OpMode and Aftertreatment Temperature 80
Figure 2-9: NOx Emission Rates from 42 Vehicles with 0.20 g/bhp-hr FEL LHD Engines by
MOVES OpMode and Aftertreatment Temperature 81
Figure 2-10: Time Fraction from 81 Vehicles with 0.20 g/bhp-hr FEL HHD Engines by MOVES
OpMode and Aftertreatment Temperature 82
Figure 2-11: Time Fraction from 20 Vehicles with 0.20 g/bhp-hr FEL MHD Engines by MOVES
OpMode and Aftertreatment Temperature 82
Figure 2-12: Time Fraction from 42 Vehicles with 0.20 g/bhp-hr FEL LHD Engines by MOVES
OpMode and Aftertreatment Temperature 83
Figure 2-13: Total NOx Contribution from 81 vehicles with 0.20 g/bhp-hr FEL HHD Engines by
MOVES OpMode and Aftertreatment Temperature 84
xxvii
-------
Figure 2-14: Total NOx Contribution from 20 vehicles with 0.20 g/bhp-hr FEL MHD Engines by
MOVES OpMode and Aftertreatment Temperature 84
Figure 2-15: Total NOx Contribution from 42 vehicles with 0.20 g/bhp-hr FEL LHD Engines by
MOVES OpMode and Aftertreatment Temperature 85
Figure 2-16: Brake-specific NOx by Vehicle Speed Bins for 93 Vehicles with HHD Diesel
Engines and an FEL of 0.20 g/bhp-hr 86
Figure 2-17: Brake-specific NOx by Vehicle Speed Bins for 26 Vehicles with MHD Diesel
Engines and an FEL of 0.20 g/bhp-hr 87
Figure 2-18: Brake-specific NOx by Vehicle Speed Bins for 49 Vehicles with LHD Diesel
Engines and an FEL of 0.20 g/bhp-hr 87
Figure 2-19: Brake-specific NOx by Window Average Power Bins for 85 Vehicles with HHD
Diesel Engines and an FEL of 0.20 g/bhp-hr 88
Figure 2-20: Vehicle Speed Profile of HD Duty Cycles 89
Figure 2-21: MOVES OpMode Time Fraction for each Simulated HD Combination Long-Haul
Duty Cycle 90
Figure 2-22: Brake-specific NOx emissions by simulated cycle for HHD diesel engines with
NOx FEL of 0.20 g/bhp-hr 91
Figure 2-23: Distance-specific NOx emissions by simulated cycle for HHD diesel engines with
NOx FEL of 0.20 g/bhp-hr 91
Figure 2-24: Number of windows in each bin from 168 HDUIT shift days, sorted in order (the
first 30% are shown). The black lines indicate 2400 windows and 1200 windows. Note some
HDUIT tests have multiple bins containing fewer than 2400 windows 96
Figure 2-25: Range of NOx rates in windows in each bin, using data from 93 HDIUT results in
the HHD, 0.2 FEL category. The heavy line is the median value from all tests, with error bars
represent the 25th and 75th percentile of the data 98
Figure 2-26: Estimated Projected Ford E-450 ORVR Results based on Extrapolation 103
Figure 2-27: Estimated Projected Isuzu ORVR Results based on Extrapolation 103
Figure 3-1: Developmental Cummins X15 Engine equipped with individual cylinder deactivation
undergoing engine dynamometer testing as part of the CARB Stage 3 research at SwRI.... 109
Figure 3-2: Schematic layout (not to scale) of the dual-SCR EAS tested as part of the CARB
Stage 3 research at SwRI 110
Figure 3-3: Developmental EAS with light-off SCR installed in engine dynamometer test cell at
SwRI (upper left, upper right) and details of the downstream, single "box" unit (lower left,
lower right) Ill
Figure 3-4: The average pressure drop across the DPF on the SET for the degreened
aftertreatment and the equivalent of 290,000 miles of operation aftertreatment 116
Figure 3-5: Schematic layout (not to scale) of the dual-SCR EAS tested as part of the EPA Stage
3 research at SwRI 116
Figure 3-6: CARB Southern Route Cycle 118
Figure 3-7: Grocery Delivery Truck Cycle 119
Figure 3-8: Drayage Truck Cycle 119
xxviii
-------
Figure 3-9: Euro-VI ISC Cycle 120
Figure 3-10: ACES 4-hour Cycle 120
Figure 3-11: EAS with close-coupled light-off SCR 123
Figure 3-12: EAS with light-off SCR integrated into an under-cab mounting position. This
system is designed to be installed in a Navistar Daycab which is shown in the upper right. 124
Figure 3-13: EPA developmental MY2018 Cummins X15 Heavy HDE 125
Figure 3-14: Variable reluctance sensor for valve position measurement. The final installation
will include valve position measurement at the valve bridges for each of the six cylinders. 125
Figure 3-15: Example of engine-speed, engine load, and resulting SCR inlet temperature used
over the DAAAC 126
Figure 3-16: FTP composite and SET NOx emissions results including IRAF for the EPA Stage
3 developmental engine and emissions control system versus equivalent miles of operation.
130
Figure 3-17: LLC NOx emissions results including IRAF for the EPA Stage 3 developmental
engine and emissions control system versus equivalent miles of operation 130
Figure 3-18: FTP-75, Cold start three phase test cycle 144
Figure 3-19: EPA Highway Fuel Economy Cycle 145
Figure 3-20: EPA 4 phase LA92 test cycle 146
Figure 3-21: Super cycle, GEM greenhouse gas cycle 147
Figure 3-22: FTP-75 Cold start cumulative HC comparison 148
Figure 3-23: FTP-75 Cold start cumulative NOx comparison 148
Figure 3-24: FTP-75 Catalyst light-off time comparison 149
Figure 3-25: Extended idle catalyst cool down comparison 150
Figure 3-26: Comparison of emissions performance for MY 2019 certified HD SI engines;
Overall Average (grey dash) includes data from all six engines, Subset Average (yellow dash)
includes data from the three engines with the best NOx performance (Cert E4-E6) 153
Figure 3-27: Comparison of operating modes in the fuel mapping-based SET 156
Figure 3-28: Engine RPM Down-Speeding FTP CO Comparison 161
Figure 3-29: Engine RPM Down-Speeding SET CO Comparison 162
Figure 4-1: Important Factors Involved in Seeing a Scenic Vista (Malm, 2016) 191
Figure 4-2: Mandatory Class I Federal Areas in the U.S 192
Figure 4-3: Nitrogen and Sulfur Cycling, and Interactions in the Environment 195
Figure 5-1: Percent change in in-use emission rates for 2010 standard (0.2 g/hp-hr) compliant
HHD vehicles, compared to the percent change in the 2010 duty-cycle standard across
MOVES operating modes 215
Figure 5-2: Duty-cycle-based running NOx emissions, ERduty in use, for HHD diesel for the
control scenarios 218
Figure 5-3: Base NOx rates and off-cycle NOx standard compliant emission rates for HHD diesel
223
XXIX
-------
Figure 5-4: Comparison of Running NOx emission rates for diesel-fueled HHD compliant with
the MY2030 proposed Option 1 duty-cycle and off-cycle standards 224
Figure 5-5: Estimated zero-mile NOx emission rates for HHD diesel vehicles due to the proposed
Option 1, Option 2 and Alternative duty-cycle and off-cycle standards 225
Figure 5-6: Methodology to model the effects of tampering and mal-maintenance (T&M) on
emission rates according to warranty and useful life 227
Figure 5-7: NOx g/mile emission rates estimated from MOVES for HHD diesel long-haul
combination trucks for the model year 2027 fleet across vehicle age for the baseline and
control scenarios 233
Figure 5-8: NOx g/mile emission rates estimated from MOVES for HHD diesel long-haul
combination trucks for the model year 2031 fleet across vehicle age for the baseline and
control scenarios 234
Figure 5-9: Estimated relationship between the HHD NOx 12-hour cold-start and the composite
FTP NOx standards 236
Figure 5-10: Duty-cycle-based NOx start emissions for HHD Diesel for the baseline, proposed,
and alternative control scenarios 237
Figure 5-11: Duty-cycle-based running NOx emission rates for LHD gasoline for the control
scenarios 242
Figure 5-12: LHD gasoline Duty-cycle-based running THC emission rates for LHD gasoline for
the control scenarios 243
Figure 5-13: Modeling process of onroad emissions as part of the input for air quality modeling
248
Figure 5-14: Estimated zero-mile emission rates for LHD45 diesel vehicles due to the proposed
and alternative duty-cycle and off-cycle standards 255
Figure 5-15: Estimated zero-mile emission rates for MHD diesel vehicles due to the proposed
and alternative duty-cycle and off-cycle standards 256
Figure 5-16: National NOx Emission Reductions from Heavy-duty Vehicles in Calendar Years
2030, 2040, and 2045 — Proposed Option 1 Program Reductions from the Baseline for Each
Fuel Type Category by Emission Process 257
Figure 5-17: National NOx Emission Reductions from Heavy-duty Vehicles in Calendar Years
2030, 2040, and 2045 — Proposed Option 1 Program Reductions from the Baseline for Each
Fuel Type Category by HD Regulatory Class 257
Figure 5-18: National VOC Emission Reductions from Heavy-duty Vehicles in Calendar Years
2030, 2040, and 2045 — Proposed Option 1 Program Reductions from the Baseline for Each
Fuel Type Category by Emission Process 258
Figure 5-19: National VOC Emission Reductions from Heavy-duty Vehicles in Calendar Years
2030, 2040, and 2045 — Proposed Option 1 Program Reductions from the Baseline for Each
Fuel Type Category by HD Regulatory Class 258
Figure 5-20: National Exhaust PM2.5 Emission Reductions from Heavy-duty Vehicles in
Calendar Years 2030, 2040, and 2045 — Proposed Option 1 Program Reductions from the
Baseline for Each Fuel Type Category by Emission Process 259
XXX
-------
Figure 5-21: National Exhaust PM2.5 Emission Reductions from Heavy-duty Vehicles in
Calendar Years 2030, 2040, and 2045 — Proposed Option 1 Program Reductions from the
Baseline for Each Fuel Type Category by HD Regulatory Class 259
Figure 5-22: National CO Emission Reductions from Heavy-duty Vehicles in Calendar Years
2030, 2040, and 2045 — Proposed Option 1 Program Reductions from the Baseline for Each
Fuel Type Category by Emission Process 260
Figure 5-23: National CO Emission Reductions from Heavy-duty Vehicles in Calendar Years
2030, 2040, and 2045 — Proposed Option 1 Program Reductions from the Baseline for Each
Fuel Type Category by HD Regulatory Class 260
Figure 5-24: Comparison of Calendar Year 2045 Onroad Heavy-Duty NOx Emissions from
Different Engine Operational Process for the Baseline, Proposed Option 1, and Proposed
Option 2 Standards 261
Figure 5-25: National Heavy-duty Vehicle NOx Emissions (Annual US Tons) For Calendar
Years Between 2027 and 2045 262
Figure 5-26: National Heavy-Duty Vehicle VOC Emissions (Annual US Tons) For Calendar
Years Between 2027 and 2045 263
Figure 5-27: National Heavy-duty Vehicle PM2.5 (Exhaust Only) Emissions (Annual US Tons)
For Calendar Years Between 2027 and 2045 264
Figure 5-28: National Heavy-Duty Vehicle CO Emissions (Annual US Tons) For Calendar Years
Between 2027 and 2045 265
Figure 5-29: National Heavy-duty Vehicle NOx Emission Reductions (Annual US Tons)
Relative to the Baseline Case For Omnibus Nationwide Scenarios (Symbols) As Compared
with EPA Proposed Option 1 (Line) 278
Figure 6-1: 8-Hour Ozone Nonattainment Areas (2008 Standard) 280
Figure 6-2: 8-Hour Ozone Nonattainment Areas (2015 Standard) 280
Figure 6-3: Counties Designated Nonattainment for PM2.5 (1997, 2006, and/or 2012 standards)
282
Figure 6-4: Map of the CMAQ 12 km modeling domain (noted by the purple box) 285
Figure 6-5: Projected Change in 8-hour Ozone Design Values in 2045 due to Proposed Rule . 289
Figure 6-6: Projected Change in Annual PM2.5 Design Values in 2045 due to Proposed Rule . 292
Figure 6-7: Projected Change in 24-hour PM2.5 Design Values in 2045 due to Proposed Rule. 295
Figure 6-8: Projected Absolute Change in Annual Ambient N02 Concentrations in 2045 299
Figure 6-9: Percent Change in Annual Ambient NO2 Concentrations in 2045 299
Figure 6-10: Absolute Change in Annual Ambient CO Concentrations in 2045 300
Figure 6-11: Percent Change in Annual Ambient CO Concentrations in 2045 301
Figure 6-12: Changes in Ambient Benzene Concentrations in 2045 due to Proposed Rule:
Absolute Changes in |ig/m3 (left) and Percent Changes (right) 302
Figure 6-13: Changes in Ambient Naphthalene Concentrations in 2045 due to Proposed Rule:
Absolute Changes in |ig/m3 (left) and Percent Changes (right) 302
Figure 6-14: Changes in Ambient Acetaldehyde Concentrations in 2045 due to Proposed Rule:
Absolute Changes in |ig/m3 (left) and Percent Changes (right) 303
XXXI
-------
Figure 6-15: Changes in Ambient Formaldehyde Concentrations in 2045 due to Proposed Rule:
Absolute Changes in |ig/m3 (left) and Percent Changes (right) 303
Figure 6-16: Absolute Change in Annual Deposition of Nitrogen in 2045 304
Figure 6-17: Percent Change in Annual Deposition of Nitrogen in 2045 305
Figure 6-18: Distributional maps of populated 12km grid cells across the contiguous United
States in 2045. Darker areas represent the location of grid cells within the highest 5 percent of
baseline concentrations for (a) PM2.5 and (b) ozone 307
Figure 7-1: Costs Relative to First Year Costs using Different Seed Volume Factors 321
Figure 7-2: Repair & Maintenance Cost/Mile Curve (2018 dollars) 345
Figure 7-3: Emission Repair Cost/Mile Curve (2018 dollars) 347
Figure 7-4: Emission Repair Cost Curve in the Baseline Case (dashed line) and the proposed
Option 1 (solid line) for a Short-haul Combination Truck equipped with a HHD Class-8
Diesel Engine 348
Figure 8-1: Stylized Relationship between the PM2.5 Concentrations Considered in Epidemiology
Studies and our Confidence in the Estimated PM-related Premature Deaths 390
Figure 8-2: Estimated Percentage of PM2.5-Related Deaths (Turner et al. 2016) and Number of
Individuals Exposed (30+) by Annual Mean PM2.5 Level in 2045 391
Figure 11-1: Secondary Vehicle Manufacturers, Estimated Impacts as a Percent of Annual
Revenues 432
XXXll
-------
Executive Summary
The Environmental Protection Agency (EPA) is proposing changes to heavy-duty highway
engine and vehicle provisions. This draft RIA is generally organized to provide overall
background information, methodologies, and data inputs, followed by results of the various
analyses. A summary of each chapter of the draft RIA follows.
Chapter 1 describes key technologies that manufacturers could use to meet more stringent
emissions standards for oxides of nitrogen (NOx), particulate matter (PM), hydrocarbons (HC),
and carbon monoxide (CO). The chapter introduces technologies specific to compression-
ignition engines and spark-ignition engines, and also discusses fuel considerations, advanced
powertrain technologies, and emission monitoring technologies that may apply across engine
types.
Chapter 2 describes the existing test procedures as well as the development process for the
test procedures being proposed for spark and compression ignition engine compliance. This
includes the determination of emissions from both engines and hybrid powertrains as well as the
development of new duty cycles.
Chapter 3 describes the technology feasibility demonstration programs, including engine
technologies and emission control strategies for reducing NOx, PM, NMHC, and CO. The
technologies presented represent potential ways that the industry could meet the proposed
stringency levels, and they provide the basis for the technology costs and benefits analyses.
Chapter 4 presents a discussion of the health effects associated with exposure ambient
concentrations of ozone, PM, NO2, CO, and air toxics. The discussion of health impacts is
mainly focused on describing the effects of air pollution on the population in general.
Additionally, children are recognized to have increased vulnerability and susceptibility related to
air pollution and other environmental exposures; this and effects for other vulnerable and
susceptible groups are discussed. The chapter also discusses the environmental effects associated
with pollutants affected by this proposed rulemaking, specifically PM, ozone, NOx and air
toxics.
Chapter 5 presents our analysis of the national emissions impacts of the proposed and
alternative standards for calendar years 2027 through 2045. In this chapter we quantify emissions
from NOx, volatile organic compounds (VOC), PM2.5, CO, and others. The onroad national
emissions inventories were estimated using an updated version of EPA's Motor Vehicle
Emission Simulator (MOVES) model. Table ES-1 summarizes the projected reductions in heavy-
duty emissions from the proposed Options 1 and 2 in 2045. In addition to describing the national
emission inventories, this chapter describes the methods used to estimate the spatially and
temporally-resolved emission inventories used to support the air quality modeling analysis
documented in Chapter 6.
1
-------
Table ES-1: Projected Heavy-Duty Emission Reductions in 2045 from the Proposed Options 1 and 2
Standards
Pollutant
Percent Reduction in Highway Heavy-duty Emissions
Proposed Option 1
Proposed Option 2
NOx
61%
47%
Primary PM2 5
26%
24%
VOC
21%
20%
CO
17%
16%
Chapter 6 presents information on air quality, including a discussion of current air quality,
details related to the methodology used for the air quality modeling analysis, and results from the
air quality modeling analysis. Reductions in emissions of NOx, VOC, PM2.5, and CO from the
proposed standards are projected to lead to decreases in ambient concentrations of ozone, PM2.5,
NO2, and CO. Specifically, the proposed standards would significantly decrease ozone design
value concentrations across the country, with a population-weighted average decrease of 2 ppb in
2045. Ambient PM2.5, NO2 and CO concentrations are also predicted to improve in 2045 as a
result of the proposed rule. The largest predicted improvements in both ozone and PM2.5 are
estimated to occur in areas with the worst baseline air quality, where a substantially larger
number of people of color are expected to reside. The emission reductions provided by the
proposed standards would be important in helping areas attain the National Ambient Air Quality
Standards (NAAQS) and prevent future nonattainment. In addition, the proposed standards are
expected to result in improvements in nitrogen deposition and visibility but are predicted to have
relatively little impact on ambient concentrations of air toxics.
Chapter 7 presents estimates of the costs associated with the emissions-reduction
technologies that manufacturers could add in response to the proposed standards. We present
these not only in terms of the upfront technology costs per engine as presented in Chapter 3 of
this draft RIA, but also how those costs would change in the years following implementation.
We present the costs associated with the proposed program elements of extended regulatory
useful life and warranty. These technology costs are presented in terms of direct manufacturing
costs and associated indirect costs such as warranty and research and development (R&D). The
analysis also includes estimates of the possible operating costs associated with the proposed
changes—the addition of new technology and extension of warranty and useful life periods. All
costs are presented in 2017 dollars consistent with AEO 2018 unless noted otherwise.
Table ES-2 presents the technology costs, operating costs and the sum of the two for each
proposed option in 2045.
2
-------
Table ES-2: Total Program Costs: Undiscounted Annual Costs in 2045 and Annualized Costs through 2045 at
3% and 7% Discount Rates (Billions of 2017 dollars)
Total
Technology
Costs
Total
Operating
Costs
Sum
Proposed Option 1
2045 Annual
$1.6
$0.72
$2.3
Proposed Option 2
2045 Annual
$1.4
$1.5
$2.9
Proposed Option 1
Present Value, 3%
$23
$4.2
$27
Proposed Option 2
Present Value, 3%
$21
$9.2
$30
Proposed Option 1
Present Value, 7%
$17
$2.6
$19
Proposed Option 2
Present Value, 7%
$15
$5.5
$21
Proposed Option 1
Annualized, 3%
$1.6
$0.29
$1.9
Proposed Option 2
Annualized, 3%
$1.4
$0.64
$2.1
Proposed Option 1
Annualized, 7%
$1.6
$0.25
$1.9
Proposed Option 2
Annualized, 7%
$1.5
$0.54
$2.0
Chapter 8 describes the methods used to estimate health benefits from reducing
concentrations of ozone and PM2.5. For the proposed rulemaking, we have quantified and
monetized health impacts in 2045, representing projected impacts associated with a year when
the program would be fully implemented and when most of the regulated fleet would have turned
over. We also discuss unquantified benefits associated with the standards that, if quantified and
monetized, would increase the total monetized benefits. Overall, we estimate that the proposed
program would lead to a substantial decrease in adverse PM2.5- and ozone-related health impacts
in 2045. Table ES-3 presents our estimates of total monetized benefits for proposed Options 1
and 2.
Table ES-3: 2045 Annual Value, Present Value and Equivalent Annualized Value of Benefits of the Proposed
Options 1 and 2 (billions, 2017$)a b
Proposed Option 1
Proposed Option 2
3%
7%
3%
7%
Discount
Discount
Discount
Discount
2045
$12-$33
$10 -$30
$9.1 -$26
$8.2 - $23
Present Value (2027-2045)
$88 - $250
$52-$150
$71 - $200
$41 -$120
Annualized Value
$6.0 -$17
$4.7-$13
$5.0-$14
$4.0-$11
a All benefits estimates are rounded to two significant figures; numbers may not sum
due to independent rounding. The range of benefits in this table are two separate
estimates and do not represent lower- and upper-bound estimates, though they do
reflect a grouping of estimates that yield more and less conservative benefits totals.
The benefits in 2045 are presented in annual terms and are not discounted. However,
all benefits in the table reflect a 3 percent and 7 percent discount rate used to account
for cessation lag in the valuation of avoided premature deaths associated with long-
term exposure.
b The benefits associated with the standards presented here do not include the full
complement of health, environmental, and climate-related benefits that, if quantified
and monetized, would increase the total monetized benefits.
Chapter 9 compares the estimated range of total monetized health benefits to total costs
associated with the proposed criteria pollutant program Options 1 and 2. This chapter also
3
-------
presents the range of monetized net benefits (benefits presented in Chapter 8 minus costs
presented in Chapter 7) associated with the same scenarios (see Table ES-4). EPA expects that
implementation of the proposed rule would provide society with a substantial net gain in welfare,
notwithstanding the health and other benefits we were unable to quantify (see draft RIA Chapter
8.8 for more information about unquantified benefits).A
Table ES-4: 2045 Annual Value, Present Value and Equivalent Annualized Value of Costs, Benefits and Net
Benefits of the Proposed Options 1 and 2 (billions, 2017$)a b
Proposed Option 1
Proposed Option 2
3%
7%
3%
7%
Discount
Discount
Discount
Discount
Benefits
$12 -$33
$10-$30
$9.1 -$26
$8.2 - $23
2045
Costs
$2.3
$2.3
$2.9
$2.9
Net Benefits
$9.2-$31
$8.1 -$28
$6.2 - $23
$5.3 -$21
Benefits
$88 - $250
$52-$150
$71 - $200
$41-$120
Present Value
Costs
$27
$19
$30
$21
Net Benefits
$61 - $220
$33 -$130
$41 -$170
$21 -$96
Equivalent
Benefits
$6.0-$17
$4.7-$13
$5.0-$14
$4.0 -$11
Annualized
Costs
$1.9
$1.9
$2.1
$2.0
Value
Net Benefits
$4.1 -$15
$2.9-$12
$2.9-$12
$2.0 - $9.3
a All benefits estimates are rounded to two significant figures; numbers may not sum due
to independent rounding. The range of benefits (and net benefits) in this table are two
separate estimates and do not represent lower- and upper-bound estimates, though they do
reflect a grouping of estimates that yield more and less conservative benefits totals. The
costs and benefits in 2045 are presented in annual terms and are not discounted. However,
all benefits in the table reflect a 3 percent and 7 percent discount rate used to account for
cessation lag in the valuation of avoided premature deaths associated with long-term
exposure.
b The benefits associated with the standards presented here do not include the full
complement of health, environmental, and climate-related benefits that, if quantified and
monetized, would increase the total monetized benefits.
Chapter 10 provides an economic analysis of the impacts of the proposed standards on
vehicle sales and employment. This proposed rulemaking is considered economically significant,
because it is expected to have an annual impact on the economy of $100 million or more but is
not expected to have measurable inflationary or recessionary effects. This chapter presents a
peer-reviewed analysis to develop a relationship between estimated changes in vehicle price due
to a new regulation and corresponding changes in vehicle sales (i.e., pre- and low-buy
elasticities). We suggest an approach to quantify potential impacts on vehicle sales due to new
emission standards; we also provide an example of how the results could be applied to the final
regulatory analysis for this rule in draft RIA Chapter 10.1. Our example results for proposed
Option 1 suggest pre- and low-buy for Class 8 trucks may range from zero to approximately two
percent increase in sales over for a period of up to 8 months before the 2031 standards begin
(pre-buy), and a decrease in sales from zero to approximately two percent over a period of up to
12 months after the 2031 standards begin (low-buy). The employment assessment focuses on the
motor vehicle manufacturing and the motor vehicle parts manufacturing sectors, with some
A EPA does not expect the omission of unquantified benefits to impact the Agency's evaluation of regulatory options
since unquantified benefits generally scale with the emissions impacts of the proposed Options.
4
-------
assessment of impacts on additional sectors likely to be most affected by the proposed standards.
The employment assessment includes EPA's qualitative and quantitative estimates of the partial
employment impacts of this proposed rule on regulated industries and an examination of
employment impacts in some closely related sectors.
Chapter 11 presents our analysis of the potential impacts of the proposal on small entities
that would be subject to the highway heavy-duty engine and vehicle provisions of this proposed
rule. These are: heavy-duty vehicle manufacturers, heavy-duty secondary vehicle manufacturers,
and heavy-duty alternative fuel engine converters. Other entities that would be subject to the rule
are either not small (e.g., engine and incomplete vehicle manufacturers) or are not expected to
incur any burden from the proposed rule (e.g., in sectors other than highway heavy-duty engines
and vehicles). Our analysis estimates that no small entities would experience an impact of 3% or
more of their annual revenue as a result of our proposal.
5
-------
Chapter 1 Technology to Control Emissions from Heavy-Duty Engines
This chapter describes key technologies that manufacturers could use to meet more stringent
emissions standards for oxides of nitrogen (NOx), particulate matter (PM), hydrocarbons (HC),
and carbon monoxide (CO). The chapter introduces technologies specific to compression-
ignition engines and spark-ignition engines, and also discusses fuel considerations, advanced
powertrain technologies, and emission monitoring technologies that may apply across engine
types.
1.1 Compression-Ignition Engine Technologies
The following sections describe the compression-ignition engine technologies that we are
considering for reducing criteria pollutant emissions as part of this proposed rulemaking. Many
of the technologies are described with respect to diesel fuel, but they are expected to be broadly
applicable to all fuels used in compression-ignition engines. Our compression-ignition engine
feasibility demonstration for this proposal is based on some of the technologies presented in this
section.
Chapter 3 of this draft RIA describes the existing and proposed test procedures for diesel-
ignition engine certification. Chapter 4 describes the compression-ignition engine feasibility
demonstration program, including a description of the specific technology package we are
evaluating, and the effectiveness of those technologies over our current and proposed test
procedures, and our projected cost of those technologies.
1.1.1 Current Catalyst Technologies
This section addresses technologies that, based on our current understanding, would be
available in the 2024 to 2030 timeframe to reduce emissions and ensure robust in-use
compliance. The following discussion introduces the technologies and emission reduction
strategies we are considering for the proposed rulemaking, including thermal management
technologies that can be used to better achieve and maintain adequate catalyst temperatures, and
next generation catalyst configurations and formulations to improve catalyst performance across
a broader range of engine operating conditions.
Modern diesel engines rely heavily upon catalytic exhaust aftertreatment systems (EAS) to
meet exhaust emission standards. Current (MY2018-2020) heavy-duty diesel EAS consist of a
diesel oxidation catalyst (DOC) followed by a catalyzed diesel particulate filter (CDPF), a urea
injector, a urea mixer or other decomposition component, and then one or more selective
catalytic reduction (SCR) catalyst substrates (Figure 1-1, Figure 1-2). Such systems are capable
of reducing PM emissions by greater than 95% under most operating conditions and are capable
of reducing NOx emissions by 90 to 98% at exhaust temperatures above 250 °C.
Unreacted ammonia downstream of the SCR is typically referred to as "ammonia slip". An
ammonia slip catalyst can be zone-coated onto the outlet of the rearmost SCR catalyst substrate
(the case for most LHDDE and MHDDE and some HHDDE applications) or can be coated onto
a separate catalyst substrate (some HHDDE applications) and uses platinum-group-metal (PGM)
exchanged zeolites to promote reaction of ammonia remaining downstream of the SCR catalysts.
Ammonia is an important air toxic compound and can also contribute to secondary PM
formation. The use of closed-loop feedback electronic control of urea dosing using zirconia NOx
6
-------
sensors for NOx feedback and the use of an ammonia slip catalyst (ASC) together can reduce
ammonia emissions from modern EAS-equipped heavy-duty diesel engines to less than 4
mg/bhp-hr.1 Some LHDDE applications using chassis dynamometer certification place the urea
injector and SCR between the DOC and CDPF or combine SCR and CDPF functionality into one
catalyst, sometimes referred to as selective catalytic reduction on filter (SCRF).
DOC
CDPF
Urea Injector
SCR + ASC
(zone coated)
Mixer
SCR
lexhaust flow
17
Same components integrated
into box-style muffler
Figure 1-1: Functional schematic showing relative positioning of exhaust emission control components
arranged within an in-line exhaust system (top) and integrated into a box-style system (bottom).
7
-------
Figure 1-2: Integrated series heavy-duty truck exhaust emission control systems from Cummins Emission
Solutions (top) and box-style system from Eberspacher (bottom), with cut-away showing some of the internal
components (bottom right).®
The DOC, SCR, and ASC typically use cordierite ceramic flow-through monolithic substrates
that are wash-coated with active materials. The CDPF uses a wall-flow substrate made of either
cordierite, silicon carbide (SiC), or aluminum titanate (AhTiOs) for exhaust filtration (or
"trapping") of particulate matter that is coated with active materials. Alternating cells of the
wall-flow substrate are blocked, forcing the exhaust to flow through the porous substrate wall.
The particulate matter, consisting primarily of elemental carbon soot, is filtered from the exhaust
flow onto and within the wall of the CDPF and can then be oxidized to CO2 using either passive
regeneration with nitrogen dioxide (NO2) or active regeneration with excess oxygen in the
exhaust. Passive regeneration of the CDPF depends on oxidation of a fraction of nitric oxide
(NO) emissions in the exhaust to NO2. Soot oxidation using NO2 occurs at exhaust temperatures
" Disclaimer: Any mention of trade names, manufacturers or products does not imply an endorsement by the United
States Government or the U.S. Enviromnental Protection Agency. EPA and its employees do not endorse any
commercial products, services, or enterprises.
8
-------
of approximately 250 °C, thus it does not require external heat addition to the exhaust under
most operating conditions. Active regeneration using excess oxygen in the exhaust occurs at
exhaust temperatures above 500 °C - 600 °Cc, and thus requires adding heat to the exhaust.
This can be accomplished using one of several different approaches:
• Late, in-cylinder, post-injection of fuel after the primary combustion event and
subsequent heat addition from the exothermic reaction of the excess fuel over the
DOC and CDPF
• Direct injection of diesel fuel into the exhaust, with exothermic reaction of the fuel
over the DOC and CDPF
• Use of an exhaust-integrated, external combustion burner system
Selective catalytic reduction (SCR) reduces nitrogen oxides NOx (consisting of both NO and
NO2) to N2 and water by using ammonia (NH3) as the reducing agent. The SCR catalyst coatings
used for post-2010 model-year heavy-duty diesel applications in the U.S. are typically copper
(Cu) exchanged or iron (Fe) exchanged zeolites, (e.g., Fe-ZSM-5), and most SCR coatings for
recent (MY2018-2020) applications are one of several different Cu and/or Fe-exchanged
chabazite zeolite structures (e.g., Cu-SSZ-13). The method for supplying ammonia to the SCR
catalyst is to inject a mixture of 32.5% urea in water solution into the exhaust stream. In the
presence of high temperature exhaust gasses (> 180 - 250 °C)D, the urea decomposes to form
both NH3 and iso-cyanic acid (HNCO) by thermolysis, with subsequent hydrolysis of the HNCO
to form additional NFb:
CO(NH2)2 —> NH3 + HNCO
HNCO + H2O —> NH3 + CO2
The "standard SCR reaction" of NO (the predominant NOx species from diesel combustion)
over transition-metal zeolite or vanadium SCR catalysts can be represented as:
4NH3 + 4NO + O2 —* 4N2 + 6H2O
Improved reaction kinetics can be achieved at low exhaust temperatures (<300 °C) by
oxidizing a portion of the NO in exhaust using a platinum-group-metal (PGM) coated diesel
oxidation catalyst (DOC) to achieve a 1:1 molar ratio of NO:N02. The resulting "fast SCR
reaction" is:
2NH3 + NO + NO2 —> 2N2 + 3H20
c The temperature at which soot oxidation temperature occurs differs depending on the catalytic coating used on the
CDPF.
D Note that the urea decomposition temperature is dependent upon spray atomization, exhaust flow turbulence, and
exhaust flow rate.
9
-------
An NO2 SCR reaction with limited NO availability also occurs, but at a significantly slower
reaction rate than the standard SCR reaction and is sometimes referred to as the "slow SCR
reaction":
8NH3 + 6 NO2 —> 7N2 + I2H2O
The details of SCR reactions for NOx reduction over transition-metal zeolite SCR can be
better represented as a series of interrelated reactions that are part of a more complex redox
cycle, such as the one proposed by Rudolf and Jacob.2
Urea dosing control takes into account a number of different factors, including:
• The stoichiometry of NOx reduction by NH3 (1:1 or 4/3:1 molar ratio)
• Molar ratio of N0:N02 at the inlet of the SCR catalyst
• The amount of NH3 stored and released from the zeolites
0 Thermal desorption of stored NH3 can allow NOx reduction to occur at exhaust
temperatures that are often too low for urea injection and decomposition
• The degree of urea/exhaust mixture preparation of the system design
0 Droplet formation and evaporation
0 Induced turbulent mixing of aqueous urea and exhaust to aid droplet breakup
• The efficiency of urea decomposition to NH3 (> 95-98% at >250 °C is typical for
modern injector/mixer designs)
• The probability forming solid deposits at low exhaust temperatures from partially
decomposed urea
0 Urea injection at exhaust temperatures below approximately 180 to 200 °C can
result in significantly increased deposit formation depending on mixture
preparation and other factors
0 Urea injector fouling can occur from deposit build up on the urea injector tip and
other exhaust system surfaces
0 Deposits can temporarily deactivate active catalytic surfaces, requiring higher
temperature operation in order to remove the deposits
Copper (Cu) exchanged chabazite zeolites such as Cu-SSZ-13 have demonstrated good
hydrothermal stability, good low temperature performance, and represent a large fraction of the
transition-metal zeolite SCR catalysts used in heavy-duty applications.3 Improvements to both
the coating processes and the substrates onto which the zeolites are coated have improved the
low-temperature and high-temperature NOx conversion, improved selectivity of NOx reduction
to N2 (i.e., reduced selectivity to N2O), and improved the hydrothermal stability. Improvements
in SCR catalyst coatings over the past decade have included:4'5'6'7'8
• Increased washcoat thickness
10
-------
• Optimization of Silicon/Aluminum (Al) and Cu/Al ratios
• Increased Cu content and Cu surface area
• Optimization of the relative positioning of Cu2+ ions within the zeolite structure
• The introduction of specific co-cations
• Co-exchanging of more than one type of metal ion into the zeolite structure
In the absence of more stringent NOx standards, these improvements have been realized
primarily as reductions in SCR system volume, reductions in system cost, and improvements in
durability since the initial introduction of metal-exchanged zeolite SCR in MY2010. Sales-
weighted average engine-displacement-specific catalyst volumes for MY2019 MHDDE and
HHDDE are shown in Table 1-1.
Table 1-1: Engine-displacement specific catalyst substrate volume for MY2019 MHDDE and HHDDE.
Component
Mil DDI Specific
Volume*
IIIIDDI Specific
Volume*
DOC
0.61
0.74
CDPF
1.39
1.49
SCR
2.11
2.24
ASC
0.38
0.40
Notes:
*Specific Volume = (catalyst total substrate volume) / (engine's piston swept displacement)
NOx reductions greater than 95% are possible with modern SCR systems over a broad range
of operating conditions and at relatively high hours of operation, however SCR functionality is
particularly reduced at lower exhaust temperatures due to difficulties with low-temperature urea
decomposition and due to slower SCR reaction kinetics (Figure 1-3). In the figure, the initial
data point at 140 °C SCR inlet temperature reflects NOX reduction with stored ammonia only
(no urea injection). The hours in the legend represent of operation over an accelerated aging
cycle that included both thermal and chemical effects. The 1,430 hours of aging represented
approximately 8,000 hours of equivalent engine operation. Reduced oxidation of NO to NO2
over the DOC and DPF at low exhaust temperatures (e.g., 200 to 250 °C) reduces the ability to
take advantage of the "fast SCR reaction". As previously mentioned, current SCR systems limit
urea injection to temperatures above 180 °C to 200 °C to prevent urea injector and catalyst
deposits. NOx reduction reactions at temperatures below approximately 200 °C are thus reliant
on use of NH3 stored within the zeolite structure. During extended operation at low exhaust
temperatures, stored NH3 is eventually depleted and if exhaust temperatures cannot be increased
sufficiently to allow initiation of urea injection and effective decomposition to NH3, then NOx
reduction eventually ceases.
11
-------
SCR Space Velocity (hr-1) » 6,500 7,000 13,000 17,000 25,000 33,000 39,000 48,0C
Figure 1-3: NOx reduction efficiency of an early, developmental Cu-zeolite SCR formulation relative to DOC
and SCR inlet temperatures and SCR space velocity (adapted from McDonald et al. 20119).
Compression-ignition engine exhaust temperatures are relatively low following cold starts,
during coasting downhill, during sustained idle, or at low vehicle speeds and during light load
operation. Technologies that accelerate warm-up from a cold-start and maintain catalyst
temperature above 200°C can help achieve further NOx reduction from SCR systems under those
types of operation. Technologies that improve urea decomposition to NH3 at temperatures below
200°C can also be used to reduce NOx emissions under cold start, coasting, light load, and low
speed conditions.
1.1.2 Catalyst Durability
The regulatory full-useful-life for HHDDE emissions compliance with the fully-phased-in
2007 Heavy-duty Standards is 435,000 miles, ten years, or 22,000 hours of operation. Zeolite-
based SCR systems have demonstrated high levels of NOx reduction efficiency at the end of
regulatory useful life. The aging mechanisms of diesel exhaust aftertreatment systems are
complex and include both chemical and hydrothermal changes. Aging mechanisms on a single
component can also cascade into impacts on multiple catalysts and catalytic reactions within the
system due to the interrelated nature of catalytic reactions over upstream components on other
aftertreatment components further downstream. Some aging impacts are fully reversible (i.e.,
conditions occur that can fully mitigate the aging impact). Other aging impacts are only partially
reversible, irreversible, or can only be reversed with some form of intervention (e.g., changes to
engine calibration to alter exhaust temperature and/or composition).
12
-------
DOCs undergo reversible aging due to adsorption of hydrocarbons, soot and sulfur species;
and irreversible aging due to phosphorus (P) poisoningE and thermal sintering of platinum group
metals (PGM) and other active materials. The catalytic materials on DPFs undergo similar aging
impacts, and also continuously accumulate metallic ash, which typically accumulates towards the
rear of the DPF channels. As accumulated ash migrates towards the front of the DPF channels,
exhaust backpressure increases. Heavy-duty diesel exhaust systems, particularly those of
HHDDE, are designed to allow CDPF removal for ash maintenance, often at approximately
250,000- to 350,000-mile intervals. Systems for LHDDE and MHDDE are typically designed
with sufficient CDPF capacity to forgo ash maintenance within the current regulatory
requirements for full useful life. Ash maintenance involves removal of the CDPF and
application of a either a dry cleaning process (e.g., reverse flushing of the CDPF with
compressed air) or a wet cleaning process (e.g. reverse flushing of the CDPF with water or with
a specific aqueous cleaning solution).
Aging of zeolite SCR is more complex. Hydrothermal aging of Cu-SSZ-13 SCR catalysts
impacts both catalyst acidity and NFb adsorption, transforms active Cu sites into less active
species, and causes Cu migration from exchanged positions within the zeolite structure and
subsequent formation of aggregated CuO.10 The severity of hydrothermal aging increases in the
presence of sulfur.11
Chemical poisoning of SCR can occur from fuel and lubricant contaminants, or via
degradation of upstream components. Sources of chemical poisoning include:
• Lubricant consumption
° Zinc dialkyldithiophosphate anti-wear, antioxidant, and corrosion inhibiting
additives
Phosphorus (P)
Zinc (Zn)
° Sulfur (S)
• Fuel
° S
° Trace contaminants from biodiesel blending (alkali metals, e.g. Na, K)
° Adsorption of hydrocarbon species into zeolite structure and subsequent blockage
of pores through soot formation12
• Migration of metals from upstream components
° PGM from the DOC and CDPF
° Transition metal (e.g. Fe, Cu) oxides from upstream SCR components or along
SCR substrates in series
Hydrothermal and chemical aging impacts on the DOC can also impact SCRNOx reduction,
particularly at low temperatures, via inhibition of NO to NO2 oxidation necessary for the fast
SCR reaction. The potential for future SCR durability improvements fall into the following
categories:
E The sources of P poisoning are from lubricating oil consumption and P-containing lubricating oil additives, such as
zinc dithiophosphate.
13
-------
• Designing excess capacity into the catalyst (e.g., increased catalyst volume, increased
catalyst cell density, increased active material content and surface area)
• Use of a small-volume initial "sacrificial" substrate to adsorb chemical catalyst
poisons upstream of the initial DOC or SCR substrate
• Continued improvements to zeolite materials
° Further optimization of Silicon/Aluminum (Al) and Cu/Al ratios
° Exchange of beneficial co-cations
° Co-exchanging of more than one type of transition metal into the zeolite structure
° Reducing pore size to inhibit HC adsorption and pyrolysis
• Direct hydrocarbon dosing downstream of the light-off SCR during active CDPF
regeneration to reduce exposure of the light-off SCR to hydrocarbons and fuel
contaminants
• Use of washcoat additives, changes to substrate porosity, and other improvements to
increase PGM dispersion, reduce PGM particle size, reduce PGM mobility and
reduce agglomeration within the DOC and CDPF washcoatings,
• Improvements to catalyst housings and substrate matting material to minimize
vibration and prevent exhaust gas leakage around the substrate
• Reducing SCR and DOC sectional density, either through increased porosity or
decreased cell wall thickness, thus lowering substrate mass and improving warm-up
characteristics.
• Adjusting engine calibration and emissions control system design to minimize
operation that would damage the catalyst (e.g., improved control of CDPF active
regeneration, increased passive CDPF regeneration, HC dosing downstream of initial
light-off SCR, direct temperature sensor feedback control of active regeneration and
chemical deSOx)
• Use of specific engine calibration strategies for chemical deSOx of SCR (e.g., high
temperature operation with urea dosing)13 to remove strongly-bound sulfur
compounds from zeolite SCR
• Diagnosis and prevention of upstream engine malfunctions that can potentially
damage exhaust aftertreatment components
Increased SCR catalyst capacity, along with incremental improvements to current zeolite
coatings would be primary strategies for improving NOx control over a longer regulatory useful
life requirement. SCR capacity can be increased by approximately 40 to 50% with the use of a
light-off SCR substrate combined with a downstream substrate with a moderate volume increase
and with moderately increased catalytic activity from continued incremental improvements to
chabazite and other zeolite coatings used for SCR. Total SCR volume would thus increase by
approximately 50% to 80% relative to today's systems. SCR capacity can also be increased in
the downstream SCR system using thin-wall (4 to 4.5 mil), high cell density (600 cells-per-
square-inch) substrates.
Chemical aging of the DOC, CDPF, and SCR can be reduced by the presence of an upstream
light-off SCR or use of a small "sacrificial" substrate to adsorb chemical poisons. Transport and
adsorption of sulfur (S), P, calcium (Ca), zinc (Zn), sodium (Na), and potassium (K) compounds
and other catalyst poisons are more severe for the initial catalyst within an emissions control
system and tend to reduce in severity for catalysts positioned further downstream. Chemical
deSOx strategies can be used to remove strongly-bound sulfur from zeolite SCR13. This involves
14
-------
creating a strongly reducing environment via dosing of urea in excess of the typical 1:1 NH3 to
NOx ratio at temperatures of approximately 500 °C to 550 °C. Further evolutionary
improvements to the DOC washcoating materials to increase PGM dispersion and reduce PGM
mobility and agglomeration are also anticipated for meeting increased useful life requirements.
The primary strategy for maintaining CDPF function to a longer useful life would be through
design of integrated systems that facilitate easier removal of the CDPF for ash cleaning at regular
maintenance intervals. Accommodation of CDPF removal for ash maintenance is already
incorporated into many existing diesel exhaust system designs.F Incremental improvements to
catalyst housings and substrate matting material are also expected to be necessary for all catalyst
substrates within the system. Integration into a box-muffler type system, is currently being used
by a number of manufacturers and this approach is expected to continue for all catalyst
components (except possibly for an initial close-coupled/light-off SCR) in order to improve
passive thermal management and improve access to the DPF for ash maintenance.
1.1.3 Improving SCR NOx Reduction at Low Exhaust Temperatures
The improvement of SCR NOx reduction under low-speed (<1200 rpm) light-load (< 5-bar
BMEP) conditions or immediately following cold starts will require improvements to both active
and passive thermal management of the EAS.
1.1.3.1 Active Thermal Management
Active thermal management involves using engine hardware and associated control systems
to maintain and/or increase exhaust temperatures. This can be accomplished through a variety of
means, including engine throttling, heated aftertreatment systems, and exhaust flow bypass
systems. Later combustion phasing can also be used for active thermal management.
Diesel engines operate at very low fuel-air ratios (i.e., with considerable excess air), and
particularly so at low load (<5-bar BMEP) conditions. This causes relatively cool exhaust to
flow through the exhaust system at low loads, which cools the catalyst substrates. This is
particularly the case at idle. It is also significant at moderate-to-high engine speeds with little or
no engine load, such as when a vehicle is coasting down a hill. Air flow through the engine can
be reduced by induction and/or exhaust throttling. All modern heavy-duty diesel engines are
equipped with an electronic throttle control (ETC) within the induction system and most are
equipped with a variable-geometry-turbine (VGT) turbocharger, and these systems can be used
to throttle the induction and exhaust system, respectively, at light-load conditions. However,
throttling reduces volumetric efficiency0, and thus has a trade-off relative to CO2 emissions and
fuel consumption.
Heat can be added to the exhaust and the EAS by burning fuel in the exhaust system or by
using electrical heating, both of which can increase the SCR efficiency. Burner systems use an
additional diesel fuel injector in the exhaust to combust fuel and create additional heat energy in
the exhaust flow. Electrically heated catalysts use electric current applied to a metal foil
monolithic structure in the exhaust to add heat to the exhaust flow. In addition, heated and
higher-pressure urea dosing systems improve the decomposition of urea at low exhaust
F Video by Eberspaecher demonstrating DPF removal for ash cleaning maintenance: https://youtu.be/lf_vysKbfaA
G Relative efficiency of the air-exchange process in an internal combustion engine
15
-------
temperatures and thus allow urea injection to occur at lower exhaust temperature (i.e., at
approximately 135 °C to 140 °C ). At light-load conditions with relatively high flow/low
temperature exhaust, considerable fuel energy or electric energy would be needed for these
systems. This would likely cause a considerable increase in CO2 emissions and fuel
consumption with conventional designs.
Exhaust flow bypass systems can be used to manage the cooling of exhaust during cold start
and low load operating conditions. For example, significant heat loss occurs as the exhaust gases
flow through the turbocharger turbine. Turbine bypass valves allow exhaust gas to bypass the
turbine and avoid this heat loss at low loads when turbocharger boost requirements are low. In
addition, an EGR flow bypass valve would allow exhaust gases to bypass the EGR cooler when
EGR cooling is not required, such as immediately following a cold start or under cold ambient
conditions. EGR cooler bypass is currently used in light-duty diesel and LHDDE applications.
Variable valve actuation (VVA) systems can also be used for active thermal management.
VVA includes a family of valvetrain designs that alter the timing and/or lift of the intake and
exhaust valves. Use of VVA can reduce pumping losses, increase specific power, and control
the level of residual gases in the cylinder.
VVA has been adopted in light-duty vehicles to increase an engine's efficiency and specific
power. It has also been used as a thermal management technology to open exhaust valves early
to increase heat rejection to the exhaust and heat up exhaust catalysts more quickly. This VVA
strategy, called early exhaust valve opening (EEVO), has been applied to the Detroit DD814 to
aid in CDPF regeneration, but a challenge with this strategy for maintaining aftertreatment
temperature is that it reduces cycle thermal efficiency, and thus can contribute to increased CO2
emissions.
Cylinder deactivation (CDA), late intake valve closing (LIVC), and early intake valve closing
(EIVC) are three VVA strategies that can also be used to reduce airflow through the exhaust
system at light-load conditions, and have been shown to reduce the CO2 emissions and fuel
consumption trade-off compared to use of the ETC and/or VGT for throttling.15'16'17'18
Since we are particularly concerned with catalyst performance at low loads, EPA will be
evaluating two valvetrain-targeted thermal management strategies that reduce air aspiration of
engines at light-load conditions (i.e., less than 3-4 bar BMEP): CDA and LIVC. Both strategies
force engines to operate at a higher fuel-air ratio in the active cylinders for a given load demand,
which increases exhaust temperatures, with the benefit of little or no fuel consumption increase
and with potential for fuel consumption decreases under some operating conditions. The key
difference between these two strategies is that CDA completely removes airflow from one or
more deactivated cylinders with the potential for exhaust temperature increases of up to 80 °C at
light loads, while LIVC reduces airflow from all cylinders with up to 40 °C hotter exhaust
temperatures.16'17'18
One of the challenges of CDA is that it requires proper integration with the rest of the
vehicle's driveline. This can be difficult in the vocational vehicle segment where an engine is
often sold by the engine manufacturer (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 use of CDA requires fine tuning of the engine calibration as the engine moves into
and out of deactivation to achieve acceptable noise, vibration, and harshness (NVH). Mitigation
16
-------
strategies include changes to driveline dampening, motor mount and/or chassis dampening, and
the use of dynamic CD A with individual cylinder deactivation control.
LIVC may provide emission reductions similar to fixed CD A, with the added benefits of no
significant NVH concerns. A production-level LIVC system may also be cost-competitive or
lower in cost compared to CD A, depending on how LIVC is implemented and depending on
specific engine valvetrain design characteristics.
1.1.3.2 Passive Thermal Management
Passive thermal management involves changes or modifications to component designs to
increase and maintain the exhaust gas temperatures without the use of active thermal
management. It is done primarily through insulation and/or reducing the mass of EAS and other
exhaust system components so that less exhaust energy input is required to reach catalyst light-
off temperatures and/or the exhaust temperatures at which urea dosing can commence.19 Passive
thermal management strategies generally have little to no impact on CO2 emissions or can
improve CO2 emissions if used to replace an active thermal management strategy. The use of
passive thermal management strategies for improving catalyst light-off in light-duty gasoline
applications has led to significant reductions in cold-start exhaust emissions.20 Passive thermal
management design elements can be equally applied to EAS systems used in heavy-duty
applications.
More specifically, using a smaller sized, initial SCR catalyst within the EAS with a high-
porosity, lower density substrate reduces its mass and reduces catalyst warmup time. Moving the
SCR catalyst nearer to the turbocharger outlet effectively reduces the available mass prior to the
SCR inlet, minimizing heat loss and reducing the amount of energy needed to warm components
up to normal operating temperatures. Reducing the mass of the exhaust system and insulating
between the turbocharger outlet and the inlet of the SCR system using an air-gap or other
insulation would reduce the amount of thermal energy lost through the walls. Close coupling of
catalysts is near ubiquitous in modern light-duty EAS. The use of air-gap construction is also
common in light-duty applications.
Dual-walled manifolds and exhaust pipes utilizing a thin inner wall and an air gap separating
the inner and outer wall may be used to simultaneously insulate the exhaust system and reduce
the thermal mass, minimizing heat lost to the walls and decreasing the time necessary to reach
operational temperatures after a cold start. Mechanical insulation applied to the exterior of
exhaust components, including exhaust catalysts, is readily available and can minimize heat loss
to the environment and help retain heat within the catalyst as operation transitions to lighter
loads and lower exhaust temperatures. Integrating the DOC, DPF, and SCR substrates into a
single exhaust assembly can also assist with retaining heat energy.
EPA is evaluating several passive thermal management strategies in the diesel technology
feasibility demonstration program, including a light-off SCR located closer to the exhaust turbine
(see draft RIA Chapter 1.1.3.3), use of an air-gap exhaust manifold and downpipe, use of lower
mass, high-porosity and/or thin-wall substrates for the DOC and light-off SCR substrates, and
use of an insulated and integrated single-box system for the DOC, DPF, and downstream
SCR/ASC. We will evaluate their combined ability improve catalyst warmup and maintain
higher exhaust temperatures under light-load conditions, which are anticipated to significantly
reduce NOx emissions during both in-use operation and over the regulatory engine dynamometer
17
-------
test cycles used for engine certification (see draft RIA Chapter 4.1 for detailed discussion of our
diesel technology demonstration programs).
1.1.3.3 Advanced SCR System Development
A recent development in SCR system architecture is the development of light-off or dual SCR
systems, which is a variation of passive thermal management.16-21-22 This system maintains a
layout similar to the conventional SCR configuration discussed earlier, but integrates an
additional small-volume SCR catalyst, which is in some cases also close-coupled to the
turbocharger's exhaust turbine outlet (Figure 1-4). This small SCR catalyst may be configured
with or without an upstream DOC, and with or without a small sacrificial substrate to adsorb
chemical poisons upstream of an initial SCR substrate. A recent example of this system's
architecture was demonstrated as part of "Stage 3" of the California Air Resources Board
(CARB) - Heavy-duty Low-NOx Test Program.23 The CARB Stage 3 research program is
summarized within Chapter 4.1.1.1. EPA is evaluating dual-SCR catalyst system technology
similar to the CARB "Stage 3" system as part of a diesel technology feasibility demonstration
program (see Chapter 4.1.1.2 for more detail).
The benefits of this design result from its ability to warm up the initial light-off SCR substrate
faster as a result of it being relatively low mass and being the first catalyst downstream of the
turbocharger with the EAS. Such light-off SCR catalysts can also be designed to have smaller
substrates with lower bulk density. The reduced mass reduces thermal inertia and allows faster
warmup. The design also positions the urea injection and mixing as the very first components in
the system, thus allowing faster heat up of the urea injector and urea mixer when implementing
active thermal management measures. These designs also require less input of heat energy into
the exhaust to maintain exhaust temperatures during light-load operation. Urea injection to the
close-coupled light-off SCR can also be reduced or terminated once the second, downstream
SCR reaches operational temperature, thus allowing additional NOx to reach the DOC and CDPF
to promote passive regeneration (soot oxidation) on the CDPF, reducing fuel consumption and
CO2 emissions. Very close-coupling of the light-off SCR to the exhaust turbine is possible when
using heated urea dosing system since such systems enable a relatively short mixing length
between the urea dosing system and the inlet of the light-off SCR (see Chapter 4.1.1.2 for more
detail).
18
-------
Urea Injector
Cu-SSZ-13 SCR Cu-SSZ-13 SCR + ASC
(zone coated)
Figure 1-4: Potential layout of a 2027+ dual-SCR system in an in-line configuration (top) and comparable
components integrated to improve passive thermal management (bottom).
One potential concern about this technology is the durability challenge associated with
placing an SCR catalyst upstream of the CDPF. To address this concern, two light-off SCR
system designs will be hydrothermally and chemically aged to an equivalent of 850,000 miles as
part of the EPA Heavy-duty Diesel Low NOx Demonstration Program. Please refer to Chapter
4.1.1.2 for additional information regarding this test program.
1.1.4 Closed Crankcases
During combustion, gases can leak past the piston rings sealing the cylinder and into the
crankcase. These gases are called blowby gases and generally include unburned fuel and other
combustion products. Blowby gases that escape from the crankcase are considered crankcase
emissions (see 40 CFR 86.402-78). Current regulations restrict the discharge of crankcase
emissions directly into the ambient air, and blowby gases from gasoline engine crankcases have
been controlled for many years by sealing the crankcase and routing the gases into the intake air
through a positive crankcase ventilation (PCV) valve. However, in the past there have been
concerns about applying a similar technology for diesel engines. For example, high PM
emissions venting into the intake system could foul turbocharger compressors. As a result of this
concern, diesel-fueled and other compression-ignition engines equipped with turbochargers (or
other equipment) were not required to have sealed crankcases (see 40 CFR 86.007-11(c)). For
these engines, manufacturers are allowed to vent the crankcase emissions to ambient air as long
as they are measured and added to the exhaust emissions during all emission testing to ensure
compliance against the emission standards.
Because all new highway heavy-duty diesel engines on the market today are equipped with
turbochargers, they are not required to have closed crankcases under the current regulations.
Manufacturer compliance data show approximately one-third of current highway heavy-duty
diesel engines have closed crankcases, indicating that some heavy-duty engine manufacturers
have developed systems for controlling crankcase emissions that do not negatively impact the
turbocharger. EPA is proposing provisions to require a closed crankcase ventilation system for
all highway compression-ignition engines to prevent crankcase emissions from being emitted
directly to the atmosphere (See Preamble Section III). These emissions could be routed upstream
of the aftertreatment system or back into the intake system.
19
-------
1.1.4.1 Emissions from Open Crankcases
EPA conducted emissions testing of open crankcase systems on two low mileage, modern
heavy-duty diesel trucks.24 The testing was conducted at EPA's National Vehicle and Fuel
Emissions Laboratory. The two vehicles were tested on a heavy-duty chassis dynamometer
where the crankcase flow and emissions were measured separately from the tailpipe exhaust
emissions. The vehicles were tested over a variety of operating conditions. The cycles included
the ARB Transient cycle with a cold start, repeat ARB Transient cycles, a 10 minute idle cycle,
and a highway cycle at 55 mph and 65 mph.
The crankcase emission rates were calculated for THC, NMHC, CFU, NOx, CO2, and CO
using the densities found in 40 CFR 86.144-94. The average crankcase and tailpipe emission
rates for each of the two trucks (NVFEL 1 and NVFEL 2) by test phase are show in Figure 1-5.
The error bars represent the standard error of the mean. As shown, the crankcase THC and CO
emissions are a notable fraction of the tailpipe exhaust emissions. Table 1-2 includes the average
crankcase emission rates across the cycles for each truck.
Table 1-2: Average Crankcase Emission Rates (gram/hour)
THC
CH4
NOx
CO
Truck 1
0.305
0.001
1.19
0.212
Truck 2
0.067
0.026
1.09
0.832
20
-------
7
b
4 5
m Tailpipe CO Ca/hr)
¦ Oanfceoire CO ls/f>r>
L 1.
Tran*j«r-rt
w/ itart
Tr#nw«r»l tfe
NVFEL 1
l Trill pip#" THG (fj/hrJ
I Cr*nkum« IHC (n/hr|
i
1.
1,
NVFEL I
Tailpipe CO,(g/t>r}
¦ CririfcCAi* CO,
3
2.5
m Tailpipe- CfrMfi/hr>
¦ Cwiki;#** CH„
1.
it
NVFEl 2
TrarahBWt kJle
NV FEL2
I I i
Tr*ri»t«ni fr«rraicr^ u
NVFEL2
~ .Si
O
1 JO
130
ICO
30
I
CO
40
?o
1.
If jKrvlf
W/SU I
Wle
NVFFI 1
¦ TAnlpiptf- HO* iJt/lir)
= Crankcase NO* (g/hr)
I
. i
NVKL 1
NVFEL2
Figure 1-5: Tailpipe Exhaust and Crankcase Emission Rates from Two Heavy-Duty Diesel Trucks
We were unable to measure PM emissions from the crankcase as part of the EPA test
program. Therefore, in our MOVES model we will continue to use the PM emission rates
measured in the Advanced Collaborative Emissions Study (ACES) Phase 1 test program.2? The
average PM emission rate of the four 2007 MY heavy-duty diesel engines was 32.1 mg/hour
21
-------
1.1.4.2 Description of Closed Crankcase Technologies
Crankcase emissions are controlled through the use of closed crankcase filtration systems or
by routing unfiltered blow-by gases directly into the exhaust system upstream of the emission
control equipment. Closed crankcase filtration systems work by separating oil and particulate
matter from the blow-by gases through single or dual stage filtration approaches, routing the
blowby gases into the engine's intake manifold and returning the filtered oil to the oil sump.
These systems are required for new heavy-duty diesel vehicles in Europe starting in 2000. Oil
separation efficiencies in excess of 90 percent have been demonstrated with production ready
prototypes of two stage filtration systems. By eliminating 90 percent of the oil that would
normally be vented to the atmosphere, the system works to reduce oil consumption and to
eliminate concerns over fouling of the intake system when the gases are routed through the
turbocharger.
An alternative approach would be to route the blow-by gases into the exhaust system
upstream of the catalyzed diesel particulate filter which would be expected to effectively trap and
oxidize the engine oil and diesel PM. This approach may require the use of low sulfur engine oil
to ensure that oil carried in the blow-by gases does not compromise the performance of the
sulfur-sensitive emission control equipment.
Our feasibility analysis is based on the use of closed crankcase system that includes
technologies designed to filter crankcase gases sending the clean gas to the engine intake for
combustion and returning the oil filtered from the gases to the engine crankcase. These systems
are proven in use.
1.1.5 Opposed-Piston Diesel Engines
While not part of EPA's planned technology demonstration program for the this rulemaking,
the agency is tracking ongoing work to develop opposed-piston diesel engine technology for
heavy-duty on-highway vehicle applications.26 27 One example of work with this technology is a
project to develop and demonstrate a 10.6 liter, 450 hp opposed-piston diesel engine and related
aftertreatment technologies for Class 8 line-haul tractors operating at certified emission
performance levels of 0.02 g NOx/bhp-h (90 percent below US 2010) and 432 g C02/bhp-h
(2027 HD Phase 2 engine standard).28 In addition to the emissions demonstration work, a high
level cost study has been conducted by FEV that indicates the direct and indirect cost of an
opposed-piston engine are less than that of a conventional HD diesel engine.29 This project is
supported by a variety of public and private sector partners including: Achates Power, Aramco
Services Company, BASF, CALSTART, CARB, Corning, Delphi, Eaton, Faurecia, Federal
Mogul, PACCAR/Peterbilt, Sacramento Metropolitan Air Quality Management District, San
Joaquin Valley Air Pollution Control District, South Coast Air Quality Management District,
Southwest Research Institute, Tyson Foods, and Walmart.30
Opposed-piston engine technology has not yet been proven feasible in Class 8 on-highway
applications, but if feasibility is shown, then the technology could provide another pathway to
ultra-low NOx, high efficiency engine technology for heavy-duty vehicle fleets. If the
demonstration project cited above is successful, then it may lead to early-commercial
deployment of opposed-piston diesel engines for heavy-duty on-highway applications in the
near-term. As such, it may be reasonable to anticipate commercialization of heavy-duty opposed-
piston diesel engine technology by model year 2027.
22
-------
1.2 Spark-Ignition Engine Technologies
The following sections describe the spark-ignition engine technologies we considered to
reduce criteria pollutant emissions (NMHC, CO, NOx, and PM) as part of this rulemaking. Many
of the technologies are described with respect to gasoline fuel, but they are expected to be
broadly applicable to all fuels used in spark-ignition engines. Our spark-ignition engine
feasibility demonstration for this proposal is based on some of the technologies presented in this
section.
Chapter 2.3 of this draft RIA describes the existing and proposed test procedures for spark-
ignition engine certification. Chapter 3.2 describes the spark-ignition engine feasibility
demonstration program, including a description of the specific technology packages we
evaluated, the effectiveness of those technologies over our current and proposed test procedures,
and our projected cost of those technologies.
1.2.1 Technology Description for NMHC, CO, and NOx Control
A range of technology options exist to reduce NMHC, CO, and NOx emissions from both
heavy-duty highway gasoline fueled spark ignition and diesel engines to levels below the current
EPA 2007/2010 standards. Available options include modifications to the engine calibration,
engine design, exhaust system, and aftertreatment system design. The different available options
each contain specific benefits and limitations. This section describes the technical challenges to
reducing emissions from current levels, describes available technologies for reducing emissions,
estimates the potential emissions reduction of the different technologies, describes if there are
other ancillary benefits to engine and vehicle performance with the technology, and reviews the
limits of each technology. Except where noted, these technologies are applicable to all spark-
ignition engines covered by this rule. Unique compression-ignition technologies are addressed in
Section 1.1.
1.2.1.1 Summary of the Technology Challenge for NMHC, CO, and NOx Control
Historically, heavy-duty spark-ignition engine FTP tests have shown that the majority of
NMHC, CO, and NOx emissions occur during the cold start phase; however, emissions during
warmed-up and hot operation, specifically during high-load operation, can significantly
contribute to emissions. Additionally, as described in Chapter 3.2, in-vehicle testing has
indicated that sustained low load conditions such as prolonged idling can result in emission
increases due to reduced aftertreatment temperatures (i.e. cool-off). The proposed standards in
this rulemaking target high-load NMHC, CO and NOx emission control performance.
Specifically, significant quantities of NMHC and CO emissions can be produced if enrichment
events occur regularly during high-load operation. Control of NOx emissions during high-load
operation requires designs that provide sufficient catalyst volume to handle the higher exhaust
gas flow rates and also precise control of closed loop fuel biasing for the catalyst to maintain
peak NOx efficiency.
In order to achieve significantly lower NMHC, CO and NOx emissions over the FTP,
manufacturers can change the design of their exhaust and catalyst systems, as well as adopt
calibration strategies to reduce catalyst light-off times and reduce warmed-up and hot running
emissions. Design changes to reduce catalyst light-off time (e.g. closer catalyst placement) can
23
-------
also result in higher catalyst temperatures during high-load operation. To achieve lower NMHC
and NOx levels, manufacturers will need to develop and implement technologies and calibration
strategies to manage catalyst temperatures during high-load operation while minimizing fuel
enrichment.
For the catalyst to effectively reduce NMHC, CO and NOx emissions it must reach a light-off
temperature of approximately 350 °C. Emissions during the catalyst warm up period can be
reduced by reducing the emissions produced by the engine during the catalyst warm up phase.
Emissions can also be reduced by shortening the time period required for the catalyst to reach the
light-off temperature and maintaining sufficient catalyst temperature during low load and idle
operation. Reducing warmed-up NOx emissions requires improving the efficiency of the catalyst
system using improved catalyst loading and washcoat technologies in addition to more precise
calibration and software controls. NOx emissions performance will also generally be reduced
from a reduction in the sulfur content of the fuel.
To achieve the proposed heavy-duty emission standards, it is anticipated that engine
manufacturers will focus on four areas to reduce emissions:
• Minimizing the emissions produced by the engine before the catalyst reaches the
light-off temperature
• Reducing the time required for the catalyst to reach the light-off temperature and
staying above the light-off temperature throughout all operation
• Improving the NOx efficiency of the catalyst during warmed-up operation at medium
and high loads
• Minimizing or eliminating enrichment in high-load operation.
We describe strategies to address these four areas in the following sections.
1.2.1.2 Reducing Engine-Out Emissions
During the first minutes of the cold start FTP, the engine is operating either at idle or low
speed and load. The engine start temperature is between 20 and 30 °C (68 and 86 °F). At these
temperatures and under these low loads the cold engine produces lower concentrations of NOx
than NMHC. As the engine warms up and as the load increases the concentration of NOx
produced by the engine increases and the concentration of NMHC decreases.
The design of the air induction system, combustion chamber, spark plug, and fuel injection
system determine the quantity of fuel required for stable combustion to occur in the cold engine.
Optimizing the performance of these components can provide reductions in the amount of fuel
required to produce stable combustion during these cold operating conditions. Reductions in the
amount of fuel required leads to reductions in cold start NMHC emissions.
The design considerations to minimize cold start emissions are also dependent on the fuel
injection method. Port fuel injected (PFI) engines have different design constraints than gasoline
direct injection (GDI) spark ignition engines. For both PFI and GDI engines, however, attention
to the details affecting the in-cylinder air/fuel mixture can reduce cold start NMHC emissions.
24
-------
For example, it has been shown that cold start NMHC emissions in PFI engines can be
reduced by reducing the size of the fuel spray droplets and optimizing the spray targeting. Fuel
impinging on cold engine surfaces in the cylinder does not readily vaporize and does not
combust.31 Improving injector targeting to reduce the amount of fuel reaching the cylinder walls
reduces the amount of fuel needed to create a combustible air fuel mixture. Reducing the size of
the spray droplets improves the vaporization of the fuel and the creation of a combustible
mixture. Droplet size can be reduced by modifying the injector orifice plate and by increasing the
fuel pressure. Reducing droplet size and improving fuel vaporization during cold start has been
shown to reduce cold transient emissions by up to 40 percent during the cold start phase of the
light-duty FTP emission test.
The mixture formation process in a GDI engine is different than a PFI engine. In a PFI engine
the fuel can be injected prior to or during the intake valve opening to prepare the fuel in an
optimal manner for emission controls. The fuel generally has time to evaporate during the intake
stroke as the fuel and air are drawn into the cylinder and is mixed with the incoming air. In
addition, as the engine combustion heat from the previous firing events warms the intake valve
and other surfaces in the area, the fuel can be injected into the intake runner and engine heat can
assist in evaporating the fuel prior to the intake valve opening. The GDI engine injects fuel at
higher fuel pressures than PFI engines directly into the combustion chamber. In a GDI engine,
the fuel droplets need to evaporate and mix with the air in the cylinder in order to form a
flammable mixture. Injecting directly into the cylinder reduces the time available for the fuel to
evaporate and mix with the intake air in a GDI engine compared to a PFI engine. An advantage
of the GDI design is that the fuel spray does not impinge on the walls of the intake manifold or
other surfaces in the cylinder.
GDI systems stagger the injection timing event. At least one study has indicated that
significant reductions in hydrocarbon emissions can be achieved by splitting the injections
during the cold start of a GDI engine. An initial injection occurs during the intake stroke and a
second injection is timed to occur during the compression stroke. This injection method reduced
unburned hydrocarbon emissions 30 percent compared to a compression stroke only injection
method.32
These are two examples of specific engine design characteristics, fuel injector design and fuel
system pressure on PFI engines and injection timing on GDI engines, which can be used to
reduce cold start NMHC emissions significantly during the engine warm up prior to the catalyst
reaching the light-off temperature.
Optimizing the fuel injection system design and calibration is anticipated to be used in all
vehicle classes, including heavy-duty vehicles. It is anticipated that these described
improvements, along with improvements to other engine design characteristics, will be used to
reduce cold start emissions for passenger cars, LDTs, MDPVs, and HDTs in coming model
years, which would pave the way for them to be applied to heavy-duty engines.
Because the engine is relatively cold and the operating loads are low during the first 50
seconds of operation, the engines typically do not produce significant quantities of NOx
emissions during this time. In addition, manufacturers tend to retard the combustion timing
during the catalyst warm up phase. Retarding combustion timing has been shown also to reduce
the concentration of NMHC in the exhaust. This calibration method further reduces peak
combustion temperatures while increasing the exhaust gas temperature compared to optimized
25
-------
combustion timing. The increased exhaust gas temperature leads to improved heating of the
catalyst and reduced catalyst light-off times. Retarding combustion and other technologies for
reducing catalyst light-off time are discussed in the following section.
1.2.1.3 Reducing Catalyst Light-Off Time
The effectiveness of current engine emission control systems depends in large part on the time
it takes for the catalyst to light-off, which is typically defined as the catalyst reaching a
temperature of 350°C. In order to reduce catalyst light-off time, it is expected manufacturers will
use technologies that will improve heat transfer to the catalyst during the cold start phase and
improve catalyst efficiency at lower temperatures. Technologies to reduce catalyst light-off time
include calibration changes, thermal management, close-coupled catalysts, catalyst PGM
loading, and secondary air injection. The technologies are described in greater detail below.
1.2.1.3.1 Calibration Changes
Engine calibration changes may be employed to increase the temperature and mass flow of
the exhaust with the goal of reducing the amount of time required for the catalyst to reach the
critical light-off temperature. By reducing the time required for the catalyst to light-off, these
changes can effectively reduce NMHC, CO and NOx emissions. Since the catalyst system in an
SI engine is the predominant method to control emissions and is responsible for over a 95%
reduction from the engine out emissions, any acceleration in the warm-up of the catalyst system
translates into immediate emission reductions at the tailpipe.
Retarding combustion in a cold engine by retarding the spark advance is a well-known
method for reducing the concentration of NMHC emissions in the exhaust and increasing the
exhaust gas temperature.33 34 The reduction in NMHC concentrations is due to a large fraction of
the unburned fuel within the cylinder combusting before the flame is extinguished at the cylinder
wall. Reductions of total hydrocarbon mass of up to 40 percent have been reported from these
studies evaluating the effect of spark retard on exhaust emissions.
In addition to reducing the NMHC exhaust concentration, retarding the spark advance reduces
the torque produced by the engine. In order to produce the same torque and maintain the engine
speed and load at the desired level when retarding the spark advance, the air flow into the engine
is increased causing the manifold pressure to increase which can also improve combustion
stability. Retarding the combustion process also results in an increase in the exhaust gas
temperature. The retarded ignition timing during the cold start phase in addition to reducing the
NMHC emissions increases the exhaust mass flow and exhaust temperature. These changes lead
to a reduction in the time required to heat the catalyst.
The torque produced by the engine will begin to vary as the spark retard amount reaches
engine combustion limits. As the torque variations increase, the combustion process is
deteriorating, and the engine performance begins to degrade due to the partial burning. It is the
level of this variability which defines the absolute maximum reduction in spark advance that can
be utilized to reduce NMHC emissions and reduce the catalyst light-off time.
Retarding combustion during cold start can be applied to spark-ignition engines in all vehicle
classes. The exhaust temperatures and NMHC emission reductions will vary based on engine
design. This calibration methodology is anticipated to be used to improve catalyst warm- up
times and reduce cold start NMHC emissions for all vehicle classes.
26
-------
With the penetration of variable valve timing technology increasing in gasoline-fueled
engines, additional work is being performed to characterize the impact of valve timing on cold
start emissions. Calibration changes to the valve timing during the cold start phase can lead to
additional reductions in cold start NMHC emissions.35
1.2.1.3.2 Exhaust System Thermal Management
This category of technologies includes all design attributes meant to conduct combustion heat
into the catalyst with minimal cooling. This includes insulating the exhaust piping between the
engine and the catalyst, reducing the wetted area of the exhaust path, reducing the thermal mass
of the exhaust system, and/or using close-coupled catalysts (i.e., the catalysts are packaged as
close to the engine cylinder head as possible to mitigate the cooling effects of longer exhaust
piping). By reducing the time required to achieve catalyst light-off, thermal management
technologies reduce NMHC, CO and NOx emissions.
Moving the catalyst closer to the cylinder head is a means that manufacturers have been using
to reduce both thermal losses and the catalyst light-off time. Many vehicles today use close-
coupled catalysts, a catalyst which is physically located as close as possible to the cylinder head.
Moving the catalyst from an underbody location closer to the cylinder head reduces the light-off
time significantly.
Another means for reducing heat losses is to replace cast exhaust manifolds with thin-wall
stamped manifolds. Reducing the mass of the exhaust system reduces the heat losses of the
system. In addition, an insulating air gap can be added to the exhaust system which further
reduces the heat losses from the exhaust system. Insulating air gap manifolds are also known as
dual-wall manifolds.
With thin- and dual-wall exhaust manifolds, close-coupled catalyst housings can be welded to
the manifold. This reduces the needed for manifold to catalyst flanges which further reduces the
thermal inertia of the exhaust system. Close coupling of the catalyst and reducing the thermal
mass of the exhaust system significantly reduces the light-off time of the catalyst compared to an
underbody catalyst with flanges and pipes connected to a cast exhaust manifold.
Using close-coupled catalysts reduces the heat losses between the cylinder head and catalyst.
While reducing the time required to light-off the catalyst the close-coupled catalyst can be
subject to higher temperatures than underbody catalysts during high-load operating conditions.
To ensure the catalyst does not degrade, manufacturers currently use fuel enrichment to maintain
the exhaust temperatures below the levels which would damage the catalyst. It is anticipated that
to meet the proposed heavy-duty emission standards, manufacturers will ensure that fuel
enrichment is minimized on the FTP. Calibration measures, other than fuel enrichment, may
therefore be needed to ensure the catalyst temperature does not exceed the maximum limits.
Another technology beginning to be used for both reducing heat loss in the exhaust and
limiting exhaust gas temperatures under high-load conditions is integrating the exhaust manifold
into the cylinder head. Honda utilized this technology on the Insight's 1.0 L VTEC-E engine.
The advantage of this technology is that it minimizes exhaust system heat loss during warm-up.
In addition, with the exhaust manifold integrated in the cylinder head, the cooling system can be
used to reduce the exhaust temperatures during high-load operation. It is anticipated that
27
-------
manufacturers will further develop this technology as a means to both quickly light-off the
catalyst and reduce high-load exhaust temperatures.
We expect thermal management to be an effective strategy for manufacturers to lower
NMHC, CO and NOx emission levels. Our feasibility demonstration described in Chapter 3.2
evaluates catalysts located closer to the engine as a method of thermal management. We expect
that manufacturers will further optimize the thermal inertia of the exhaust system to minimize the
time needed for the catalyst to achieve the light-off temperature, while ensuring the high-load
performance does not cause thermal degradation of the catalyst system. It is expected that
methods and technologies will be developed to reduce the need to use fuel enrichment to reduce
high-load exhaust temperatures.
Optimizing the catalyst location and reducing the thermal inertia of the exhaust system are
design options manufacturers can apply to all vehicle classes for improving engine cold start
emission performance. It is not anticipated that heavy-duty vehicles with spark-ignition engines
will utilize catalysts that are very close-coupled to the exhaust manifold (i.e., will not use a
strategy similar to close-coupled catalyst locations found on passenger cars and light-duty
trucks). The higher operating loads of these heavy-duty engines results in durability concerns due
to high thermal loading. It is expected that manufacturers will work to optimize the thermal mass
of the exhaust systems to reduce losses along with optimizing the underbody location of the
catalyst. These changes are expected to improve the light-off time while not subjecting the
catalysts to the higher thermal loadings from a close coupled location.
1.2.1.3.3 Catalyst Design Changes
There are several different catalyst design changes that can be implemented to reduce the time
for the catalyst to light-off Changes include modifying the substrate design, replacing a large
volume catalyst with a cascade of two or more catalysts, and optimizing the loading and
composition of the platinum group metals (PGM).
Progress continues to be made in the development of the catalyst substrates which provide the
physical support for the catalyst components, which typically include a high surface area alumina
carrier, ceria used for storing oxygen, PGM catalysts, and other components. A key design
parameter for substrates is the cell density. Today, catalyst substrates can be fabricated with cell
densities up 1,200 cells per square inch (cpsi) with wall thicknesses approaching 0.05 mm.
Increasing the surface area of the catalyst improves the performance of the catalyst. Higher
substrate cell densities increase the surface area for a given catalyst volume. Higher surface areas
improve the catalyst efficiency and durability reducing NMHC and NOx emissions.
The key limitation of the higher cell density substrates is increased exhaust system pressures
at high-load conditions. The cell density and substrate frontal area are significant factors that
need to be considered to optimize the catalyst performance while limiting flow loss at high-load
operation.
Engine speeds and load are low during the first 50 seconds of the FTP test and it is a
challenge to achieve catalyst light-off during the cold start operation. One method for reducing
the catalyst light-off time is to replace a single catalyst with two catalysts which when combined
total the same volume as the single catalyst. Having a two-catalyst system that includes a close-
coupled, front catalyst with comparatively reduced-volume reduces the heat needed to reach the
28
-------
light-off temperature for the front catalyst due to its location and reduced thermal mass. The
front catalyst of the two-catalyst system will reach operating temperature before the individually
larger volume single catalyst, reducing the light-off time of the system.
All other parameters held constant, increasing the PGM loading of the catalyst also improves
the efficiency of the catalyst. The ratio of PGM metals is important as platinum, palladium, and
rhodium have different levels of effectiveness promoting oxidation and reduction reactions.
Therefore, as the loading levels and composition of the PGM changes, the light-off performance
for both NMHC and NOx need to be evaluated. Improved catalyst substrates and PGM loadings
designs additional effective approaches to reduce emissions and we anticipate manufacturers will
incorporate advanced catalyst designs in their future emission strategies.36 We used an advanced
catalyst formulation in our HD SI feasibility demonstration for this rule (see Chapter 3.2).
1.2.1.3.4 Secondary Air Injection
By injecting air directly into the exhaust stream, close to the exhaust valve, combustion
(hydrocarbon oxidation) can be maintained within the exhaust, creating additional heat and
thereby further increasing the catalyst temperature. The air/fuel mixture must be adjusted to
provide a richer exhaust gas for the secondary air to be effective.
Secondary air injection systems are used after the engine has started and once exhaust port
temperatures are sufficiently high to sustain combustion in the exhaust port. When the
secondary air pump is turned on the engine control module increases the amount of fuel being
injected into the engine. Sufficient fuel is added so that the air/fuel ratio in the cylinder is rich of
stoichiometric. The exhaust contains significant quantities of CO and hydrocarbons. The rich
exhaust gas mixes with the secondary air in the exhaust port and the combustion process
continues, increasing the temperature of the exhaust and rapidly heating the manifold and close-
coupled catalyst.37-38
Engines which do not use secondary air injection can only operate rich of stoichiometry for a
minimal amount of time after a cold start as the added enrichment would cause increased NMHC
emissions. The richer cold start calibration used with vehicles that have a secondary air injection
system provides a benefit, as combustion stability is improved. In addition, the richer calibration
is not as sensitive to changes in fuel volatility. Less volatile fuels found in the market may result
in poor start and idle performance on engines calibrated to run lean during the cold operation.
Engines which use secondary air and have a richer warm up calibration would have a greater
combustion stability margin. Manufacturers may perceive this to be a benefit for the operation of
their vehicles during the cold start and warm up phase.
Historically, secondary air injection has also been used to control CO and NMHC emissions
during high load rich operation. These designs incorporated a mechanically driven air pump that
would continuously inject air into part of the catalyst to oxidize the CO and NMHC emissions, a
particularly important emission control technology used during enrichment operation. With the
improvements in electronic engine controls, all manufacturers discontinued the use of secondary
air injection for the purpose of high load enrichment and instead used software models and other
algorithms that maintain stoichiometric operation for slightly longer periods of operation than
previously. These newer strategies however were designed to only provided a temporary
emission control benefit of stoichiometric operation since the engines eventually go into
29
-------
enrichment modes for either power improvement or thermal protection at which point the CO
and NMHC emissions are no longer controlled.
1.2.1.4 Improving Catalyst NOx Efficiency during Fully Warmed-up Operation
For engines certified to the EPA 2007/2010 emission standards, significant quantities of NOx
emissions are produced by engines during warmed-up engine operation on the FTP. The
stabilized NOx emission levels will need to be reduced to achieve the proposed heavy-duty
NMHC, CO and NOx emission standards. Improving the NOx performance of the engine can be
achieved by improving the catalyst efficiency during warmed-up operation. As previously
described, the performance of the catalyst can be improved by modifications to the catalyst
substrate, increasing cell density, increasing PGM loadings, and, particularly important, reducing
the sulfur level of gasoline. Three-way catalyst efficiency is also affected by frequency and
amplitude of the air/fuel ratio. For some engines warmed-up catalyst NOx efficiency can be
improved by optimizing the air/fuel ratio control and limiting the amplitude of the air fuel ratio
excursions. It is anticipated that a combination of changes will be made by manufacturers,
including further improvements to air/fuel ratio calibration and catalyst changes including cell
density and PGM loadings.
1.2.1.5 Reducing Enrichment
Heavy-duty vehicles tend to operate at high loads and catalyst durability can be a concern due
to the increased thermal loading as the catalyst is moved closer to the cylinder head. Moving the
catalyst closer to the exhaust manifold could result in increasing the time spent in fuel
enrichment modes to ensure the catalyst temperatures are maintained below design thresholds,
which if allowed to operate too hot could reduce the durability of the catalyst. Using fuel
enrichment to control catalyst temperature while effective, causes significant increases in criteria
pollutant emissions and also significant increases in fuel consumption.
1.2.1.5.1 Exhaust Gas Temyerature Measurement
The methodology for determining when temperatures in the exhaust and in the catalyst are
high enough to initiate thermal protection (i.e., enrichment) is almost exclusively done using
software modeling of the thermally-limited components. This methodology can be effective at
triggering enrichment when needed; however, if it is implemented in an excessively conservative
manner where the temperature prediction is higher than the actual critical component
temperature, it can result in unnecessary enrichment episodes which lead to substantial increases
in all the emissions. Since the gasoline heavy-duty engines are designed for work and are
expected to operate regularly at high loads where the exhaust temperatures become important
concerns for component durability, any improvement in the accuracy of the methodology to
provide enrichment protection will result in both reductions in emissions and improvements to
durability.
A potential improvement over the current practice of using temperature modeling is use of
hardware in the form of temperature measurement sensors in the exhaust at the most critical
locations, generally at the catalyst inlet. While some light-duty variations are available,
currently, only one gasoline HD product, the Nissan NV series of cargo vans with the 5.6L V8
has implemented temperature sensors.39 However, temperature measurement sensors are very
common in all diesel applications. While this improvement to the accuracy of temperature
30
-------
measurements in the exhaust may not result in emission reductions during the limited operation
range of today's certification test cycles, it will likely provide "real world" emission reductions
compared to modeled temperature strategies that may conservatively trigger enrichment episodes
prematurely.
1.2.1.5.2 Continuous Stoichiometric Operation
One concept that has been used in other non-road sectors (e.g., large SI engines operating
indoors or confined areas) is requiring the engine to always operate at stoichiometry where the
three-way catalyst is generally at peak efficiency for all emissions. To apply this strategy to
heavy-duty highway SI engines, manufacturers would have to prevent the engine from entering
enrichment. Enrichment could be avoided with upgrades to materials of specific components that
are currently limited by high temperature constraints, or using the large degree of modern engine
control authority to prevent the engine from entering areas of operation that require enrichment.
Modern gasoline engines have several engine hardware components and calibration strategies
that could be used to reduce temperatures by modifying combustion or load characteristics, such
as EGR, valve timing, electronic throttle airflow, cylinder deactivation, and other available
methods. Some of these methods to remain in stoichiometric air-fuel control may result in a
governing or detuning of the engine after a period of time to avoid prolonged high power
operation where stochiometric operation cannot be maintained due to increasing exhaust
component temperatures; however, this may be an acceptable approach for a sector where
sustained absolute engine power may not be necessary or as important as lower emissions and
better fuel economy.
While we were unable to directly control enrichment in our HD SI feasibility demonstration,
in Chapter 3.2 we show that emissions can be well-controlled when an engine maintains
stoichiometric operation. We expect engine manufacturers will continue to optimize their engine
calibrations and limit enrichment as an effective means of reducing NMHC and CO emissions to
meet future standards.
1.2.1.6 Additional Emission Control Strategies
A strategy that may provide some degree of emission reductions involves down-speeding or
governing of the engine operating range to keep the engine speeds and loads in areas where
engine hardware and exhaust temperatures minimize needed enrichment for thermal protection.
This strategy will allow the emission controls to remain in stoichiometric air-fuel control (i.e.
closed loop) where the catalysts can maintain peak efficiency for NMHC, CO and NOx for a
broader range of operation.
The down-speeding approach was a technology discussed to reduce CO2 emissions for the HD
Phase 2 GHG rule40 and has been successfully implemented in both gasoline and diesel
applications to reduce GHG emissions. With the advent of modern electronic controls of both
engine and transmission, the opportunity exists to precisely keep the engine in the optimal
operating areas for reduced GHG and criteria emissions. Multi-speed automatic transmissions
available in recent years in HD applications with wide ratios containing 6 or more forward gears
have provided the opportunity to operate the engine in a more optimal fashion with little or no
loss of vehicle performance and capabilities. This strategy is currently used by at least one HD
gasoline engine manufacturer as indicated by its advertised maximum rated test speed (i.e. peak
horsepower) of approximately 4000 RPM compared to the much higher speeds, over 4700 RPM
31
-------
of other HD applications.41 This lower speed is made possible by transmission strategies
preventing over-speeding, which allows the emission controls to operate in a much more
desirable and lower emitting area of engine operation. We evaluated engine down-speeding as
part of the HD SI feasibility demonstration presented in Chapter 3.2.
1.2.2 Technology Description for PM Control
Particulate matter emitted from internal combustion engines is a multi-component mixture
composed of elemental carbon (or soot), semi-volatile organic compounds (SVOC), sulfate
compounds (primarily sulfuric acid) with associated water, nitrate compounds and trace
quantities of metallic ash. At temperatures above 1,300 K, fuel hydrocarbons without access to
oxidants can pyrolyze to form particles of elemental carbon. Fuel pyrolysis can occur as the
result of operation at richer than stoichiometric air-to-fuel ratio (primarily PFI gasoline GDI
engines), direct fuel impingement onto surfaces exposed to combustion (primarily GDI and
diesel engines), and non-homogeneity of the air-fuel mixture during combustion (primarily diesel
engines). Elemental carbon particles that are formed can be oxidized during later stages of
combustion via in-cylinder charge motion and reaction with oxidants.
SVOCs are composed primarily of organic compounds from lubricant and partial combustion
products from fuel. PM emissions from SVOC are typically gas phase when emitted from the
engine and contribute to PM emissions via particle adsorption and nucleation after mixing with
air and cooling. Essentially, PM-associated SVOC represent the condensable fraction of NMHC
emissions. Sulfur and nitrogen compounds are emitted primarily as gaseous species (SO2, NO
and NO2). Sulfate compounds can be a significant contributor to PM emissions from stratified
lean-burn gasoline engines and diesel engines, particularly under conditions where PGM
containing exhaust catalysts used for control of gaseous and PM emissions oxidize a large
fraction of the SO2 emissions to sulfate (primarily sulfuric acid). Sulfate compounds do not
significantly contribute to PM emissions from spark-ignition engines operated at near
stoichiometric air-fuel ratios due to insufficient availability of oxygen in the exhaust for
oxidation of SO2 over PGM catalysts.
Elemental carbon PM emissions can be controlled by:
• Reducing fuel impingement on piston and cylinder surfaces
• Inducing charge motion and air-fuel mixing via charge motion (e.g., tumble and
swirl) or via multiple injection (e.g., GDI and diesel/common rail applications)
• Injection strategies that eliminate opportunity for PM forming conditions (open valve
injection on PFI)
• Reducing or eliminating operation at net-fuel-rich air-to-fuel ratios (PFI gasoline and
GDI applications)
• Use of wall-flow or partial-wall-flow exhaust filters (GPF)
SVOC PM emissions can be controlled by:
• Reducing lubricating oil consumption
32
-------
• Improvements in exhaust catalyst systems used to control gaseous NMHC emissions
(e.g., increased PGM surface area in the catalyst, improvements in achieving catalyst
light-off following cold-starts, etc.)
1.2.3 Technologies to Address Evaporative Emissions
As exhaust emissions from gasoline engines continue to decrease, evaporative emissions
become an increasingly significant contribution to overall HC emissions from gasoline-fueled
vehicles. To evaluate the evaporative emission performance of current production heavy-duty
gasoline vehicles, EPA tested two heavy-duty vehicles over running loss, hot soak, three-day
diurnal, on-board refueling vapor recovery (ORVR), and static test procedures. These engine-
certified "incomplete" vehicles meet the current heavy-duty evaporative running loss, hot soak,
and three-day diurnal emission requirements. However, as they are certified as incomplete
vehicles, they are not required to control refueling emissions and do not have ORVR systems.
Results from the refueling testing confirm that these vehicles have much higher refueling
emissions than gasoline vehicles with ORVR controls.42-43 The results for the ORVR tests are
shown in Table 1-3. A discussion on the test procedure limitations and estimated modeled results
from this test program is in Section 2.3.2.
Table 1-3: ORVR results with modeled values for test procedure limitations
SHED (grams)
Modeled
SHED
(grams)
Average (g/gal)
Modeled
Average
(g/gal)
Current
Refueling
Standard
(g/gal)
Ford E-450
114
168
2.3
3.34
0.2
108
2.2
Isuzu NPR
72
86
2.8
71
2.8
Opportunity exists to extend the usage of the refueling evaporative emission control
technologies already implemented in complete heavy-duty gasoline vehicles to the engine-
certified incomplete gasoline vehicles in the over-14,000 lb. GVWR category. The primary
technology we are considering is the addition of ORVR, which was first introduced to the
chassis-certified light-duty and heavy-duty applications beginning in MY 2000 (65 FR 6698,
February 10, 2000). An ORVR system includes a carbon canister, which is an effective
technology designed to capture HC emissions during refueling events when liquid gasoline
displaces HC vapors present in the vehicle's fuel tank as the tank is filled. Instead of releasing
the HC vapors into the ambient air, ORVR systems recover these HC vapors and store them for
later use as fuel to operate the engine.
The fuel systems on these over-14,000 pound GVWR incomplete heavy-duty gasoline
vehicles are similar to complete heavy-duty vehicles that are already required to incorporate
ORVR. These incomplete vehicles may have slightly larger fuel tanks than most chassis-
certified (complete) heavy-duty gasoline vehicles and are somewhat more likely to have dual
fuel tanks. These differences may require a greater ORVR system storage capacity and possibly
some unique accommodations for dual tanks (e.g., separate fuel filler locations), but we expect
they will maintain a similar design. Figure 1-6 presents a schematic of a standard ORVR system.
33
-------
fj
¦J
In
* \
Carbon
Canister
|t"*i
1
y
L
Multi-Function
Control Valves
Fill Limit
Vent Vstve
Filler Pipe
Inlet
Check
Valve
Figure 1-6: Schematic of an ORVR system11
1.2.3.1 Filler Pipe and Seal
In an ORVR system, the design of the filler pipe, the section of line connecting the point at
which the fuel nozzle introduces fuel into the system to the gas tank, is integral to how fuel
vapors displaced during a fuel fill will be handled. The filler pipe is typically sized to handle the
maximum fill rate of liquid fuel allowed by law while also integrating one of two methods to
prevent fuel vapors from exiting through the filler pipe to the atmosphere: a mechanical seal or a
liquid seal approach. A dual fuel tank chassis configuration may require a separate filler pipe and
seal for each fuel tank.
The mechanical seal is typically located at the top of the filler neck at the location where the
fuel nozzle is inserted into fuel neck. The hardware piece forms a seal against the fuel nozzle by
using some form of a flexible materi al (usually a plastic materi al) that makes direct contact with
the fuel station fuel-filling nozzle to prevent fuel vapors from exiting the filler pipe as liquid fuel
is pumped into the fuel tank. In the case of capless systems, this seal may be integrated into the
spring-loaded seal door that opens when the nozzle is inserted into the filler pipe receptacle.
There are concerns with a mechanical seal's durability due to wear over time, and its ability to
maintain a proper seal with unknown service station fill nozzle integrity and variations beyond
design tolerances.
H Stant ORVR System https://www.stant.com/
34
-------
The liquid seal approach uses the size and bends of the filler pipe to cause a condition where
the entire cross-section of the filler pipe is located in the fuel tank or close to the entry into the
fuel tank and is full of the incoming liquid fuel preventing fuel vapors from escaping up and out
through the filler pipe. By creating a solid column of liquid fuel in the filler pipe, the liquid seal
approach does not require a mechanical contact point with the fill nozzle to prevent escape of
vapors. The liquid seal has been the predominant sealing method implemented in the regulated
fleet in response to the ORVR requirements.
1.2.3.2 ORVR Flow Control Valve
As described above, the sealing of the filler pipe prevents the fuel vapors from escaping into
the ambient air; however, the fuel vapors that are displaced by the incoming liquid fuel need to
be routed to the canister. In order to properly manage the large volume of vapors during
refueling that need to be controlled, most ORVR systems have implemented a flow control valve
that senses that the fuel tank is getting filled with fuel and triggers a unique low-restriction flow
path to the canister. This flow path is specifically used only during the refueling operation and is
unique in that it provides the ability to quickly move larger volumes of fuel vapors into the tank
than normally required under other operation outside of refueling events. The flow control valve
will allow this larger flow volume path while refueling but then return to a more restrictive vapor
flow path under all other conditions, including while driving and while parked for overnight
diurnal s.
The flow control valve is generally a fully-mechanical valve system that utilizes connections
to the fuel tank and filler pipe to open and close vapor pathways with check valves and check
balls and pressure switches via diaphragms. The valve may be integrated into the fuel tank and
incorporate other aspects of the fuel handling system ("multi-function control valve" in Figure
1-6) including roll-over valve, fuel and vapor separators to prevent liquid fuel from reaching the
canister, and other fuel tank vapor control hardware. Depending on the design, the filler pipe
may also be integrated with the flow control valve to provide the necessary pressure signals. A
dual fuel tank chassis configuration may require a separate flow control valve for each fuel tank.
1.2.3.3 Canister
The proven technology to capture and store fuel vapors has been activated charcoal. This
technology has been used in vehicles for over 50 years to reduce evaporative emissions from
sources such as fuel tanks and carburetors. When ORVR was originally discussed, existing
activated charcoal technology was determined to be the appropriate technology for the capture
and storage of refueling related fuel vapors. This continues to be the case today, as all known
ORVR-equipped vehicles utilize some type of activated charcoal.
The activated charcoal is contained in a canister, which is made from a durable material that
can withstand the fuel vapor pressures, vibration, and other durability concerns. For vehicles
without ORVR systems, canisters are sized to handle evaporative emissions for the three-day
diurnal test with the canister volume based on the fuel tank capacity. A dual fuel tank chassis
configuration may require a separate canister for each fuel tank.
1.2.3.4 Purge Valve
The purge valve is the electro-mechanical device used to remove fuel vapors from the fuel
tank and canister by routing the vapors to the running engine where they are burnt in the
35
-------
combustion chamber. This process displaces some amount of the liquid fuel required from the
fuel tank to operate the engine and results in a small fuel savings. The purge valve is controlled
by the engine or emission control electronics with the goal of removing the necessary amount of
captured fuel vapors from the canister in order to prepare the canister for subsequent fuel vapor
handling needs of either the next refueling event or vapors generated from a diurnal event. All
on-road vehicles equipped with a canister for evaporative emissions control utilize a purge valve.
Depending on the design, a dual fuel tank chassis configuration may require a separate purge
valve for each fuel tank.
1.2.3.5 Design considerations for Unique Fuel Tanks
The commercial truck market gasoline applications incorporate several fuel tank options that
may require unique ORVR design considerations. While most commercial vehicle fuel tanks are
similar to the already ORVR-compliant complete vehicles in the 8500 to 14,000 GVWR class,
some of the commercial vehicles include larger tank sizes (50 to 70 gallons) or may have a dual
tank option. As described above, the canister sizing will be a function of the required amount of
fuel vapor handling during refueling. Larger fuel tanks will require larger canisters with more
activated charcoal than historically found in other gasoline vehicles. Some design challenges
will likely exist in designing the canister system to handle the large vapor volumes while
balancing the restriction to flow through the larger activated charcoal containing canisters.
Dual fuel tank systems, which have very limited availability, may also require some unique
design considerations. Typically, the canister is located in very close proximity to the fuel tank
to properly manage the refueling fuel vapors efficiently with minimal distance between the tank
and canister. Dual fuel tanks may require duplicate ORVR systems to have the necessary
flexibility to manage the refueling vapors, particularly since the fuel tanks are filled
independently through separate filler pipe assemblies.
A small portion of the commercial truck market gasoline applications have fuel tanks that are
similar in design to diesel fuel tanks located on the outside of the frame. These tanks are
typically cylindrical or rectangular in shape with the gas cap directly on the top of the tank and
do not have a fill neck. These type of fuel tanks may require unique approaches such as a
mechanical seal built into the fuel tank filling location where the fuel cap is normally located, or
they may require a design that adds a filler pipe for a liquid seal approach.
1.3 Fuels Considerations
Both the compression- and spark-ignition engine technologies discussed above are capable of
running on alternative fuels (e.g., natural gas, biodiesel). We have typically applied the gasoline-
and diesel-fueled engine standards to alternatively-fueled engines based on the combustion cycle
of the alternatively-fueled engine: applying the gasoline-fueled engine standards to spark-
ignition engines and the diesel-fueled engine standards to compression-ignition engines. This
approach is often called "fuel neutral." The sections below discuss some of the available
alternative fuels in more depth.
1.3.1 Natural gas
With relatively low natural gas prices (compared to their peak values) in recent years, the
heavy-duty industry has become increasingly interested in engines that are fueled with natural
gas. It has some emission advantages over diesel, with lower engine-out levels of both NOx and
36
-------
PM. Several heavy-duty CNG engines have been certified with NOx levels better than 90 percent
below US 2010 standards. However, because natural gas must be distributed and stored under
pressure, there are additional challenges to using it as a heavy-duty fuel.
Liquified Petroleum Gas (LPG) is also used in certain lower weight-class urban applications,
such as airport shuttle buses, school buses, and emergency response vehicles. LPG use is not
extensive, nor do we project it to grow significantly in the proposed rulemaking timeframe.
1.3.2 Biodiesel
Over the last decade, biodiesel content in diesel fuel has increased under the Renewable Fuels
Standard. In 2010, less than 400 million gallons of biodiesel were consumed in the U.S.,
whereas in 2018, over 2 billion gallons of biodiesel were being blended into U.S. diesel fuel.
While the biodiesel content in diesel fuel averaged around 3.5 percent in 2018, biodiesel levels
range from 0 to 20 volume percent in highway diesel fuel. As discussed further below, with
increasing volumes of biodiesel in the fuel stream, greater attention has been given to the
influence of biodiesel on highway diesel fuel quality. The following sub-sections discuss how
biodiesel is produced and the importance of purification processes for removing potential metal
containments, standards for biodiesel fuel quality, potential impacts of metals in biodiesel fuel on
diesel engines, aftertreatment systems, and emissions, as well as efforts by EPA and others to test
the metal content of biodiesel samples across the country.
1.3.2.1 Biodiesel Production and a Potential for Metals
Biodiesel can be made from various renewable sources, such as vegetable oil, animal fat, or
waste cooking oil. It is produced through transesterification of the oil or fat with methanol,
which results in mono-alkyl esters and the co-product glycerin. The process occurs in the
presence of a catalyst, typically sodium or potassium methoxide or hydroxide.44 Following
transesterification and separation of the glycerin, the biodiesel must be purified, which is usually
done by extracting, distilling, or filtering the impurities into water.
The purification process is essential to address the potential for metal and other impurities in
biofuel. There are number of potential sources of metal contamination in biofuel production.
These include:
1) Vegetable oil seeds used to produce feedstock contain high concentrations of sodium (Na),
potassium (K), calcium (Ca), magnesium (Mg), and potassium (P), as well as aluminum (Al),
iron (Fe), manganese (Mn), zinc (Zn), and smaller concentrations of other metals.45
2) The potassium and sodium methoxide catalysts which break down triglycerides to methyl
esters (NaOH and KOH can also be used) can contribute metals to biodiesel. These metals can
form soaps with free fatty acids, and the soaps in both the metal esters and glycerine forms are
reacted with acid (hydrochloric acid) to convert the soaps to free fatty acids so they can be more
easily removed. Sodium hydroxide is added to neutralize any acid added to eliminate soaps.
3) Methyl esters are washed, distilled or filtered to remove the metals added as catalysts. The
wash water is recycled, and metal ions can accumulate in the wash water. Hard wash water
containing CaCC>3, Mg(OH)2, CaSC>4 is found in Rocky Mountain states and the Midwest, and
these water-soluble compounds can accumulate in the residual water found in biodiesel.
37
-------
4) The medium used to filter methyl esters could also contribute to metals in the biodiesel.
The filter material is typically made up of diatomaceous earth which is primarily silica
containing alumina, iron oxide and calcium oxide. In addition, small amounts of calcium or
magnesium can be added to the fuel from the purification process.46'47
1.3.2.2 Standards for Biodiesel Fuel Quality
Biodiesel quality, including metal content, is regulated by ASTM D6751-19 for B100 fuels.
ASTM D6751-19 sets a limit of 5 ppm for combined Na and K (group 1A metals) and a limit of
5 ppm for combined Ca and Mg (group 2A metals) using the EN14538 inductively coupled
plasma optical emission spectroscopy (ICP-OES) measurement method.48 ASTM D6751-18
also places a 10-ppm limit on P (group 5 metal) using the ASTM D4951 inductively coupled
plasma atomic emission spectroscopy (ICP-AES) measurement method.49 The limits on metals
in ASTM D6751 are meant to be protective when biodiesel is used in blends (e.g., B20, BIO).
Fuel quality for biodiesel blends in the B6 to B20 range is regulated by ASTM D7467-19.50 This
specification does not contain a metal limit for these biofuel blends because, as the method
states, the concentration would likely be too low to measure using the ICP-OES method
specified (EN 14538). Similarly, D975 regulates BO to B5 and does not have a metals
specification (just a total ash % limit of 0.01%).51 Thus, the basis for control of metals in
biodiesel blends is control of the B100 blend stock. The thought being that if the B100 fuel is
under the ASTM D6751-19 limit, the combined Na + K and Mg + Ca will be below 1 ppm
respectively for B20 and lower blends. Yet, the actual metal content of today's fuels can be
challenging to quantify when it is lower than the 1 ppm level specified for B20 and lower blends,
because of the detection limit of the current test methods. The detection limit of the EN14538 is
1 ppm for each metal, and the method includes a statement if the metal is below the limit of
detection of the method, then it is not included in the reporting calculation. Efforts to quantify
biodiesel metal contents below 1 ppm are discussed in 1.3.2.5 below.
1.3.2.3 Potential Impacts of Metals on Engine and Emission Control Devices
Across a range of concentrations, metals in biodiesel can be present as ions, abrasive solids or
soluble metallic soaps. Abrasive solids can contribute to wear of fuel system components,
pistons and rings, as well as contribute to engine deposits. Soluble metallic soaps have little
impact on wear but may contribute to diesel particulate filter plugging and engine deposits.
Metal accumulation in diesel particulate filters can increase pressure drops and result in shorter
times between maintenance intervals.52'53 A level of 1 mg/kg (1 part per million) of trace metal
in the fuel result in an estimated accumulation of about 22 g of trace metal in diesel particulate
filters per 100,000 miles (assuming a fuel economy of 15 mpg and 100% trapping efficiency).52
Metallic fuel contaminants can also accumulate on fuel injectors, or be converted to oxides,
sulfates, hydroxides or carbonates in the combustion process, which forms an inorganic ash that
can deposit onto the exhaust emission control devices found in modern diesel engines.54 Alkali
metals are well known poisons for catalysts used in emission control devices, and have been
shown to negatively impact the mechanical properties of ceramic substrates.55'56 Alkali metal
hydroxides such as Na and K are volatilized in the presence of steam and can, therefore,
penetrate the catalyst washcoat or substrate.
38
-------
1.3.2.4 Potential for Emissions Impacts of Metals in Biodiesel
Numerous studies have collected and analyzed emission data from diesel engines operated on
biodiesel blended diesel fuel with controlled amounts of metal content.53'57'58'59'60'61'62'63'64 Some
of these studies show an impact on emissions, while others do not. However, four factors need
to be considered when reviewing these studies:
1. These studies were conducted using accelerated aging protocols and exposure to these
metals from the fuel consumed in a more conventional manner could cause different effects (the
effects could be greater or less) than what these studies show,
2. The emissions testing studies were designed to test the effect of metal content in biodiesel
if the metal content was at the ASTM limit, however, as shown below, biodiesel likely contains a
lower metal content than the standard.
3. Different manufactures use different catalyst formulations and different physical layouts
for emission aftertreatment systems, and while one manufacturer might be less susceptible to
metals contamination, others may be more affected. This issue relates to factor #4:
4. When these studies examined the effect of metals on heavy-duty engines, they studied the
impact of these metals based on current engine and aftertreatment configurations over the current
regulatory useful life. This proposed rule would require heavy-duty engines to comply with a
more stringent NOx standard and a longer useful life. Certainly, a longer useful life would
expose the aftertreatment devices to increased amounts of metals (many of today's engines often
operate beyond the current regulatory useful life and would already be exposed to more metals
than during their regulatory useful life). Also, the engine manufacturers may change the
composition and configuration of their aftertreatment devices to comply with the proposed
standards, which could affect how fuel metals would affect the aftertreatment devices.
Brookshear et al. 2012 studied the impact of Na on heavy-duty diesel engine aftertreatment
devices.59'1 In this accelerated aging study, they doped a B20 fuel to 5,000 ppm each of Na and S
and aged to an equivalent 435,000 miles. They found impacts on SCR function if the SCR was
positioned before the DPF. There was no impact on the DOC or DPF.
Lance et al. 2016 also studied the effect of Na on heavy-duty diesel engine aftertreatment.6 u
They doped their B20 fuel with Na to a level of 14 ppm or 14 times the pseudo 1 ppm limit of a
B20 fuel and accelerated aged the aftertreatment out to 435,000 miles. The results indicated an
acceleration of DPF ash buildup and platinum group metal migration from the DOC/CDPF to the
SCR. The results of the system performance, including degradation in performance, are shown
in Figure 1-7. The results indicated that the degradation in NOx performance can be attributed to
degradation of all aftertreatment components.
1 Author affiliations: University of Tennessee and Oak Ridge National Lab
1 Author affiliations: Oak Ridge National Lab, Cummins, and National Renewable Energy Lab
39
-------
0.50
0.45
_ 0.40 ¦
j=
£¦0.35
1
-0.30
*
c
? 0.25
Z, 0.20
«
5 0.15
£
" 0.10
0.05
0.00
Figure 1-7: SCR performance over the hot start HDDE FTP61'K
Williams et al. 2011 studied the effect of Na and Ca on a 2008 non-road 8.8L Caterpillar
diesel engine, a MAN D2066 10.5 L diesel engine, and a 2008 Cummins 8.3L diesel engine64,L
They doped their B20 fuel to 27 times the pseudo 1 ppm Na and Ca limit of a B20 fuel and
accelerated aging of the emission control systems out to 150,000 and 435,000 miles. The results
showed no significant degradation in the thermo-mechanical properties of cordierite, aluminum
titanate, or silicon carbide DPFs after exposure to 150,000-mile equivalent biodiesel ash and
thermal aging. It is estimated that the additional ash from 150,000 miles of biodiesel use would
also result in moderate increases in exhaust backpressure for a DPF. A decrease in DOC activity
was seen after exposure to 150,000-mile equivalent aging, resulting in higher HC slip and a
reduction in NO2 formation. The exposure of a cordierite DPF to 435,000-mile equivalent aging
resulted in a 69% decrease in the thermal shock resistance parameter. The metal-zeolite SCR
catalyst experienced a slight loss in activity after exposure to 435,000-mile equivalent aging.
This catalyst, placed downstream of the DPF, showed a 5% reduction in overall NOx conversion
activity over the HDDT test cycle.
I x
I
v.
c
M
T5
35
et
<
•a
< X
u <3
C f
Cl
O C
Q
"8 =:
5 K
£ -J
%'
— —
c 2
M&.
5 -
u n
c 2
c -
U T3
g ^
- <
C 0.
0
o
K Note that DG indicates that the specific component is not an aged part, but a new degreened part
L Author affiliations: National Renewable Energy Lab, Manufacturers of Emission Controls, BASF, Caterpillar, and
Umicore AG
40
-------
Williams et al. 2013 studied the effect of Na, K and Ca on a 2011 LD 6.7L diesel engine
aftertreatment.54'M They doped their B20 fuel to 14 times the pseudo 1 ppm Na and Ca limit of a
B20 fuel and accelerated aged the emission control systems out to 150,000 miles. The authors
aged sets of production exhaust systems that included a DOC, SCR catalyst, and DPF. Four
separate exhaust systems were aged, each with a different fuel: ULSD containing no measurable
metals, B20 containing sodium, B20 containing potassium, and B20 containing calcium.
Analysis of the aged catalysts included Federal Test Procedure emissions testing with the
systems installed on a Ford F250 pickup, bench flow reactor testing of catalyst cores, and
electron probe microanalysis (EPMA). The thermo-mechanical properties of the aged DPFs were
also measured.
EPMA imaging of aged catalyst parts found that both the Na and K penetrated into the
washcoat of the DOC and SCR catalysts, while Ca remained on the surface of the washcoat.
Bench flow reactor experiments were used to measure the standard NOx conversion, NFb
storage, and NFb oxidation for each of the aged SCR catalysts. Flow reactor results showed that
the first inch of the SCR catalysts exposed to Na and K had reduced NOx conversion through a
range of temperatures (Figure 1-8 and Figure 1-9) and also had reduced NFb storage capacity.
The SCR catalyst exposed to Ca had similar NOx conversion and NFb storage performance
compared to the catalyst aged with ULSD.
Chassis dynamometer vehicle emissions tests were conducted with each of the aged catalyst
systems installed onto a Ford F250 pickup. Regardless of the evidence of catalyst deactivation
seen in flow reactor experiments and EPMA imaging, the vehicle successfully passed the 0.2
gram/ mile NOx emission standard with each of the four aged exhaust systems. This indicates
that if catalyst volumes are chosen to account for degradation, the emission control system can
accommodate some loss in catalyst activity since deactivation occurred only in the first inch of
the catalyst and did not affect overall NOx emissions.54
M Author affiliations: National Renewable Energy Lab, Oak Ridge National Lab, Manufacturers of Emission
Controls, BASF, Ford, and University of Tennessee
41
-------
-10
100 150 200 250 300 350 400 450 500 550 600 650
Catalyst Bed Temperature (°C)
Figure 1-8: SCR NOx conversion for the first inch of aged SCR catalysts.54
100
90
80
g 70
| 60
£ 50
4>
£ 40
o
O 30
O 20
10
0
-10
100 150 200 250 300 350 400 450 500 550 600 650
Catalyst Bed Temperature (°C)
Figure 1-9: SCR NOx conversion for the seventh inch of aged SCR catalysts.54
1.3.2.5 Testing for Metals in Biodiesel
The National Renewable Energy Laboratory (NREL) has conducted several studies on the
metal content of biodiesel. The NREL studies generally look at the fuel quality in fuel samples
taken across from the country. In some cases, samples are taken at the refinery where B100 was
sampled and in other cases they are taken at the pump. These studies provided analysis of metal
content for fuel samples collected and analyzed in the 2006, 2007, 2008, 2011, and 2016
timeframes. The 2006, 2007, and 2011 studies analyzed the metal content of B100 fuel, while
the 2008 and 2016 studies analyzed the metal content of biodiesel blends up to approximately
B20. Some samples taken during the 2006 study were acquired prior to the finalization of the
combined Na + K limit in ASTM D6751 (May 2006).
42
-------
These results indicate that over the 2006 to 2018 time frame the incidence of off specification
biofuel decreased over time. A summary of the off-specification samples can be found in Table
1-4.
Table 1-4: NREL Fuel Samples off Specification for ASTM D6751 (or equivalent B20) limit for Na + K and
Ca + Mg.
Number
Number
of
of
NREL Fuel Study
Year
Biodiesel Content
Samples
off spec
for Na
+ K
Samples
off spec
for Ca
+ Mg
Total Number of
Samples
2006 (pre-D6751)
B100
7
1
24
2006 (post-D6751)
B100
0
1
15
2007
B100
3
3
55
2008
B0.2 to B90
6
0
34
2011
B100
1
0
67
2016
B0 to B22
0
1
35
2017
B100
1
0
459
2018
B100
0
0
491
The NREL studies prior to 2016 focused on identifying gross exceedances of the blends and
blend stocks. As noted above, the analytical method specified in ASTM D6751-18 (EN 14538)
affords a detection limit of 1 ppm, which is adequate to ensure whether or not biofuel blend
stocks (B100) are compliant with the Na + K and Ca + Mg limits. In addition to determining
compliance with the ASTM D6751-18 limit, it is also important to determine the actual metal
content of these fuels in order to assess the levels that aftertreatment systems will be exposed to
over their full useful life. The 2016 NREL study used measurement equipment and procedures
capable of testing down to lower metal levels, which is useful for understanding the actual metal
content of biodiesel blends. Table 1-5 summarizes the test procedures and level of detection
(LOD) levels for each year of the NREL studies.
Table 1-5: Test procedure LODs for NREL studies.
Year of NREL Study
Test Procedure
Level of Detection (ppm)
Na
K
Ca
Mg
2006
ASTM D5185
?
?
?
?
2007
ASTMD7111
1
1
0.1
0.1
2008
ASTMD7111
0.5
0.5
0.5
0.1
2011
ASTMD7111
1
1
0.1
0.1
2016
UOP-389
0.1
0.1
0.1
0.1
2016
ICP-MS
0.029
0.001
0.0005
0.001
2017
EN 14538
1
1
1
1
2018
EN 14538
1
1
1
1
43
-------
In the 2016 study, NREL took steps to improve their understanding of testing accuracy and
improve the testing resolution of the analysis by utilizing three different measurement methods,
two of which afforded very low detection limits. In this study, biodiesel blends were analyzed
for metal content using the Universal Oil Products Method 389 (UOP-389), Microwave Plasma
Atomic Emission Spectroscopy (MP-AES), and Inductively Coupled Plasma Mass Spectroscopy
(ICP-MS) methods.70 This allowed for comparisons of metal content from the three different
testing methods. Almost all the results for the MP-AES testing method were below the LOD,
though, so this method will not be further discussed. The results of the study for UOP-389 and
the ICM-MS are presented in Table 1-6.
44
-------
Table 1-6: NREL 2016 Metals results for UOP-389 and ICP-MS70
State
Biodiesel
Content
Error
UOP-389 ICP-AES
ICP-MS
Na
K
Ca
Mg
Na
K
Ca
Mg
Fe
Vol %
Vol %
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
FL-1 (fleet)
22.01
1.76
0.14
0.04
0.08
0.01
(0.029)
0.021
0.008
0.026
0.065
FL-2
18.38
1.59
0.04
0.01
0.05
0.01
(0.029)
0.033
0.002
0.014
0.146
GA
19.36
1.63
0.06
0.02
0.06
0.01
(0.029)
0.034
0.002
0.011
1.44
NC-1 (fleet)
20.4
1.68
0.06
0.01
0.09
0.02
(0.029)
0.02
0.002
0.01
0.004
NC-2
0
0.71
0.02
0.01
0.02
0.01
No Data
No Data
No Data
No Data
No Data
PA-1
17.73
1.56
0.09
0.01
0.05
0.02
(0.029)
0.014
0.002
0.009
0.012
PA-2 (fleet)
20.31
1.68
0.08
0.02
0.04
0.03
(0.029)
0.031
0.002
0.019
0.017
MA
21.35
1.73
0.04
0.01
0.04
0.01
(0.029)
0.02
0.002
0.015
0.016
VA (fleet)
18.56
1.6
0.08
0.02
0.06
0.02
(0.029)
0.028
0.012
0.011
0.086
IA (fleet)
16.1
1.48
0.09
0.03
0.04
0.01
(0.029)
0.1
0.017
0.014
0.249
IL
15.3
1.44
0.07
0.01
0.03
0.01
(0.029)
0.02
0.007
0.019
0.005
IN
20.34
1.68
0.2
0.07
0.1
0.03
(0.029)
0.237
0.106
0.031
0.075
KS
17.27
1.53
0.09
0.02
0.53
0.06
(0.029)
0.11
1.366
0.166
1.557
KY
20.11
1.67
0.05
0.01
0.03
0.01
(0.029)
0.022
0.002
0.015
0.069
MI
20.51
1.69
0.11
0.02
0.06
0.01
(0.029)
0.045
0.122
0.025
3.117
MN-1
10.68
1.22
0.04
0.01
0.04
0.01
(0.029)
0.016
0.002
0.025
0.172
MN-2
10.61
1.22
0.02
0.01
0.03
0.01
(0.029)
0.016
0.002
0.048
0.003
MO
19.44
1.64
0.09
0.01
0.05
0.01
(0.029)
0.034
0.01
0.012
0.019
OH-1
20.4
1.68
0.05
0.01
0.05
0.01
(0.029)
0.013
0.002
0.011
0.033
OH-2
8.84
1.13
0.01
0.01
0.03
0.01
(0.029)
0.015
0.002
0.007
0.003
TN-1
20.24
1.68
0.1
0.01
0.06
0.01
(0.029)
0.019
0.003
0.006
0.085
TN-2 (fleet)
20.02
1.67
0.24
0.02
0.08
0.01
(0.029)
0.029
0.012
0.009
0.361
LA
12.26
1.3
0.04
0.01
0.03
0.01
No Data
No Data
No Data
No Data
No Data
TX-1
17.14
1.53
0.06
0.01
0.08
0.01
(0.029)
0.01
0.002
0.005
0.003
TX-2
1.4
0.78
0.07
0.01
0.04
0.01
(0.029)
0.012
0.002
0.006
0.004
TX-3
1.2
0.77
0.04
0.01
0.03
0.01
(0.029)
0.011
0.002
0.005
0.004
CO
21.99
1.76
0.09
0.01
0.05
0.01
(0.029)
0.023
0.094
0.035
0.037
ID (fleet)
19.64
1.65
0.22
0.03
0.12
0.03
(0.029)
0.085
0.065
0.034
0.044
AZ (fleet)
20.39
1.68
0.07
0.01
0.04
0.01
(0.029)
0.085
0.065
0.034
0.044
CA-1
18.84
1.61
0.05
0.02
0.03
0.01
(0.029)
0.016
0.002
0.019
0.009
CA-2
22.12
1.77
0.07
0.01
0.04
0.01
(0.029)
0.016
0.002
0.011
0.03
CA-3 (fleet)
19.9
1.66
0.06
0.01
0.04
0.01
(0.029)
0.033
0.002
0.018
0.005
NM
19.61
1.65
0.1
0.02
0.05
0.01
(0.029)
0.025
0.002
0.005
0.284
OR-1
20.1
1.67
0.06
0.02
0.03
0.02
(0.029)
0.012
0.002
0.005
0.009
OR-2
20.15
1.67
0.05
0.01
0.05
0.01
(0.029)
0.021
0.002
0.005
0.011
Average
0.079
0.016
0.064
0.014
(0.029)
0.037
0.058
0.021
0.243
Average
Na + K and Ca + Mg
0.047
0.039
0.033
0.040
Note: Values in parenthesis are below the detection limit and are reported at the detection limit.
Differences in the results between the two methods could be due to interferences by other
metals, or other aspects of the molecules, with the measuring method. The measurement
45
-------
differences could also be due to the unique ionization efficiency of each element and how well
the instrument ionizes the element of interest. Even small differences could impact the results at
sub 1-ppm levels. The difference in sample preparation techniques can also have a significant
effect on the results. The UOP-389 method uses acid digestion followed by ashing, while the
ICP-MS method used a simpler preparation of sample dilution and direct analysis. The UOP-389
method was developed for the analysis of petroleum products and blending components,
including biodiesel blends, and uses a wet ashing method that is unique to this procedure, which
is why it was selected for this project.
Although several samples contained elevated amounts of Ca and Mg (KS, MI, IN, and CO)
that were well above the level of other samples, these trace levels would still be very low in the
B100 blend stock, with the exception of one sample (KS). Overall, the NREL data suggest that
metal contents in biodiesel have decreased over time and, as of 2018, are generally very low
across samples. Nevertheless, small sample sizes could be biasing the resuits.44'45'65'66'67'68'69
To that end, in 2019 an engine manufacturer raised concerns to EPA that biodiesel is the
source of high metal content in highway diesel fuel, and that higher biodiesel blends, such as
B20, are the principal problem.71 The engine manufacturer observed higher than normal
concentrations of alkali and alkaline earth metals (Na, K, Ca, and Mg) in their highway diesel
fuel samples, and observed fouling of the aftertreatment control systems of their engines, which
caused an associated increase in emissions. The engine manufacturer sampled the ash that was
fouling their fuel injectors and aftertreatment devices and determined the ash to be composed of
sodium sulfate, sodium carboxylates, and sodium chloride, which they claimed were from
biodiesel. The engine manufacturer recommends limiting biodiesel blends to 5 percent biodiesel
(B5). After hearing engine manufacturer concerns about the metal content in biodiesel in early
2019, EPA began to focus a previously developed fuel sampling program on biodiesel metal
content. Below, we summarize the information that we obtained through that sampling program.
Separate from hearing about engine manufacturer concerns with biodiesel metal content, EPA
began a process to determine the metal content of different fuels in early 2016. The most
prominent concern to EPA at that time was the blending of less refined natural gas liquids with
ethanol to produce E85 blends; however, EPA recognized the need to understand the metal
content of all fuels. EPA initiated a sampling effort in late 2016 to obtain samples of diesel fuel,
gasoline, natural gas liquids, jet fuel, biodiesel and ethanol. The samples were collected in acid-
washed glass bottles using a clean hands/dirty hands sampling procedure to reduce the chance for
contamination during the sampling process. Fuel samples were collected from both small and
large fuel production plants, in the case that facility size plays a role in the amount of metals
which ends up in the fuel. Samples were also taken in different geographical regions. Fuel
production facilities likely use different feedstocks based on their geographical region, and the
different feedstock types could affect the fuels metal content. Because of the cost and effort
involved in obtaining these fuel samples, a limited number of samples of each fuel type were
obtained. Thus, this was a screening study, which could be expanded later on if high metal
contents were detected in any of the fuels.
Approximately 100 samples were obtained across the various fuel types. A subset of these
samples (27 B100 samples) were recently sent to the California Department of Food and
Agriculture (CDFA) laboratory. These samples were analyzed for biodiesel regulated metals,
including Na, K, Ca, Mg and P, and also tested for Molybdenum (Mo), Boron (B), Barium (Ba),
46
-------
Copper (Cu), Manganese (Mn), Silica (Si), Titanium (Ti), Vanadium (V) and Zinc (Zn). The
California lab utilized the ASTM D7111-16 ICP-AES method that returned detection limits of
0.023 (Na), 0.052 (K), 0.013 (Ca), 0.004 (Mg), 0.001 (P), 0.006 (Mo), 0.013 (B), 0.001 (Ba),
0.005 (Cu), 0.001 (Mn), 0.017 (Si), 0.003 (Ti), 0.002 (V), and 0.005 (Zn) ppm.72 The results of
the analysis are shown in Table 1-7 and
Table 1-8.
Na was above the detection limit for 22 of the samples, with the highest result at 564 ppb. K
was only above the detection limit for 3 of the samples, with the highest result at 660 ppb. Ca
was above the detection limit for 9 of the samples, with the highest result at 551 ppb. Mg was
only above the detection limit for 5 of the samples, with the highest result at 133 ppb. The
highest result for combined Na and K was 744 ppb, while the highest result for combined Ca and
Mg was 662 ppb.
All of the 27 B100 fuel samples from this test program were compliant with the ASTM
D6751-18 limit of 5 ppm for Na + K and Ca + Mg respectively, and the results showed that
levels were at less than 20% of the limit for two of the samples, while the rest were at less than
10% (and in most cases well below that) of the limit. A reduction of 80% in metal content for
B20 and a reduction of 95% in metal content for B5 fuel blends would result in a maximum Na +
K content of 149 ppb and 37 ppb respectively for the B100 fuel with the highest Na + K content.
Ca + Mg would be 132 ppb and 33 ppb respectively.N
N This assumes no contribution from the diesel fuel used to formulate the blends.
47
-------
Table 1-7: EPA 2017 Metals results for ICP-AES analysis of Na, K, Ca, and Mg performed by CDFA.
Sample ID
Area of US
Na
(ppm)
K
(ppm)
Ca
(ppm)
Mg
(ppm)
Na + K
(ppm)
Ca + Mg
(ppm)
25982
East
0.171
0.211
0.131
0.133
0.382
0.264
25988
East
0.084
0.660
(0.013)
0.018
0.744
[0.031]
25998
Midwest
0.241
(0.052)
(0.013)
0.005
[0.293]
[0.018]
26004
Midwest
(0.023)
(0.052)
(0.013)
(0.004)
(0.075)
(0.017)
26006
Midwest
0.081
(0.052)
(0.013)
(0.004)
[0.133]
(0.017)
26083
Midwest
0.181
(0.052)
0.026
(0.004)
[0.233]
[0.030]
26084
Midwest
0.188
(0.052)
0.025
(0.004)
[0.240]
[0.029]
26088
Midwest
0.111
(0.052)
(0.013)
(0.004)
[0.163]
(0.017)
26090
Midwest
0.135
(0.052)
(0.013)
(0.004)
[0.187]
(0.017)
26092
Midwest
0.201
(0.052)
(0.013)
(0.004)
[0.253]
(0.017)
26095
Midwest
0.378
(0.052)
(0.013)
(0.004)
[0.430]
(0.017)
26164
South
0.193
(0.052)
0.040
(0.004)
[0.245]
[0.044]
26165
South
0.044
(0.052)
(0.013)
(0.004)
[0.096]
(0.017)
26166
South
(0.023)
(0.052)
(0.013)
(0.004)
(0.075)
(0.017)
26217
West
0.108
(0.052)
(0.013)
(0.004)
[0.160]
(0.017)
26218
West
0.564
(0.052)
0.027
(0.004)
[0.616]
[0.031]
26219
West
0.278
(0.052)
0.143
0.031
[0.330]
0.174
26248
West
0.303
(0.052)
0.551
0.111
[0.355]
0.662
26250
South
0.031
0.186
(0.013)
(0.004)
0.217
(0.017)
26253
South
(0.023)
(0.052)
(0.013)
(0.004)
(0.075)
(0.017)
26254
South
(0.023)
(0.052)
(0.013)
(0.004)
(0.075)
(0.017)
26256
South
(0.023)
(0.052)
(0.013)
(0.004)
(0.075)
(0.017)
26283
South
0.331
(0.052)
(0.013)
(0.004)
[0.383]
(0.017)
26830
South
0.379
(0.052)
0.016
(0.004)
[0.431]
[0.020]
26833
East
0.290
(0.052)
(0.013)
(0.004)
[0.342]
(0.017)
27581
East
0.226
(0.052)
0.044
(0.004)
[0.278]
[0.048]
27955
West
0.270
(0.052)
(0.013)
(0.004)
[0.322]
(0.017)
Average
0.182
0.085
0.046
0.014
0.267
0.060
* Values in (parenthesis) are below the detection limit and are reported at the detection limit.
**Na + K and Ca + Mg values in [square brackets] include one element that is below the detection limit and is
included in the calculation at the detection limit.
48
-------
Table 1-8: EPA 2017 Metals results for ICP-AES analysis of Mo, P, B, Ba, Cu, Mn, Si, Ti, V, and Zn
performed by CDFA.
Sample
Area of US
Mo
P
B
Ba
Cu
Mn
Si
Ti
V
Zn
ID
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
25982
East
0.070
0.301
2.577
0.057
0.157
0.141
2.364
0.138
0.134
0.041
25988
East
0.012
0.305
1.220
0.006
0.032
0.014
2.896
0.004
0.010
(0.005)
25998
Midwest
(0.006)
0.207
0.126
0.002
0.006
0.002
0.927
(0.003)
0.003
(0.005)
26004
Midwest
(0.006)
0.073
0.154
(0.001)
(0.005)
(0.001)
0.034
(0.003)
(0.002)
(0.005)
26006
Midwest
0.008
0.138
0.081
(0.001)
(0.005)
(0.001)
0.100
(0.003)
(0.002)
(0.005)
26083
Midwest
(0.006)
0.188
0.062
0.003
0.022
0.004
0.241
(0.003)
(0.002)
(0.005)
26084
Midwest
(0.006)
0.182
0.053
0.002
0.023
0.004
0.233
(0.003)
(0.002)
(0.005)
26088
Midwest
(0.006)
0.735
0.029
(0.001)
(0.005)
(0.001)
0.020
(0.003)
(0.002)
(0.005)
26090
Midwest
0.008
0.226
0.017
(0.001)
0.019
(0.001)
0.172
(0.003)
(0.002)
(0.005)
26092
Midwest
0.010
0.165
0.038
(0.001)
0.046
(0.001)
1.168
(0.003)
0.004
(0.005)
26095
Midwest
(0.006)
0.100
0.029
(0.001)
0.054
(0.001)
1.043
(0.003)
0.004
(0.005)
26164
South
0.022
0.232
0.022
(0.001)
0.024
0.003
0.220
(0.003)
(0.002)
(0.005)
26165
South
(0.006)
0.949
(0.013)
(0.001)
(0.005)
(0.001)
0.046
(0.003)
(0.002)
(0.005)
26166
South
(0.006)
0.019
0.014
(0.001)
(0.005)
(0.001)
0.032
(0.003)
(0.002)
(0.005)
26217
West
0.011
0.101
0.024
(0.001)
0.035
(0.001)
0.350
(0.003)
0.021
(0.005)
26218
West
(0.006)
(0.001)
(0.013)
(0.001)
0.025
(0.001)
(0.017)
(0.003)
(0.002)
0.049
26219
West
(0.006)
0.302
0.014
(0.001)
0.035
(0.001)
0.104
(0.003)
(0.002)
(0.005)
26248
West
(0.006)
0.365
(0.013)
0.004
0.037
(0.001)
1.776
(0.003)
(0.002)
(0.005)
26250
South
(0.006)
(0.001)
(0.013)
(0.001)
(0.005)
(0.001)
0.130
(0.003)
(0.002)
(0.005)
26253
South
(0.006)
0.025
0.023
(0.001)
(0.005)
(0.001)
1.077
(0.003)
(0.002)
(0.005)
26254
South
(0.006)
0.021
0.027
(0.001)
(0.005)
(0.001)
1.239
(0.003)
(0.002)
(0.005)
26256
South
(0.006)
(0.001)
(0.013)
(0.001)
(0.005)
(0.001)
0.306
(0.003)
(0.002)
(0.005)
26283
South
(0.006)
0.133
(0.013)
(0.001)
0.193
(0.001)
0.592
(0.003)
0.020
(0.005)
26830
South
(0.006)
0.193
(0.013)
(0.001)
0.066
(0.001)
1.516
(0.003)
0.004
(0.005)
26833
East
0.012
0.237
(0.013)
(0.001)
0.035
0.007
0.397
(0.003)
0.013
(0.005)
27581
East
(0.006)
0.038
(0.013)
(0.001)
(0.005)
(0.001)
0.461
(0.003)
0.004
(0.005)
27955
West
(0.006)
(0.001)
(0.013)
(0.001)
(0.005)
(0.001)
0.115
(0.003)
(0.002)
(0.005)
Average
0.010
0.194
0.172
0.004
0.032
0.007
0.651
0.008
0.009
0.008
* Values in parenthesis are below the detection limit and are reported at the detection limit.
The California Air Resources Board (ARB) CDFA inspectors carried out a biodiesel sampling
campaign throughout California during the spring and fall of 2019 collecting three hundred fifty-
five (355) biodiesel and diesel fuel samples from both #2 diesel labeled pumps and biodiesel
labeled pumps in the state of California.73
These samples were analyzed by the same lab as the 27 EPA samples mentioned above and
afforded the same detection limits. The primary focus of analysis was to examine the average
and observed range of concentration for Na, K, Ca, Mg and P of the biodiesel samples and the
diesel samples.
Statistical analysis of the samples showed that the Na, K, Ca, Mg and P concentrations in all
of the 355 collected fuel samples across California were significantly lower than the worst case
expected concentrations for a B20 fuel blended from B100 blend stock that is at the ASTM
49
-------
D6751-18 limit. Only three P samples, one Mg + Ca sample, and thirteen Na + K samples across
the entire sample set exceeded worst case expected absolute concentrations for a B5 blended
from B100 blend stock that is at the ASTM D6751-18 limit.
Na was the most abundant metal observed and was above the detection limit for 273 of 355
samples with sample 30077 exhibiting the highest result at 837 ppb. The rest of the metals were
largely below detection limits. K was only above the detection limit for 14 of 355 samples with
sample 15162 exhibiting the highest result at 172 ppb. Ca was above the detection limit for 24
of 355 samples with sample 30062 exhibiting the highest result at 168 ppb. Mg was only above
the detection limit for 32 of 355 samples with sample 15162 exhibiting the highest result at 238
ppb. Sample 30077 exhibited the highest result for combined Na and K at 889 ppb, while sample
30062 exhibited the highest result for combined Ca and Mg at 353 ppb. P was above the
detection limit for 92 of 355 samples, with sample 30077 exhibiting the highest result at 862 ppb.
Tables containing the maximum and average concentrations with standard deviation can be
found in the ARB comments to the ANRPM in the docket.73
A review of the NREL, EPA, and ARB data sets indicate that biodiesel fuel is compliant with
the ASTM D6751-18 limits for Na, K, Ca, and Mg. While the test results indicate that there is
an occasional B100 blend stock that is off specification with respect to the ASTM D6751-18
limits, and occasional BXX blends that are off specification to the pseudo limits, these
occurances are the exception. The NREL 2016, EPA, and ARB data sets all use measurement
methods that afford low levels of detection (sub-100 ppb), and these data sets further indicate
that the Na, K, Ca, and Mg content of biodiesel blends is extremely low in general, on the order
of less than 100 ppb. While these metals are present in biodiesel blends and testing has shown
that exposure to metals can adversely affect emission control system performance, data suggest
that the low levels measured in today's fuels are not enough to adversely affect system
performance out to full regulatory useful life, provided that the engine manufacturer properly
sizes the catalysts to account for the low4evel exposure.
1.4 Advanced Powertrain Technologies
This section discusses powertrain technologies capable of reducing NOx emissions from
heavy-duty vehicles. In particular, we provide a technology description and emissions
performance discussion, as well as an overview of current and future markets for heavy-duty
hybrid electric vehicles (HEVs), battery electric vehicles (BEVs), and fuel cell electric vehicles
(FCEVs). These powertrain technologies (HEVs, BEVs, FCEVs) are collectively referred to as
advanced powertrain technologies where appropriate in this discussion.
1.4.1 Hybrid
1.4.1.1 Technology Description
Heavy-duty HEVs are those that are propelled by both an on-board engine using a
consumable fuel (e.g., diesel internal combustion engine, hydrogen fuel cell) and an energy
storage device (e.g., battery, capacitor). HEV technologies that recover and store braking energy
using a driveline-coupled electric machine and battery storage have been used extensively in
light-duty applications as fuel saving features and are being adopted in certain heavy-duty
applications. Heavy-duty hybrid technologies can be described along a continuum of "mild" to
"strong" in terms of the extent to which they utilize an energy source other than a combustion
50
-------
engine; they also differ in the coupling and placement of the electric machine within the
powertrain. The specific definitions of "mild" and "strong" hybrids can be challenging to solidify
given variation in how different manufacturers employ individual and combinations of hybrid
technologies. The purpose of this discussion is to provide an overview of the types of
technologies employed across the mild to strong hybrid spectrum, rather than focus on defining
the technologies per say.°
For instance, North American Council for Clean Freight Efficiency (NACFE) defines "mild"
hybrids as those that use a starter or generator combined with a small battery pack to supplement
main engine power in select circumstances; most mild hybrids use 48V electrical batteries or
batteries with lower voltage and can provide fuel economy benefits in the range of 10% or less.74
The 48V battery is a lithium-ion battery that pairs with an electric motor to supplement the
typical combustion engine and 12V battery in conventional vehicles; the 48V battery and
associated network can power a variety of components (e.g., electric turbos, EGR pumps, AC
compressors, heated catalysts, cooling fans, oil pumps, coolant pumps), along with supplying
power to active chassis systems and regenerative braking.79 In contrast, NACFE defines
"strong" hybrids as those that utilize a larger electric machine paired with a smaller engine and
larger battery pack (relative to mild hybrids); strong hybrids generally use 300 to 500V electrical
batteries, and thus can propel the vehicle down the roadway at moderate speeds without engine
power. 74 Across the spectrum of hybrid technologies, there is a wide variety of powertrain
configurations that can range from electrifying an axle or adding a 48V battery for accessory
loads (mild hybrids) to a powertrain designed such that the electrical propulsion is able to meet
the duty-cycle demands in nearly all driving conditions (strong, plug-in hybrid).74 Mild hybrid
configurations that may be of growing interest in the industry include the use of an electric
motor/generator, regenerative braking, electric boosts and advanced batteries, as well as
advanced stop-start systems that employ a 48V belt-driven starter-generator.79 Detailed
descriptions of different hybrid configurations and components is available in the National
Academies of Sciences report, "Reducing Fuel Consumption and Greenhouse Gas Emissions of
Medium- and Heavy-Duty Vehicles, Phase Two: Final Report".75
1.4.1.2 Emissions Performance
Emissions impacts of heavy-duty hybrids vary across this spectrum of mild to strong, as well
as with the level of integration and design of the technology. Heavy-duty hybrid technologies
have the potential to decrease or increase NOx emissions depending on how they are designed.
For example, a hybrid system can reduce NOx emissions if it eliminates idle operation or uses
the recovered electrical energy to heat aftertreatment components. In contrast, data show that
some hybrid technologies can produce higher engine-out NOx emissions due to higher engine
speeds combined with lower torque relative to a conventional engine; lowering the torque (load)
on the engine can also reduce the engine's ability to maintain sufficiently high aftertreatment
temperatures during low-load operation.76 The combination of higher engine out NOx with lower
aftertreatment efficiency resulted in up to 50% higher NOx emissions in MY2010 to 2012 heavy-
duty hybrid vehicles.76 Improvements in integration and design of hybrid powertrain and engine
components can address the potential for elevated NOx emissions from hybrid vehicles. EPA
0 EPA defines "mild hybrid" as a hybrid engine or powertrain with regenerative braking capability where the system
recovers less than 20 percent of the total braking energy over the transient cycle defined in Appendix I of 40 CFR
part 1037. (40 CFR 1036.801)
51
-------
recently worked with stakeholders to adapt a test procedure that was initially developed for HD
vehicles such that it now allows hybrid powertrains to be certified to the HD GHG engine
standards; this powertrain test procedure was finalized in the HD GHG Phase 2 technical
amendments rule (86 FR 34321, June 29, 2021). Under this proposed rulemaking, EPA is
proposing that this powertrain test procedure can also be to meet the criteria pollutant engine
standards. See preamble Section III.B.2.V for additional discussion on this proposal, and Chapter
2.1.1.1.3 details on the powertrain test procedure.
We did not include heavy-duty hybrid technologies in the proposed rulemaking's emissions
inventory analysis. Hybrids were excluded for two reasons: 1) the variability in heavy-duty
hybrid emissions based on design, integration, and use of the technology (outlined above), and 2)
uncertainty in the which types of hybrids will grow in which segments of the heavy-duty market
(discussed further below).
1.4.1.3 Current and Future Markets
As noted above mild heavy-duty hybrid powertrain options such as 48V battery systems to
electrify certain components or electrified axles are of increasing interest to heavy-duty vehicle
market.79 According to a recent NACFE report, there were no Class 3 through 8 HEV models
available in late 2019, but at least one HEV Class 5 delivery truck is now available.74'77 A variety
of hybrid transit buses are available and have been in production since the early 2000s 74
The choice to use a hybrid technology in heavy-duty delivery vehicles will depend on a
variety of factors that fall into several categories, including: duty cycle (e.g., length of typical
route, freight weight requirements), regional requirements (e.g., zero emissions zones, weather
and terrain conditions), and costs (e.g., capital costs, operational costs of fuel and equipment).74
Some data suggest that these and other factors may be leading fleets to use and plan for hybrid
heavy-duty vehicles. In 2019, more than 30% of fleets responding to the National Truck and
Equipment Association (NTEA) fleet purchasing survey stated that they currently operate
alternative-fuel trucks, with electric hybrid being one of the predominate alternative-fuel
options.78 In 2019, 19% of survey respondents indicated that they are electrifying their systems,
with 77% of those respondents indicating that they intend to expand their use of electrification.78
Of fleets electrifying their vehicles, the telecom/utility market shows a higher electrification
penetration rate than other markets according to NTEA 2019 survey responses. Truck mounted
equipment (e.g., bucket and crane operations) are the most commonly electrified systems,
although that may be shifting towards "environmental control" systems.p'78 Idle reduction is
another area of growth within the use of heavy-duty hybrid technology, with 56% of NTEA
survey respondents indicating they are incorporating idle reduction technology in 2019. Fleets
choose to purchase vehicles with idle reduction, electrification, or other alternative fuel options
based primarily on reductions in operating and life cycle costs compared to conventional
vehicles. 78 Tax or grant incentives, payback period, and the ability to lease vehicles also play a
role in fleets' purchasing decisions on alternatively fueled trucks.*2'78
p The category "environmental control systems" is not defined in NTEA survey results.
Q Note that "automation/advanced technology features" were also included in the survey question on purchasing
drivers for alternatively fueled trucks (NTEA 2019)
52
-------
Hybrid technology in the heavy-duty industry is projected to increase significantly over the
next several years as a result of the HD Phase 2 GHG standards, as well as higher electrical
demands from safety and entertainment features.79 Members of the Manufacturers of Emissions
Control Association (MECA) project that 48V systems as potentially feasible by 2024 for some
engine families, with a wider range of offerings available by 2027.79 Mild-hybridization with a
48V system could provide a synergistic benefit with the CDA technology discussed in Section
1.1.3 above, as well as other engine management strategies (e.g., start-stop capability)79 The
total cost of ownership (TCO) for hybrid technologies, and its relation to diesel vehicles, will
vary based on the specifics of the hybrid system (e.g., battery size, battery voltage, motor power
and level of integration).
1.4.2 Battery-Electric and Fuel Cell
1.4.2.1 Technology description
Similar to light-duty battery-electric vehicles, heavy-duty battery electric vehicles (BEVs)
utilize a traction battery pack to store electricity for use by an electric traction motor, which
transfers mechanical energy to an electric transmission in order to drive the wheels of the
vehicle.74 Fuel cell electric vehicles (FCEV) utilize a proton-exchange membrane fuel cell
(PEMFC) stack that converts hydrogen and oxygen to electricity for use by an electric traction
motor, which transfers mechanical energy to an electric transmission in order to drive the wheels.
The FCEV is similar to a BEV in that it utilizes all-electric drive, but that electricity is derived
via an electrochemical reaction of hydrogen with oxygen in air across a membrane electrolyte
assembly.74 80 FCEVs are often similar to strong hybrids since a lithium-ion battery pack is
required in order to buffer power demand to the PEMFC and to provide regenerative braking.
FCEV powertrain architectures similar to a BEV using a PEMFC as a range extender are also
possible.81-82-83 High voltage electric machines are necessary to provide sufficient torque for the
full range of vehicle operation; thus bus voltage is also similar to BEV applications. RHEVs
Battery technology is rapidly evolving with different battery chemistries and battery pack
designs coming into the market.74 While manufacturers can build off experience with light-duty
electric vehicle batteries, there are differences in battery design for heavy-duty vehicles.
Considerations for heavy-duty vehicle battery design include: the energy-to-weight ratio, energy-
to-volume ratio, specific power (amount of current acceptable), battery lifetime (calendar years
and charge cycles), recharging time, temperature management (both cooling and heating), and
safety (both during use and at end-of-life).74-84-92 Detailed discussion of battery design is outside
the scope of this document, but we will briefly discuss a few key aspects of these considerations.
The energy to weight or volume ratios impact vehicle weight and the amount of cargo a vehicle
can carry, whereas specific power can influence charging time. Information on expectations for
battery lifetime in heavy-duty applications is nascent, but data from light-duty vehicles suggests
that batteries can operate at greater than 90% of the original capacity after approximately
150,000 miles (80% original capacity is a general target for end of useful life in a commercial
vehicle battery; total lifetime mileage of heavy-duty trucks varies by weight class and application
but can range from 250,000 to more than 800,000 miles [see Chapter 3 for more discussion on
mileage accumulation and Regulatory Useful Life]).92 Ambient temperatures can influence
R A number of PEMFC concepts under consideration for future HD truck applications use a larger battery pack and
are technically PHEVs, so they would, in that case, use grid electricity.
53
-------
battery life and performance, where increases in auxiliary power demand for cabin heating or
cooling combined with lower efficiency in battery chemistry result in reduced driving range.85
Once depleted, the amount of time required to recharge a battery will vary based not only on
battery specific power, but also the battery size (e.g., 100 kWh, 550 kWh), battery state of charge
(e.g., 20%, 50%), battery type, and type of charging equipment available.74'92 The current SAE
J1772 standard defines multiple charging levels. These include AC Level 1 (1.92 kW), AC
Level 2 (19.2kW, provisionally 22 kW), DC Level 1 (up to 48 kW) and DC Level 2 (up to 400
kW), all of which use the CCS1 connector within North America. With these variables in mind,
charging times for a medium duty electric truck would be longer on an AC Level 2 Charger (e.g.,
1 hour of AC Level 2 charging to go 10 to 20 mi) than on a fast charger at DC Level 2 (e.g., 20
minutes or less to go at least 60 mi). 104>86>87
Within the EU, ACEA has developed recommended specifications for opportunity charging,
or "OppCharge" using 150 kW to 600 kW automated overhead charging for urban bus
applications.88 The OppCharge system is already in service for urban bus charging in the EU.
Comparable overhead charging systems may also be applied to other BEV applications. In the
U.S., SAE has established a new standard, J3105, for overhead charging of heavy-duty vehicles,
including urban buses and heavy-duty trucks.89 The SAE J3105 standard provides two levels of
DC charging: J3105 Level 1 at up to 600 kW and J3105 Level 2 at up to 1.2 MW.
Tesla additionally claims a charge rate of up to 1.6 MW for their proprietary "Megacharger",
which is currently in development.90
The analogous infrastructure for a FCEV is a hydrogen refueling station, which can offer
refueling in time periods similar to filling a diesel fuel tank.s'74 The extent to which
refueling/recharging time matters will vary for different applications of heavy-duty vehicles
(e.g., delivery vans parked in a distribution center overnight versus a long-haul truck that
refuels/recharges en route to a destination).
The choice between a BEV or FCEV powertrain technology for a particular heavy-duty
vehicle application will likely depend on the specifics of an array of factors relevant to the
particular duty-cycle and operations of an individual owner-operator or fleet.74 For instance,
considerations such as charging/refueling infrastructure access, charging/refueling time, energy
conversion efficiency, vehicle weight, payload capacity, operational range, safety, noise levels,
and other driver preferences may influence the extent to which individual fleets select a BEV or
FCEV.91'92'74 Within each of these considerations there can be a spectrum of actual vehicle
characteristics that an individual fleet or purchaser would want to consider.1 For instance, for
vehicle weight, BEVs are generally expected to be heavier than diesel trucks or FCEVs due to
the larger, heavier battery; however, ICE, FCEV, and BEV weight all vary based on the specific
configuration of a model and the duty cycle it is intended to serve.74 In urban delivery
applications where a relatively smaller, lighter weight battery-pack could be used to drive a BEV
100 miles before charging, the additional vehicle weight from a battery-pack in a BEV may
sIn 2019, refueling a FCEV to go 350 miles was estimated as taking 15 minutes compared to 5 minutes to refill a
diesel fuel tank for the same trip length; note that hydrogen refilling stations can supply hydrogen at 350 or 700 bar,
which can significantly impact the amount of time required to refill (NACFE 2019)
T Further discussion on factors such as noise levels, vehicle weight, payload capacity, operational range,
charging/refueling time, energy conversion efficiency, safety, and other driver preferences is available in guidance
reports by NACFE (NACFE Class7/8Report, BEVs where they make sense, MD TCO report).
54
-------
matter less, particularly when the purchaser takes into account the lack of weight from diesel
fuel, diesel exhaust fluid, emissions control technology, exhaust systems, and other components
that are not needed in a BEV.74'92 The duty cycle of urban delivery vehicles can also fit well
with the operational characteristics of a BEV, where trucks could charge at the same location
after completing their delivery route for the day and additional energy is recovered throughout
the day from regenerative braking in stop-and-go traffic.74'92 For similar reasons, regional haul
applications in which a truck travels 300 miles or less per day may also be well-suited for BEV
technology.93'74 Transit and school buses are other applications that may be well-suited for BEV
technology given frequent regenerative braking opportunities and well-defined routes.104 In
contrast, FCEVs may be more well-suited for long-haul applications due the higher percentage of
time in steady-state highway operations and longer-range capabilities at a vehicle weight
comparable to an ICE.74 The specificity with which BEVs and FCEVs may be built for certain
applications (e.g., smaller, lighter battery packs in urban delivery trucks compared to larger
battery packs in regional haul trucks; better steady-state performance of fuel cell technology in
long-haul applications) may be seen as continuation of the trend towards greater specialization of
heavy-duty diesel trucks over the past several years (e.g., optimizing aerodynamics for long
haul); however, greater vehicle specialization for a particular application may limit the sale of
the vehicle to second and third owners.74 On the other hand, technology improvements, such as
decreases in battery costs and greater availability of DC Fast chargers, could minimize
differences between BEV models, as well as between BEVs and conventionally fueled trucks.
Improvements in fuel cell technology could similarly minimize cost differences between FCEVs
and diesel trucks.
1.4.2.2 Emissions Performance
BEVs and FCEVs do not have internal combustion engines (ICEs), and thus have zero
tailpipe emissions of NOx.u The absence of tailpipe emissions means that the emissions
performance of these heavy-duty vehicles are not dependent on the maintenance and repair of
emissions control systems, as is the case for most types of hybrids, and all ICEs. Nevertheless,
emissions from producing electricity to recharge the vehicle battery, in the case of BEVs, or
emissions from producing hydrogen, in the case of FCEVs, are part of the emissions profile for
these vehicles.v Comparisons of lifecycle emissions from BEVs, FCEVs, and ICEs vary in terms
of assumptions and conclusions, but generally suggest lower lifecycle emissions associated with
BEVs and FCEVs relative to ICEs, particularly if electricity or hydrogen fuel is generated from
renewable sources (e.g., wind, solar, biomethane, hydroelectric). w>74>104
The emissions impacts of BEVs and FCEVs are not included in the proposed rulemaking's
emissions inventory analysis because we anticipate that manufacturers will meet the proposed
standards with diesel and gasoline engine technologies. However, we recognize that BEV or
u Note that there have also been experimental bus applications using solid oxide fuel cells (SOFC) instead of
PEMFCs; SOFCs reform liquid or gaseous fossil fuels and do have tailpipe emissions.
v Other lifecycle emissions, such as those due to producing battery packs, hydrogen fuel cells, and other vehicle
components are also important to consider when looking at emissions performance over the life a vehicle, but these
factors are considered outside the scope of this analysis.
w The source and production process used to supply energy for powering BEV charging or FCEV hydrogen
production have large impacts on lifecycle emissions. For instance, hydrogen production through grid electrolysis is
very energy intensive, which could result in higher emissions. BEV charging powered by coal-driven power plants
could similarly increase lifecycle NOx emissions relative to ICEs.
55
-------
FCEV technologies are increasingly being introduced into the heavy-duty market. As such, we
were interested in evaluating the emissions impacts of BEV and FCEV technologies entering the
heavy-duty market due to market forces alone (i.e., not due to manufacturers changing their
production plans due to a change in criteria pollutant standards). As discussed in Section 0
below, there is a wide range of projections for BEV and FCEV sales in the 2027 through 2045
timeframe. Due to the wide range, we conducted a sensitivity analysis to evaluate the potential
impacts of BEV and FCEV on our emissions estimates. This sensitivity analysis is briefly
described here, with results presented in Chapter 5 Appendix 5.
The BEV and FCEV sensitivity analysis was conducted through post-processing MOVES
emissions inventories for the Baseline, the proposed Option 1 and 2, and the Alternative (i.e.,
each scenario). Specifically, we used a ratio of BEV and FCEV sales over total sales in calendar
years 2027 and 2045 to adjust each scenario in the proposed rulemaking using Equation 1-1. For
vehicle sales data we used estimates from the 2018 Energy Information Administration (EIA)
and a 2017 National Energy Laboratory (NREL) study.94'95 The EIA data is consistent with what
is used in our main MOVES emissions inventory analysis, and likely reflects a conservative
estimate of future BEV and FCEV sales. The NREL study includes "Low", "Medium", and
"High" adoption scenarios for BEVs; we used the "Medium" adoption scenario to reflect a more
moderate estimate of future BEV sales relative to the EIA data. The NREL data does not include
FCEV sales data, and we were unable to identify a comparable data source for FCEV sales
volume projections in 2027 through 2045. If additional data on FCEV sales is available when we
are conducting analyses for the final rulemaking, then we would likely evaluate using those data.
Similarly, if EPA receives feedback that the NREL "High" scenario may be more appropriate for
reflecting BEV market adoption, then we would consider that for the final rule.
Equation 1-1
Emissions_adjusted i,j = Emissions i,j x (1-BEV FCEV Ratioj)
Where:
i: pollutant
j: sourcetype
BEV FCEV Ratio: Ratio of BEV and FCEV to total vehicle population
In both the EIA and NREL datasets, vehicle sales are categorized by light-, medium-, or
heavy-heavy-duty vehicles, or buses. We apportioned vehicle sales in each of these categories to
MOVES sourcetypes based on Table 3-3 in the MOVES Population and Activity Technical
Report.96 In general, light- and medium-heavy duty vehicle sales were assigned to vocational
sourcetypes in MOVES (e.g., refuse [51], single-unit short-haul truck [52], single-unit short-haul
truck [53]), while heavy-heavy-duty vehicle sales were assigned to long-haul sourcetypes (i.e.,
combination short-haul truck [61], combination long-haul truck [62]). For buses, the NREL
dataset provided transit bus sales, so no re-assignment was needed; however, EIA does not
provide explicit data for buses and thus total heavy-duty vehicles are typically used to map buses
MOVES sourcetypes. We recognize that using total heavy-duty vehicle sales is likely a
significant underestimate of BEV transit bus sales, but since we are using EIA data as a
conservative, lower bound, estimate we did not adjust the EIA data. Table 1-9 provides a
comparison of BEV sales in 2027 and 2045 from EIA and NREL data sources. Emissions
inventory results of the sensitivity analysis are provided in draft RIA Chapter 5, Appendix 5.
56
-------
Additional discussion on BEV and FCEV emissions performance, as it relates to the potential for
emissions credits, along with a request for comment on this topic is discussed in Section IV.I of
the Preamble.
Table 1-9: BEV & FCEV Sales Percentages
2027
2045
EIA Medium-heavy Duty (BEV & FCEV)
1%
2%
NREL Medium-heavy Duty (BEV only)
5%
28%
EIA Heavy-heavy Duty (BEV & FCEV)
0.4%
0.7%
NREL Heavy-heavy Duty (BEV only)
2%
10%
EIA Buses (BEV & FCEV only)
0.8%
1%
NREL Transit Buses (BEV only)
9%
48%
1.4.2.3 Current and Future Markets
In 2019, there were approximately 60 makes and models of BEVs available for purchase, with
additional product lines in prototype or other early development stages. 104'X'Y The FCEV market
in 2019 included three commercially available models, again with additional options in prototype
or other early development stages.74'104 Current production volumes of both BEVs and FCEVs
are small with the NACFE estimating fewer than 50 FCEVs and fewer than 100 BEVs Class 7/8
trucks in production in the US in 2019.74 While a number of prototypes and demonstration
projects are underway, particularly in ports in the state of California, they are generally in early
stages of establishing feasibility and durability of the technology, as well as building out
necessary infrastructure.74 Nevertheless, some manufacturers are building off of a wealth of
experience in battery-electric passenger car and/or transit bus production, which suggests the
potential for rapid adaptation and growth of BEVs in the heavy-duty market.74 Indeed, larger
numbers of BEVs are in production the medium-heavy duty market (relative to Class 7/8) since
urban delivery is seen as a duty-cycle well-suited for BEV performance.92'92'97 BEV technology
is also increasingly used in the transit bus market, with electric buses growing from 300 to 650 in
the US between 2018 to 2019.98'z Table 1-10 and Table 1-11 provide a snapshot of BEVs in the
medium- and heavy-duty truck, and bus markets as of 2019, according to one source; however,
given the dynamic nature of the BEV and FCEV market, the number and types of vehicles
available is changing fairly rapidly.99 Similar efforts to inventory available BEV and FCEV
x The composition of all-electric truck models was: 36 buses, 10 vocational trucks, 9 step vans, 3 tractors, 2 street
sweepers, and 1 refuse truck (Nadel and Jung (2020) citing AFDC (Alternative Fuels Data Center). 2018. "Average
Annual Vehicle Miles Traveled by Major Vehicle Categories." www.afdc.energy.gov/data/widgets/10309.)
Y Note that there are varying estimates of BEV and FCEV models in the market; NACFE 2019 provided slightly
lower estimates than the those included here from Nadel and Jung 2020. A recent NREL study suggests that there
may be more models available, but it is unclear how many are no longer on the market since the inventory includes
vehicles introduced and used in commerce starting in 2012 (Smith et al. 2019).
z Note that ICCT (2020) estimates 440 electric buses were sold in the US and Canada in 2019, with 10 of those
products being FCEV pilots. The difference in estimates of number of electric buses available in the US may lie in
different sources looking at production vs. sales of units.
57
-------
vehicles are available in recent reports by NREL, ACEEE, and ICCT, as well as an interactive
webtool developed by CALSTART.84'104'100'101
Table 1-10: BEV Truck Offerings or Planned Offerings in the US (as November 2019)AA
Manufacturer
Model
Range
Battery, H2 Capacity
Availability
Delivery Vans, Shuttles, and Straight Trucks
BYD
6F
124 miles
221 kWh
Today
Chanje
V8100
150 miles
100 kWh
Today
Freightliner (Daimler)
eM2
230 miles
325 kWh
Production in 2021
GreenPower
EV Star Cargo
150 miles
118 kWh
Today
International (Navistar)
eMV
250 miles
321 kWh
Production in 2021
Ford Transit 3 5 OLID
120 miles
86 kWh
Today
Lightning Systems
Ford E-450
Ford F-59
110 miles
110 miles
129 kWh
128 kWh
Today
Today
Chevy 6500XD
130 miles
192 kWh
Today
Lion
Lion8
Unknown
480 kWh
TBD
Mitsubishi Fuso (Daimler)
eCanter
80 miles
83 kWh
Demonstration
Ford E-450
100 miles
127 kWh
Today
Motiv
Ford F-53
125 miles
127 kWh
Today
Ford F-59
90 miles
127 kWh
Today
Peterbilt (Paccar)
Model 220EV
100 miles
148 kWh
Demonstration
Phoenix Motor Cars
Ford E-450
100 miles
105 kWh
Today
Rivian
Unknown
Unknown
Unknown
Deployment in 2021
Workhorse
C1000
125 miles
70 kWh
TBD
Xos
Medium Duty
200 miles
Unknown
Today
Tractor Trucks
BYD
8TT
125 miles
409 kWh
Today
Freightliner (Daimler)
eCascadia
250 miles
550 kWh
Production in 2021
Nikola
Nikola One (sleeper)
750 miles
80 kg H2
Production in 2022
Nikola Two (day)
400 miles
1,000 kWh
Production in 2022
Peterbilt (Paccar)
Model 579
250 miles
352 kWh
Demonstration
Tesla
Semi
500 miles
Unknown
Production in 2020
Toyota/Kenworth (Paccar)
T680
300 miles
60 kg H2
Demonstration
Volvo Group
VNR
Unknown
Unknown
Production in 2020
Xos
ET One
300 miles
Unknown
Demonstration
Refuse Trucks
BYD
6R
125 miles
221 kWh
Today
8R
56 miles
295 kWh
Today
Mack (Volvo Group)
LR
Unknown
Unknown
Demonstration
Peterbilt (Paccar)
Model 520EV
80 miles
315 kWh
Demonstration
Street Sweepers
Global Environmental Products
M4 - electric
M4 - fuel cell
240 kWh
20 kg
11 hours
10 hours
Today
Today
Note: Ranges represent the maximum values reported by manufacturers. Models available "today" are defined as vehicles
eligible for California's Hybrid and Zero-Emission Truck and Bus Voucher Incentive Project incentive funding, if not
otherwise commercially available for purchase. Parent companies are listed in parentheses. Manufacturer partnerships are
designated with a slash.
AA Table adapted from UCS (2019)"
58
-------
Table 1-11: BEV Bus Offerings or Planned Offerings in the US (as November 2019)BB
Manufacturer | Model | Range | Battery, H2 Capacity | Availability
School Buses
Blue Bird
MicroBird (Type A)
Vison (Type C)
All American (Type D)
100 miles
120 miles
120 miles
88 kWh
160 kWh
160 kWh
Today
Today
Today
GreenPower
Synapse 72 (Type D)
140 miles
200 kWh
Today
IC (Navistar)
chargE (Type C)
120 miles
260 kWh
TBD
Lion
Type A
TypeC
150 miles
155 miles
129 kWh
220 kWh
Today
Today
Motiv
Type A
TypeC
75 miles
90 miles
106 kWh
127 kWh
Today
Today
Thomas Built (Daimler)
Jouley (Type C)
120 miles
220 kWh
Today
Transit Buses
BYD
K7 (30')
K9 (35')
K9 (40')
Kll (60')
135 miles
215 miles
177 miles
230 miles
196 kWh
266 kWh
352 kWh
Unknown
Today
Today
Today
Today
Complete Coach Works
ZEPS
150 miles
311 kWh
Today
El Dorado National
AXESS Fuel Cell (35')
AXESS Fuel Cell (40')
260 miles
260 miles
50 kg of H2
50 kg of H2
Today
Today
Gillig
ePlus (29')
ePlus (35')
ePlus (40')
Unknown
Unknown
Unknown
296 kWh
444 kWh
444 kWh
Today
Today
Today
GreenPower
EV250 (30')
EV300 (35')
EV350 (40')
EV400 (45')
175 miles
175 miles
185 miles
185 miles
210 kWh
260 kWh
320 kWh
320 kWh
Today
Today
Today
Today
Lightning Systems
City Transit Bus
Repower
200 miles
320 kWh
Today
Lion
LionM (26')
150 miles
160 kWh
Today
New Flyer
Xcelsior(35')
Xcelsior (40')
Xcelsior (60')
Xcelsior Fuel Cell (40')
Xcelsior Fuel Cell (60')
260 miles
225 miles
135 miles
300 miles
300 miles
545 kWh
466 kWh
466 kWh
37.5 kg of H2
60 kg of H2
Today
Today
Today
Today
Today
Proterra
Catalyst (35')
Catalyst (40')
234 miles
328 miles
440 kWh
660 kWh
Today
Today
Coach Buses
BYD
C6M (23')
C8M(35')
C9M (40')
C10M (45')
124 miles
200 miles
200 miles
230 miles
121 kWh
353 kWh
353 kWh
446 kWh
Today
Today
Today
Today
MCI (New Flyer)
D45 CRTe LE CHARGE
J4500e CHARGE
Unknown
200 miles
388 kWh
450 kWh
Production in 2020
Production in 2020
VanHool/Proterra
CX45E
Unknown
Unknown
Production in 2020
Double Decker Buses
Alexander Dennis/Proterra
500EV
200 miles
660 kWh
TBD
BYD
C10MS
230 miles
446 kWh
Today
GreenPower
EV550
240 miles
478 kWh
Today
Yard Trucks
BYD
8Y
10 hours
217 kWh
Today
Kalamar
T2E
"3 shifts"
220 kWh
Today
Orange EV
T-Series
24 hours
160 kWh
Today
Terberg
YT202
62 miles
Unknown
Today
Note: Ranges represent the maximum values reported by manufacturers. Models available "today" are defined as vehicles
eligible for California's Hybrid and Zero-Emission Truck and Bus Voucher Incentive Project incentive funding, if not
otherwise commercially available for purchase. Parent companies are listed in parentheses. Manufacturer partnerships are
designated with a slash.
59
-------
BEV and FCEV powertrain use is expected to increase in the heavy-duty market. The rate of
growth ranges widely across forecasting models; for instance, the 2018 Annual Energy Outlook
shows medium-duty BEVs making up 0.08% of total truck vehicle miles traveled (VMT) in
2030, whereas a 2018 National Renewable Energy Laboratory (NREL) study suggests 29% for
the same time frame94'95 A variety of factors will influence the extent to which BEVs and
FCEVs are available for purchase and enter the market.91'92 NACFE looked at 22 factors on
which to compare BEVs and FCEVs with heavy-duty diesel vehicles; they found that for the
Class 7/8 market, a current lack of production vehicles resulted in BEVs and FCEVs performing
worse than diesels in 2019, but each performing equal to, or better than diesel on most factors by
2030.cc'74
The lifetime total cost of ownership (TCO), which includes maintenance and fuel costs, is
likely a primary factor for fleets considering BEV and/or FCEV purchases. In fact, a 2018 survey
of fleet owners showed "lower cost of ownership" as the second most important motivator for
electrifying their fleet.DD'104 An International Council for Clean Transportation (ICCT) analysis
suggests that TCO for light- and medium heavy-duty battery-electric vehicles could reach cost
parity with diesel in the early 2020s, while heavy heavy-duty battery-electric or hydrogen
vehicles are likely to reach cost parity with diesel closer to the 2030 timeframe.102 Recent
findings from Phadke et al. suggest that BEV TCO could be 13% less than that of a diesel truck
if electricity pricing is optimized.103 Their analysis further suggests that BEV TCO could be as
low 40% less than that of a diesel truck if battery costs reach $100/kWh, electricity is provided
through solar, wind, or other sources free of GHG-emissions, and environmental benefits of
BEVs are monetized.103
As both the ICCT and Phadke et al. study suggest, fuel costs are an important part of TCO.
While assumptions about vehicle weight and size can make straight comparisons between BEVs,
FCEVs, and ICEs challenging, data show greater energy efficiency of battery-electric technology
relative to an ICE, with comparable or better energy efficiency of FCEVs relative to an
ICE.EE'74'104 Better energy efficiency leads to better fuel economy and thus lower fuel costs for
BEVs and FCEVs relative to ICEs.74'104 Maintenance and service costs are also an important
component within TCO; although there is limited data available on actual maintenance costs for
BEVs and FCEVs, early experience with BEV medium heavy-duty vehicles and transit buses
suggests the potential for lower maintenance costs after an initial period of learning to refine
both component durability and maintenance procedures.74 To facilitate fleets transitioning to
BEVs, some manufacturers are currently including maintenance in leasing agreements with
BB Table adapted from UCS (2019)"
cc Factors that NACFE considered fell into the following categories: weight, cost, maintenance effort, vehicle life,
range, "fuel" availability, and general; for additional information on the factors and how they compare in 2019 and
2030, see NACFE (2019) "Guidance Report: Viable Class 7/8 Electric, Hybrid and Alternative Fuel Tractors",
available online at: https://nacfe.org/downloads/viable-class-7-8-alternative-vehicles/
DD The primary motivator for fleet managers was "Sustainability and environmental goals"; the survey was
conducted by UPS and GreenBiz.
EE Fuel costs here refers to diesel or gasoline for an ICE and electricity or hydrogen for a BEV or FCEV,
respectively.
60
-------
fleets; it is unclear the extent to which a full service leasing model will persist or is transitioned
to a more traditional purchase after an initial period of learning.105'106
The potential for lower fuel and maintenance costs to outweigh a higher upfront cost for
BEVs and FCEVs is reflected in ICCT and other's projections of BEVs and FCEVs reaching cost
parity with diesels within the next several years; however, the current upfront cost can exceed a
diesel vehicle by 60% or more104 Upfront purchase price was listed as the primary barrier to fleet
electrification in a 2017 survey of fleet managers, which suggests that state or local incentive
programs to offset BEV or FCEV purchase costs will play an important role in the near term,
with improvements in battery and hydrogen fuel cell costs playing a role in reducing costs in the
longer-term.FF'104
Another large factor in TCO for fleet adoption of BEVs and FCEVs is charging and refueling
infrastructure, respectively.104 In early 2020, there were over 24,000 publicly accessible EV
charging stations in the US, 3,400 of which were DC Fast Chargers (with the remaining being
Level 2 charging stations); the majority of these charging stations were built for light-duty
vehicles in terms of location and power availability, thus while some of the existing Level 2
charging might be suitable for overnight charging of medium-heavy-duty vehicles, DC Fast
Charging or chargers capable of even higher peak power demand will likely be necessary for
most heavy-duty vehicles, and many fleets will generally want to have their own charging
infrastructure onsite.107'74 For FCEVs, there were 50 public and private hydrogen refueling
stations in the US in early 2020, the majority of which are located in California.107 Building out
charging and refueling infrastructure will be a critical component of BEV and FCEV
technologies continuing to enter the heavy-duty market. In the case of BEV charging
infrastructure, fleets need to work through a number of considerations, such as: space and siting
requirements (e.g., space for trucks charging over long periods of time, proximity to electrical
equipment), electrical power infrastructure (e.g., amount of power that can safely be offered to a
site without distribution grid upgrades), their fleet's energy needs, charging station maintenance,
and energy pricing (e.g.,, peak demand charges during certain times of day).104'103 Programs to
support BEV charging infrastructure are increasingly being offered through utility or other state
programs, and NACFE has published a resource to help fleets navigate the charging
infrastructure procurement process. GG'104 As noted above, some manufacturers are offering lease
options that include maintenance; these leases can also include "fuel" (electricity or hydrogen).
Nikola Motors for instance is including hydrogen refueling in the cost of leasing one of its future
FCEVs and plans to build 700 hydrogen refueling stations across the US and Canada by 2028.108
Notably, Phadke et al. find that charging cost has the largest impact on the payback period for a
BEV relative to a diesel truck.103'™'109'110
FF Other barriers that fleet managers prioritized for fleet electrification included: Inadequate charging infrastructure-
our facilities, inadequate product availability, inadequate charging infrastructure- public; for the full list of top
barriers see Nadel and Junga (2020), citing UPS and GreenBiz 2018.
GG Further discussion on factors such as noise levels, vehicle weight, payload capacity, operational range,
charging/refueling time, energy conversion efficiency, safety, and other driver preferences is available in guidance
reports by NACFE (NACFE Class7/8Report, BEVs where they make sense, MD TCO report).
1111 Note that alternative charging technologies, such as wireless or overhead charging, are being explored (see ICCT
2019 and ICCT 2017 for more discussion)
61
-------
Driver preferences such as acceleration speed and noise levels can also influence fleet
decisions to purchase BEVs and/or FCEVs.92'104 Electric motors offer a faster acceleration from
stop and faster speed going up grade, which may be attractive to delivery drivers looking to
minimize their route time.104 Similarly, BEVs are generally thought to operate at lower noise
levels relative to ICE vehicles, which could be attractive for both drivers and local
communities.104 Furthermore, BEVs and FCEVs generate occupational health benefits by
reducing driver exposure to heavy-duty vehicle emissions while on duty. State and local
activities, such as the Advanced Clean Trucks (ACT) rulemaking recently finalized in California
could also influence the market trajectory for battery-electric and fuel cell technologies.111 The
ACT requires manufacturers to sell a certain percentage of zero emission heavy-duty vehicles
(BEVs or FCEVs) for each model year, starting in MY 2024. The sales requirements vary by
vehicle class, but start at 5 to 9 percent of total MY 2024 heavy-duty vehicle sales in California
and increase up to 40 to 75 percent of MY 2035 and beyond sales.112 Outside of California,
many states also have fleet incentive programs in the form of grants, tax incentives, rebates,
exemptions, etc. that may help to accelerate fleet purchases of BEVs or FCEVs.74 In July 2020,
fifteen states and the District of Columbia signed a Memorandum of Understanding affirming a
commitment to strive towards at least 30 percent of new heavy-duty vehicle sales being zero
emission vehicles by 2030 and to reach 100 percent of new sales by 2050.113 In addition, cities
such as New York, Los Angeles, and San Francisco have committed to transitioning their transit
bus fleets to all electric in the 2030 to 2040 timeframe.104
62
-------
Chapter 2 Compliance Provisions
2.1 Compression-Ignition Engine Dynamometer Test Procedures
Test procedures are a crucial aspect of the heavy-duty criteria pollutant program. This
rulemaking is proposing to establish several new test procedures for spark and compression
ignition engine compliance. This chapter will describe the existing test procedures as well as the
development process for the test procedures being proposed. This includes the determination of
emissions from both engines and hybrid powertrains as well as the development of new duty
cycles.
2.1.1 Current CI Test procedures
Heavy-duty compression-ignition engines currently are certified for non-greenhouse gas
(GHG) pollutants using the Heavy-Duty Diesel Engine Federal Test Procedure (HDDE FTP) and
Supplemental Emission Test Ramped Modal Cycle (SET). 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 Part 86. The HDDE
FTP is defined in 40 CFR part 86 Appendix I. The SET is defined in 40 CFR 86.1362(a). All
emission measurements and calculations are defined in 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 CO is described in 40 CFR 1065.250. For measurement of
NMHC, refer to 40 CFR 1065.260. For measurement of NOx, refer to 40 CFR 1065.270. For
measurement of PM, refer to 40 CFR 1065.140, 1065.170, and 1065.290. Table 1 of 40 CFR
1065.205 provides performance specifications that we recommend analyzers meet. Note that 40
CFR 1065.307 provides linearity verifications that the system must meet. For the calculation
method for brake specific mass emissions for CO, NMHC, NOx and PM, refer to 40 CFR
1065.650.
2.1.1.1 HDDE FTP
The Heavy-Duty Diesel Engine Federal Test Procedure (HDDE 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. This procedure is well-
defined, mirrors in-use operating parameters, and continues to be appropriate for assessment of
criteria pollutant emissions from heavy duty engines.
A complete HDDE FTP involves three test sequences. First, a 20-minute test is run over the
duty-cycle with the engine at the same ambient temperature as the test cell (between 68°F and
86°F). The engine undergoes a 10-minute hot-soak following the cold-start. A 20-minute hot
start test is run over the duty-cycle following the hot-soak. The HDDE FTP emission level for
the engine is determined by weighting the cold start emissions by 1/7 (about 14 percent) and the
hot-start emission results by 6/7 (about 86 percent).
63
-------
2.1.1.2 SET
The Supplemental Emission Test Ramped Modal Cycle (SET) is a 13-mode, steady-state
engine dynamometer test that replaced the steady-state modal run SET and is based on the
European Stationary Cycle (ESC). The engine is tested on an engine dynamometer over a
sequence of steady-state modes. Emissions are collected over both the steady-state and transition
portions (20-second transitions) of the test and the results are integrated to produce a single
emission result to show compliance with the standard.
The current weighting of modes within the SET for engines complying with the 2010 NOx
and Phase 1 GHG standards is given in Table 2-1. A, B, and C speeds are determined according
to 40 CFR 1065.610.
Table 2-1: SET Mode Weighting Factors for the 2010 NOx and Phase 1 GHG Standards
Speed, % Load
Weighting factors of SET
(%)
Idle
15
A, 100
8
B, 50
10
B, 75
10
A, 50
5
A, 75
5
A, 25
5
B, 100
9
B, 25
10
C, 100
8
C, 25
5
C, 75
5
C, 50
5
Total
100
Idle Speed
15
Total A Speed
23
Total B Speed
39
Total C Speed
23
2.1.1.3 Powertrain
Powertrain test procedures were created under EPA's Heavy-duty Greenhouse Gas: Phase 2
(Phase 2) rulemaking for vehicle certification.114'115 At the time of their development, no
certification procedure existed for powertrain certification of heavy-duty hybrid vehicles to any
engine standards. The powertrain test was updated for powertrain certification of the engine to
engine standards for GHG pollutants in the Heavy-Duty Engine and Vehicle Test Procedures,
and Other Technical Amendments rulemaking (hereinafter "HD Technical Amendments").116
The powertrain certification test was finalized for certification to both the FTP and SET and is
carried out by following 40 CFR 1037.550 as described in 40 CFR 1036.505 and 40 CFR
64
-------
1036.510 and is applicable for powertrain systems with the hybrid function located in the P0, PI,
P2, and P3 positions.
The development of these test procedures required the addition of a speed and road grade
profile to the existing HDDE FTP, HDOC FTP, and newly proposed LLC duty-cycles in the
recently finalized Appendix II of 40 CFR part 1036 and to the SET in 40 CFR 1036.505.116 It
also required the development of vehicle parameters to be used in place of those in 40 CFR
1037.550; namely vehicle test mass, vehicle frontal area, vehicle drag area, vehicle coefficient of
rolling resistance, drive axle ratio, tire radius, vehicle curb mass, and linear equivalent mass of
rotational moment of inertias. Determination of system and continuous rated power along with
the maximum vehicle speed (C speed) is also required using 40 CFR 1036.527. The
combination of the generic vehicle parameters, the engine duty-cycle vehicle speed profile, and
road grade profile fully defines the system load and this is designed to match up the powertrain
load with the HDDE vFTP, HDOC vFTP, vSET, and vLLC load for an equally powered engine.
The development of these test procedures was previously described in detail in the HD
Technical Amendments.116
2.1.2 Potential updates to CI Test procedures
2.1.2.1 HDDE FTP
We are proposing no changes to the HDDE FTP weighting factors or the duty-cycle torque
values from the duty-cycle that currently applies to criteria pollutant regulations in 40 CFR part
86 Appendix I (f)(2). We are proposing a change to the speed values that does not influence the
ultimate denormalized speed, as noted below and finalized in the HD Technical Amendments,
apply to criteria pollutant certification as well. We started the migration of some heavy-duty
highway engine standard setting part test procedures to 40 CFR part 1036 in the HD Technical
Amendments. This included the migration of the HDDE FTP drive schedule to Appendix II (c)
of part 1036 in order to add vehicle speed and road grade to the duty-cycle to facilitate
powertrain testing of hybrids for compliance with the Phase 2 GHG standards.
The change that was made for GHG and that is being proposed to apply to criteria pollutant
certification as well took the normalized vehicle speeds over the HDDE FTP duty-cycle and
multiplied them by 100/112 to eliminate the need to divide by 112 in the diesel engine
denormalization equation in 40 CFR 86.1333(a)(l)(i). This eliminated the need for inclusion of
a denormalization equation in the standard setting part and allows commonization of (between
compression and spark ignition engines) the use of the denormalization equation in 40 CFR
1065.610(c)(1) (equation 1065.610-3), with no effect on stringency.
2.1.2.2 SET
The SET weighting currently used for certification of heavy-duty highway compression
ignition engines for criteria pollutants and Phase 1 GHG has a relatively large weighting in C
speed. The C speed is typically in the range of 1800 rpm for current heavy heavy-duty engine
designs. However, it is becoming much less common for engines to operate at such a high speed
in real-world driving conditions, especially during cruise vehicle speeds in the 55 to 65 mph
vehicle speed range. This trend has been corroborated by engine manufacturers' in-use data that
has been submitted to the agencies in comments and presented at technical conferences.117 Thus,
although the current criteria pollutant and HD Phase 1 GHG SET represents highway operation
65
-------
better than the FTP cycle, improvements have been made via the HD Phase 2 GHG program by
adjusting its weighting factors to better reflect modern trends in in-use engine operation. The
most recent trends for compliance with the Phase 2 GHG standards indicate that manufacturers
are configuring drivetrains to operate engines at speeds down to a range of 1050-1200 rpm at a
vehicle speed of 65mph.
To address this trend toward in-use engine down-speeding, the agencies are proposing to
apply the refined SET weighting factors and resulting SET developed for the HD Phase 2 GHG
standards to the proposed rulemaking criteria pollutant standards. The Phase 2 GHG SET mode
weightings move most of the C weighting to the A speed, as shown in Table 2-2. To better align
with in-use data, these changes also include a reduction of the idle speed weighting factor. This
proposal would apply the Phase 2 mode weightings to both criteria pollutants and the Phase 2
CO2 emission and fuel consumption standards starting in model year 2027.
Table 2-2: New SET Mode Weighting Factors in Phase 2
Speed/% Load
Weighting Factor in Phase 2
(%)
Idle
12
A, 100
9
B, 50
10
B, 75
10
A, 50
12
A, 75
12
A, 25
12
B, 100
9
B, 25
9
c, 100
2
C, 25
1
C, 75
1
C, 50
1
Total
100
Idle Speed
12
Total A Speed
45
Total B Speed
38
Total C Speed
5
The Phase 2 SET mode weighting moves most of the C speed weighting to A speed and
reduces the weighting factor on idle speed. These values are based on data from vehicle
manufacturers that have been claimed as confidential business information. These revised SET
weighting factors better reflect the lower engine speed operation of modern engines, which
frequently occurs at tractor cruise speeds.
To evaluate how current engines perform on this cycle, we tested a 2018 Detroit DD15 and a
2018 Cummins B6.7. For both engines these engines there was no significant difference for any
of the measured criteria pollutants between the two cycles. These results are summarized in
66
-------
Chapter 4.1.2 of the RIA. To assess the effect of stringency between the two SET cycles, the
CARB Stage 3 demonstration engine was also run on both versions of the SET. The results from
these tests can be found in Chapter 4.1.1.1 of the draft RIA.
2.1.2.3 LLC
Current certifications cycles (FTP and SET) and in-use not-to-exceed (NTE) compliance
requirements do not account for emissions over sustained low load operation. This is either
because the idle time in the duty-cycle is too short, or, in-use, the operation is excluded from
compliance requirements.
We are proposing a new low load certification cycle to address deficiencies in our current
certification duty-cycles and NTE in-use testing program with respect to emission control at low
load.
The National Renewable Energy Lab (NREL) and Southwest Research Institute (SwRI) under
contract with the California Air Resources Board (CARB) developed a suite of candidate low
load cycles (LLC) from urban tractor and vocational vehicle real-world activity data. The goal
of the cycle development was to develop a duty-cycle that was representative of real-world urban
tractor and vocational vehicle operations that are characterized by low engine loads, have
average power and duration adequate for demonstrating that hardware and controls needed to
deal with low load challenges are present and functional, and set an emission standard that
balances the need for NOx emission reductions and any associated GHG emission
impacts.118'119'120'121
NREL combined their Fleet DNA and CARB's heavy-duty diesel vehicle activity datasets,
incorporating a total of 751 unique vehicles across the United States, to develop the LLC. The
combined dataset included vehicles form 25 distinct locations, 26 combined vocational
designations, and 55 unique fleets incorporating both urban tractor and vocational applications.
A breakdown of the applications that were include can be found in Table 2-3.
67
-------
Table 2-3: Breakdown of vehicles from combined NREL Fleet DNA and CARB datascts.
Vehicle Application
Number of Vehicles
Parcel delivery
100
Refuse pickup
90
Line Haul
84
Beverage delivery
65
Mass transit
61
Food delivery
60
Drayage
41
Utility
32
Linen delivery
30
Transfer truck
29
Tanker
25
Telecom
24
Freight
22
Public work
13
School bus
11
Agricultural
10
Snow plow
9
Warehouse delivery
9
Construction
8
Dump truck
7
Refrigerated truck
6
Local delivery
5
Towing
4
Concrete
3
Dry van
3
Delivery
1
NREL initially developed a list of drive cycle metrics, including engine load specific
calculations to describe engine/vehicle operation. Vehicle operation was then broken down into
microtrips for operation over a given shift-day, where a single microtrip was defined as the
duration over which a vehicle speed increases from 0 mph to the time where the vehicle stops,
including operation until the vehicle speed starts to increase above 0 mph again. 10 microtrips
were averaged using moving average windows based on a sensitivity analysis of moving average
windows of 5, 10, and 15 microtrips. The total number of microtrips from the combined data
sets was approximately 1.25 million and resulted in a trimodal distribution with two large
primary peaks and a tertiary lower load peak as shown in Figure 2-1.
68
-------
30000
% average load
Figure 2-1: Window size comparison and load distribution profile.
The microtrips were then equated to operational profiles and profiles that contained average
loads of 20% or less were considered further for construction of the LLC. These remaining
profiles were then subjected to cluster analysi s to identify unique groups of operation and to
assess for outliers. K-means clustering was chosen to be applied to the dataset due to its
computational efficiency and ability to identify and remove outliers during pre-processing.
Elbow analysis was then applied to determine optimal cluster number. The resulting optimal
cluster number was three.
Representative profiles from each cluster were then selected for Greenhouse Gas Emissions
Model (GEM) simulation. Results for each cluster were ranked based on their distance to cluster
center to identify the most representative profiles. Profiles were examined for behavior and final
suitability for testing, starting with profiles closest to cluster center. Five primary modes of
operation were generally observed in the low load profiles: sustained low load, long idle,
motoring/short idle cooling (high to low load), post-cooling breakthrough (low to high load), and
mid-speed cruise motoring. Profiles found to have outlying behavior were removed and not used
for GEM modeling. These outliers were found to have one or more of the following: prolonged
periods of idle, long key off periods, and missing data. The load data broadcast by engines is not
accurate enough to allow translation of vehicle-based to engine-based profiles to create the
engine duty-cycle, so the Phase 2 GEM simulation model was used to develop the normalized
engine duty-cycle. A representative summary of ten GEM generated profiles is shown in Table
2-4.
69
-------
Initial LLC candidate duty-cycles were constructed using DRIVE to include at least one
example of each of the five primary modes of operation, incorporating five of the GEM
generated duty-cycles. It should be noted that these candidate cycles did not always include the
entire GEM generated profile if the candidate cycle could be completed in a shorter amount of
time by removing portions of the profile that did not adversely affect the target modes of
operation. Figure 2-2 gives an example of a candidate cycle incorporating representative profiles
v9892 cO, vl 1660 c5, v073 cl, v9892 cl, and vl 1806 c5.
CARB narrowed the range of candidate LLC duty-cycles to the three that best represented the
target modes of low load. The three were LLC Candidates 7, 8, and 10. Candidate cycle 7 is 90
minutes in duration, has 30 minutes of sustained low load operation, and retains the v073 cl mid-
speed cruise/motoring segment. Candidate cycle 8 is 81 minutes in duration, has 30 minutes of
sustained low load operation, and has a shorter v073 cl mid-speed cruise/motoring segment to
assess breakthrough only. Candidate cycle 10 is 70 minutes in duration, has 20 minutes of
sustained low load operation, and has a shorter v073 cl mid-speed cruise/motoring segment to
assess breakthrough only. Emission results for a representative engine compliant to EPA 2010
NOx standard for the three-candidate duty-cycles can be found in Table 2-5.
CARB recently reported that it selected candidate cycle 7, with an option to add auxiliary load
(1.5, 2.5, and 3.5 kW for LHD, MHD, and HHD engines respectfully), as the LLC that it plans to
move forward with in its upcoming heavy-duty low NOx program. This cycle was chosen by
CARB for inclusion of the highest percentage of non-idle operation when compared to the other
two candidate cycles. CARB's decision to move forward with the option to add auxiliary load
was based on making the duty-cycle more realistic with respect to real world operation as it
affords more effective function of technologies such as cylinder deactivation (CD A) during idle.
We agree with CARB's assessment of the candidate low load cycles and their adoption of
cycle 7. Thus we are proposing to adopt LLC 7, however we are proposing to require the use of
auxiliary load, to bolster the current FTP and SET duty-cycles and to better align laboratory
duty-cycles with the proposed changes we are making to in-use testing and compliance
requirements, which afford better low load NOx reduction. We are requiring the use of auxiliary
load because this results in more representative testing with respect to in-use operation. This
duty-cycle also contains vehicle speed and road grade profiles to facilitate powertrain
certification of hybrid powertrains to the engine standard. These profiles were developed in the
same manner as the HDDE FTP and HDOC FTP as well as the SET as discussed in the HD
Technical Amendments.116 The LLC can be found in 40 CFR part 1036 Appendix 11(d).
70
-------
Table 2-4: Representative Summary of GEM Generated Profiles for the Engine Duty-cycle
Profile
Vehicle
Cluster
Length
Avg
% Speed
Avg
% Torque
Repeats in
SwRI Test Runs
Class
Chassis
Engine
Trans
Gears
Vocation
1
\9892
0
800
26.9
6.9
4
8
4x2
Volvo D13
AMT
12
Food Service
2
v 11660
0
1295
21.4
6.6
3
8
6x4
Mack MP8-415C
MT
13
Dray age
3
v075
0
1130
26.3
7.4
3
8
6x4
Mack MP8-415C
AMT
10
~ravage
4
\11815
1
1949
11.5
8.8
3
8
6x4
Cummins ISX 15
MT
13
Transfer Truck
5
vll646
1
904
15.9
10.7
4
4
4x2
Cummins ISB 6.7
AT
6
Parcel Delivery
6
v073
1
1410
33.8
18.1
3
8
6x4
Mack MP8-415C
AMT
10
D ravage
7
V9892
1
1616
27
10.6
3
8
4x2
Volvo D13
AMT
12
Food Service
8
V11660
5
615
16.2
3.5
4
8
6x4
Mack MP8-415C
MT
13
~ravage
9
V11806
5
1810
7.5
6.8
3
8
6x4
Cummins ISX 12
AMT
10
Transfer Truck
10
f1X817
5
739
15.3
7.7
4
8
6x4
Cummins ISM 11
AMT
10
Transfer Truck
v9892_c0 vll660_c5 v073_cl v9892_cl vll806_c5
I * y * ii * y *— v * i
Speed ^—Torque
100
•a
-------
Table 2-5: NOx Emission Levels for a 2010 Compliant Engine on Three Candidate LLCs
Candidate #
Cycle Duration
NOx Conversion
Efficiency (%)
Engine Out NOx
(g/hp-hr)
Tailpipe NOx
(g/hp-hr)
7
90
74
3.2
0.8
8
81
77
2.9
0.7
10
70
69
3.2
1.0
2.1.2.4 Powertrain
We are proposing to allow the use of powertrain testing as an option for certification of hybrid
powertrains to criteria pollutant standards. It is envisioned that this will capture CO2 and NOx
co-optimization benefits for hybrid integration; including engine start/stop, electric motor assist,
electric vehicle mode, and brake energy recovery. The powertrain test procedures to facilitate
this type of certification testing have already been developed during the HD Technical
Amendments and are directly applicable to criteria pollutant testing. Duty-cycles that
incorporate vehicle speed and road grade for the HDDE FTP, HDOC FTP, and the SET were
finalized in the HD Technical Amendments where development of these cycles was discussed.116
The proposed addition of the low load cycle in Chapter 2.1.2.3 also included a discussion of the
vehicle speed and road grade profile development for powertrain testing.
2.2 Manufacturer-Run In-Use Testing Program for Compression-Ignition Engines
The manufacturer run in-use testing program is crucial to ensuring compliance with the
heavy-duty criteria pollutant program. This rulemaking is proposing to establish a new test
procedure for evaluating off-cycle compression ignition engine compliance. This chapter will
describe the existing test procedure as well as the development process for the test procedure
being proposed.
2.2.1 Current In-Use Program and Standards
EPA's current regulatory program for on-highway heavy-duty engines has evolved over the
past four decades from relatively simple standards and test procedures appropriate for the
mechanically controlled engines of the 1970s and 1980s, to a multi-faceted program designed to
reduce emissions from modern computer-controlled engines. However, throughout the years, the
compliance paradigm has focused heavily on the pre-production certification process, during
which the manufacturers demonstrate that their engines will meet the applicable standards and
related requirements.
Until fairly recently, compliance was determined by testing engines in a laboratory on an
engine dynamometer. Prior to model year 2004, engines were evaluated over a single transient
test cycle commonly referred to as the Heavy-Duty Diesel Engine FTP (HDDE FTP) cycle.
Beginning with the 2004 standards, EPA added the engine dynamometer-based Supplemental
72
-------
Emission Test (SET)11, and Not-to-Exceed (NTE) emission limits that are evaluated on heavy-
duty engines while in-use on the road.
Heavy-duty diesel engines are currently subject to Not-To-Exceed (NTE) standards that are
not limited to specific test cycles, which means they can be evaluated during in-use operation.
In-use data are collected by manufacturers as described in section 2.2.1. The data is then
analyzed pursuant to 40 CFR 86.1370 and 40 CFR 86.1912 to generate a set of engine-specific
NTE events, which are intervals of at least 30-seconds when engine speeds and loads remain in
the control area. The express purpose of the NTE test procedure is to apply the emission
standard to engine operation conditions that could reasonably be expected to be seen by that
engine in normal vehicle operation and use, including a wide range of real ambient conditions.
The NTE zone defines the range of engine operation where the engine must comply. The NTE
zone is based on engine speed and load and includes some carve outs that include low load
operation (excludes load points less than 30% of '/max and Pmax, and less than 15% of the
European Stationary Cycle speed), as described in 40 CFR 86.1370. In addition, there are carve
outs for altitude (> 5500 ft), maximum ambient temperature (100 °F at sea level, 86 °F at 5500
ft), aftertreatment temperature for NOx aftertreatment and oxidizing catalysts (carves out
operation at temperatures < 250°C), and provides a cold temperature operating exclusion for
EGR equipped engines (calculation based on intake manifold temperature, engine coolant
temperature, and intake manifold pressure).JJ
Heavy-duty engines are required to comply with not-to-exceed (NTE) emission limits during
in-use operation. Engine manufacturers must acquire and submit data through the manufacturer
run in-use testing program. These in-use emission limits are 1.5 (1.25 for CO) times the
laboratory certification standard or family emission limit (FEL) for NOx, NMHC, PM and CO
and can be found in 40 CFR 86.007-11. A measurement allowance value is added on to the
standard to account for measurement inaccuracies that are associated with in-use measurement
over short time periods and can be found in 40 CFR 86.1912. The engine standards and
measurement allowances are in Table 2-6.
Table 2-6: Engine Standards and In-use Measurement Allowance
NOx
PM
CO
NMHC
(g/hp-hr)
(g/hp-hr)
(g/hp-hr)
(g/hp-hr)
Engine Standards
0.20
0.01
15.5
0.14
NTE Standards
0.30
0.015
19.4
0.21
Measurement Allowance
0.15
0.006
0.25
0.01
A valid NTE event is described as an event that is 30 seconds or more in duration under
engine, aftertreatment, and ambient conditions that are within the NTE zone (i.e., do not occur
during the aforementioned exclusion conditions), see 40 CFR 86.1370 and 40 CFR 86.1912. The
engine must meet a vehicle pass ratio of 90% of valid NTE events (i.e., 90% of the valid NTE
11 The SET was later modified to run as a single continuous test (similar to how a transient
cycle is run) and renamed the Supplemental Emission Test Ramped-Modal Cycle.
11 For more on our NTE provisions, see 40 CFR 86.1370.
73
-------
events must comply with the in-use standard (0.45 g/hp-hr for NOx) for the engine to be
considered compliant) as described in 40 CFR 86.1912.
We have concerns with whether the current NTE regulations ensure that compliance can be
achieved over the entire operating regime of the engine due to low temperature aftertreatment
exclusions, and the narrow engine operation that has to be met for at least 30 consecutive
seconds, as shown in Figure 2-3. Removal of these exclusions would require the engine
manufacturer to act to maintain aftertreatment temperature at low load modes of operation, and
this in turn would lead to better in-use performance with respect to emission compliance.122
Points excluded by reason:
• Intake Manifold Temperature • Aftertreatment Out Temperature
• Power •Torque
• RPM • Duration
• NTE Event
Figure 2-3: Sample of valid and invalid NTE events, separated by exclusion zones.
Data submitted under the current regulatory program has been acquired over more than
120,000 miles of vehicle testing. The data was generated by more than 540 unique vehicles and
submitted by 14 different engine manufacturers. The percentage of test results meeting or
exceeding the 0.9 pass ratio threshold since 2005 is presented in Table 2-7. A typical data set
representing one test for an engine will contain twenty to forty valid NTE events, but under some
routes no valid NTE events are measured.
Table 2-7: Percentage of Tests Meeting or Exceeding 0.90 Pass Ratio Threshold
Constituent
Percentage of Tests Meeting
or Exceeding the 0.9 Pass Ratio
NOx
96.0
CO
99.6
PM
99.4
The average pass ratio for each of the three constituents, for all data submitted since 2005, is
presented in Table 2-8.
74
-------
Table 2-8: Average Pass Ratios for Data Submitted Since 2005
Constituent
Overall Constituent Pass Ratio
NOx
0.980
CO
0.996
PM
0.997
NTE standards have been successful in broadening the types of operation for which
manufacturers design their emission controls to remain effective. However, our analysis of
existing in-use test data indicates that less than five percent of a typical time-based dataset are
valid NTE events that are subject to the in-use NTE standards; the remaining data are excluded.
Furthermore, we found that emissions are high during many of the excluded periods of operation,
such as when the aftertreatment temperature drops below the catalyst light-off temperature. For
example, 96 percent of tests from 2014, 2015, and 2016 in-use testing orders passed with NOx
emissions for valid NTE events well below the 0.3 g/hp-hr NTE standard. When we used the
same data to calculate NOx emissions over all operation measured, not limited to valid NTE
events, the NOx emissions were more than double (0.5 g/hp-hr).123 The results were higher when
we analyzed the data to only consider NOx emissions that occur during low load events. These
results suggest there may be great potential to improve in-use performance by considering more
of the engine operation when we evaluate in-use compliance. The average value of idle NOx for
the 2014, 2015 and 2016 Test Orders is approximately 16 g/hr.
2.2.2 Information evaluated for proposed updates
2.2.2.1 CE-CERT Program description
The CE-CERT study involves instrumenting about 100 trucks and collecting in-use second by
second activity data over the course of one month from on-board GPS units and ECU scan tools.
The instrumented vehicles are equipped with MY 2010 and later heavy-duty diesel engines
utilizing SCR. Over 170 parameters are recorded including GPS based location, speed, elevation,
ECU based vehicle speed, ECU based engine parameters (RPM, MAP, MAF, load), and
aftertreatment variables (temperature, NOx ppm). The vehicles cover 19 different groups based
on vocational use, GVWR, and geographic operation.
Key findings of this study: Average speed varies from 41 mph for out-of-state line haul trucks
to 9 mph for drayage trucks. In-state line haul trucks average 32 mph because they spend less
time on freeways and double the time idling, compared to out-state line haul trucks. Most
vocational vehicles spent about 33% of time operating at idle, irrespective of the time they spent
on the freeway. SCR temperature for line haul operation has a bi-modal distribution with peaks
at 100°C and 260-280°C. Drayage truck operation in Northern California exhibits a peak SCR
operating temperature of around 110°C. Overall, the vehicles in this study spent 42-91% of their
operating time with SCR temperatures below 250°C. Based on a generic SCR emission control
efficacy versus SCR temperature curve, the average engine-out NOx reduction could be 16% and
69% for agricultural trucks and refuse trucks, respectively.
In addition to giving a picture of heavy-duty vehicle activity, the study also provided data on
NOx sensors "on-time". It is widely known that NOx sensors are not turned on when the
humidity level of the exhaust is high enough to produced condensed water, but there is little data
75
-------
on how much time the sensors are off during heavy-duty vehicle operation. This information is
useful if NOx sensors are to be used as a compliance tool for an in-use standard. The sensors
were not operational, or the tailpipe NOx sensors were not reporting valid concentrations for
40% of the operating time based on the data collected. Figure 2-4 shows the time it takes until
the sensors start reporting data from a cold-start. The majority of the engine out sensors report
data within the first 10 minutes after a cold-start. The tailpipe sensors report data within the first
30 minutes, much later than the engine out sensors. This is understandable as the catalysts act as
a heat sink, absorbing heat from the exhaust during warm-up and preventing this thermal energy
from making it to the tailpipe sensor.
35%
3" 30%
qj 25%
u
c
2 20%
tn
_C
E 15%
CT3
V? 10%
2
o
<~> 5%
0%
¦ ¦
Il
II |
11
,
.
LI
lid
JJ
jjj-1-i..j .j
6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60
Time Until NOx Sensor On (mins)
I Aftertreatment Inlet Sensor
Aftertreatment Outlet Sensor
Figure 2-4: NOx sensor time to on after engine cold-start
2.2.2.2 Summary of HDIUT Data
To evaluate the efficacy of current technology NOx emissions controls, EPA analyzed the
data from engines selected for testing in calendar years 2010 through 2016. This dataset covers
44 engine families, model years 2010 to 2015, from 11 manufacturers. The dataset includes
about 8 million seconds of quality-assured second-by-second data collected during 68,000 miles
of driving. The operational conditions include a wide range of driving speeds, transient and
steady-state conditions, engine loads, and exhaust temperature conditions that have implications
for emissions control efficacy, particularly for NOx.124 For the HHD class, out of a total 159
vehicles, 109 were line-haul, 46 were delivery, and the remaining were marked as "Other" in the
metadata. Table 2-9 illustrates the distribution of family emission limit (FEL) values for the 291
vehicles tested. An FEL is a manufacturer-specified value that represents the maximum emission
rate from the engines in that group during certification testing.
76
-------
Table 2-9: Number of Diesel Vehicles with MY 2010+ Engines by NOx FEL Group from the heavy-duty in-
use testing Program
Regulatory
Class
NOx FE1
^ Group
Total
0.20
0.35
0.50
HHD
49
0
15
64
MHD
26
23
9
58
LHD
93
31
35
159
Urban Bus
0
10
0
10
Total
168
64
59
291
The following sections describe three analyses that use the HDIUT data to investigate the
relationship between aftertreatment temperatures and NOx emissions, and identify operations
where emissions are elevated.
2.2.2.3 HDIUT Data by MOVES OpMode
For the first analysis, the data was grouped by operating mode (OpMode) used by EPA's
MOtor Vehicle Emission Simulator (MOVES) emissions inventory model. MOVES OpModes
are defined in terms of power output using a scaled tractive power (STPt) parameter, shown in
Equation 2-1, and vehicle speed.
Equation 2-1
Avt + BVf + CVf + M ¦ vt(at + g • sinOt)
Where:
STPt = the scaled tractive power at time t [scaled kW or skW]
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 test vehicle [metric ton],
/scale = fixed mass factor (LHD = 5, MHD = 7, HHD = 10),
vt = instantaneous vehicle velocity at time t [m/s],
at = instantaneous vehicle acceleration [m/s2]
g = the acceleration due to gravity [9.8 m/s2]
sin 9t = the (fractional) road grade at time t
OpMode 0 is reserved for deceleration and braking events. OpMode 1 represents idle,
defined as vehicle speeds less than 1.0 mph. The remaining OpModes are defined by STPt ranges
and vehicle speed (vt) ranges of 1 to 25 mph, 25 to 50 mph, and greater than 50 mph. Figure 2-5
is a graphical representation of the MOVES OpModes, showing their relationship to STPt and vt.
See the MOVES Technical Report on heavy-duty emission rates for more information about
OpModes.125
77
-------
" 30
Q)
24
CUD
c
CD
u
CD
O
CUD
c
Speed: [1-25) mph
o
v
Q_
CUD
C
03
o
u
1
I
I
I
Speed: [25-50) mph
o
v
Q_
CUD
C
03
o
u
1
I
I
I
I
I
I
1
Speed > 50 mph
I
I
I
I
I
1
O
Q_
.1 18
"+¦»
u
2 12
T5
-S! 6
ro
u
to
0
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES Operating Mode
Figure 2-5: MOVES Operating Modes (OpModes) by scaled tractive power and vehicle speed
Figure 2-6 shows the NOx emission rates in g/s for 65 vehicles from five manufacturers (by
color) measured during in-use testing. The engines are categorized into eight HHD engine
families (E-l through E-8 by shape) with a NOx FEL certification level of 0.20 g/bhp-hr. Real-
world operation is known to produce large variability in emission rates, compared to certification
testing that is performed in a lab. The graph shows that there is significant inter-engine family
and intra-engine family variability in these in-use tests. For example, rates for E-l are
consistently lower than E-8. There is also variability within an engine family as represented by
errors bars for each point. The spread is larger for engine families with higher emission rates, for
example E-8. Similar trends are seen in MHD and LHD engine families.
78
-------
0.25
0.20 -
o.is -
O.iO
0.05
0.00
" I
I ¦
'Hi
? t ¦
i ! I I ?
¦ * T t T
iiji"
¦ i ¦ 1
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES OpMode
Figure 2-6: MOVES OpMode Emission Rates from HHD Engine Broken Down by Engine Family
Figure 2-7, Figure 2-8, Figure 2-9 show the OpMode-based NOx emission rates for vehicles
with NOx FELs of 0.20 g/bhp-hr or better split by aftertreatment temperature. The figures do not
include all tested vehicles, as some tests did not report aftertreatment temperature. Figure 2-7 is
based on data from 81 vehicles with HHD engines, Figure 2-8 is from 20 vehicles with MHD
engines, and Figure 2-9 is from 42 vehicles with LHD engines. For all engines and most
OpModes, the NOx emission rates when a vehicle is operating with an aftertreatment
temperature below 250°C are more than double compared to operation with an aftertreatment
temperature above 250°C. The figures also show that high NOx emissions occur across all
OpModes and are not limited to low-speed or idle operation.
79
-------
0.15
0.12
J* 0.09
jag
x
2 0.06
0.03
0.00
de id
[1-25) mph
[25-50) mph
>50 mph
T < 250 C
IT> 250 C
¦ ' ¦ I i i i
JlJ
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
OpMode
Figure 2-7: NOx Emission Rates from 81 Vehicles with 0.20 g/bhp-hr FEL HHD Engines by MOVES
OpMode and Aftertreatment Temperature
jag
x
O
0.05
0.04
0.03
0.02
0.01
0.00
de id
[1-25) mph
[25-50) mph
>50 mph
T < 250 C
IT> 250 C
ji
iiLii
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
OpMode
Figure 2-8: NOx Emission Rates from 20 Vehicles with 0.20 g/bhp-hr FEL MHD Engines by MOVES
OpMode and Aftertreatment Temperature
80
-------
ba
x
O
0.05
0.04
0.03
0.02
0.01
0.00
de id
[1-25) mph
[25-50) mph
>50 mph
¦ T < 250 C
¦ T > 250 C
1
1
, ]
T |
. . 1 1
i
>, '< i 1 i
ii
i i
ll
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
OpMode
Figure 2-9: NOx Emission Rates from 42 Vehicles with 0.20 g/bhp-hr FEL LHD Engines by MOVES
OpMode and Aftertreatment Temperature
Figure 2-10, Figure 2-11, and Figure 2-12 compare the same data in terms of operational time
by aftertreatment temperature. As expected, when the vehicles are operating at idle or low
speeds, more time is spent at the lower temperature bin. However, even at high speeds, a
nontrivial amount of time is spent at aftertreatment temperatures below 250°C.
81
-------
de id
[1-25) mph
[25-50) mph
>50 mph
0.35
0.30
rc
+->
,0
•- 0.25
M-
0
§ 0.20
'¦0
2 0.15
u_
0}
1 0.10
F
0.05
0.00
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
OpMode
Figure 2-10: Time Fraction from 81 Vehicles with 0.20 g/bhp-hr FEL HHD Engines by MOVES OpMode and
Aftertreatment Temperature
¦ T < 250 C
¦ T> 250 C
T
rn
i . -
1 «¦ #- - - -a -« -I
1
1
1
j,i .
de id
[1-25) mph
[25-50) mph
>50 mph
ro
+->
£
<4-
o
c
o
"¦p
u
ro
at
E
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
OpMode
Figure 2-11: Time Fraction from 20 Vehicles with 0.20 g/bhp-hr FEL MHD Engines by MOVES OpMode
and Aftertreatment Temperature
¦ T < 250 C
¦ T > 250 C
i.
1 zi l 1.. ii i. :x i. i. i1.1, ii I# ii ii ii J
82
-------
de id
[1-25) mph
[25-50) mph
>50 mph
0.45
_ 0-40
"ro
£ 0.35
h-
*5 0.30
.1 025
"¦p
ro 0.20
^ 0.15
ji o.io
0.05
0.00
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
OpMode
Figure 2-12: Time Fraction from 42 Vehicles with 0.20 g/bhp-hr FEL LHD Engines by MOVES OpMode and
Aftertreatment Temperature
wt < r
¦ T > 250 C
I
T
L,
i
1 Ia 1 - i.
1 Ia xi ii __
i I J
. 1 1 -
By combining the NOx emission rate data with the time data, we can estimate the total NOx
contribution by operation, as shown in Figure 2-13, Figure 2-14, and Figure 2-15. The imbedded
tables in each figure display the fraction of data in each temperature bin by operation time and
NOx mass. The emissions increase from low aftertreatment temperatures is not uniform across
all operating modes. For HHD engines (Figure 2-13), the aftertreatment temperatures spent
nearly as much time below 250°C as above it, but the contribution to total NOx is much higher
from the lower temperature operation due to the higher emission rates. The MHD and LHD
engines spent much more time in low aftertreatment temperatures conditions, and it is reflected
in a higher contribution to NOx. For all engines, the low- and mid-speed operating ranges
contribute the most NOx emissions. These figures highlight the need to consider both activity
and emission rate to effectively reduce NOx.
83
-------
de id
[1-25) mph
[25-50) mph
>50 mph
T < 250 C
IT> 250 C
Li
T < 250 degC
T > 250 degC
I i i l. I. i. i. L I.
L
Ratio of Total
Time NOx
0.45
0.65
0.55
0.35
iilii il1
il
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
OpMode
Figure 2-13: Total NOx Contribution from 81 vehicles with 0.20 g/bhp-hr FEL HHD Engines by MOVES
OpMode and Aftertreatment Temperature
de id
[1-25) mph
[25-50) mph
>50 mph
T < 250 C
T > 250 C
1 T-
T < 250 degC
T > 250 degC
Ratio of Total
Time NOx
0.72
0.28
0.83
0.17
i. i. i. I- i-
jli
_la il '
ii ii H 1
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
OpMode
Figure 2-14: Total NOx Contribution from 20 vehicles with 0.20 g/bhp-hr FEL MHD Engines by MOVES
OpMode and Aftertreatment Temperature
84
-------
¦ T> 250 C
Time NOx
T < 250 degC
0.59
0.70
T > 250 degC
0.41
0.30
Ji i i 8. i L i. i. L L L la ii i, ixLa it il J i
i
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
OpMode
Figure 2-15: Total NOx Contribution from 42 vehicles with 0.20 g/bhp-hr FEL LHD Engines by MOVES
OpMode and Aftertreatment Temperature
85
-------
2.2.2.4 HDIUT Data by Speed
Figure 2-16, Figure 2-17, and Figure 2-18 show the data binned by speed for F£HD, MHD,
and LHD vehicles, respectively, with NOx FEL at or below 0.20g/bhp-hr. Operation below 5
km/hr was excluded to fully remove any idle time or low creep operation, where emission rates
are drastically higher.kk The 0.20 g/bhp-hr FEL vehicles' average NOx emission rate in the
lowest speed bin is much higher compared to higher speeds. The overall average NOx emission
rate, shown in the right-most bin, is more than twice the certification value of 0.20 g/hp-hr. The
figure suggests the difference is even greater for vehicles that spend significant time at lower
speeds. The average time fraction (diamond marker) and the average NOx fraction (square
marker) reveal the outsized contribution to total NOx from the low and medium speed bins
where the time fraction is lower than the NOx fraction.
4.00
3.50
3.00
i 2.50
Q.
^ 2.00
Ofl
x 1.50
o
1.00
0.50
0.00
O
D
1
1.40
O
~ ~
0.20 g/bhp-hr 0
O
L I
1/
0.71
0-5 km/h
~0-3 mph
5-40 km/h
" 3-25 mph
40-80 km/h
" 25-50 mph
o
o
n
O Avg bsNOx
o Avg Time Fraction
~ Avg NOx Fraction
0.24
0.45
80-145 km/h
~ 50-90 mph
0-145 km/h
~ 0-90 mph
1.0
0.9
0.S
0.7 "5
£
0.5 •-
o
0.5
0.4
0.3
0,2
0.1
0.0
c
o
+3
u
2
Figure 2-16: Brake-specific NOx by Vehicle Speed Bins for 93 Vehicles with HHD Diesel Engines and an FEL
of 0.20 g/bhp-hr
Average NOx emission rate for HHD, MHD, and LHD are 6.38 g/bhp-hr, 7.65 g/bhp-hr, and 4.61 g/bhp-hr,
respectively.
86
-------
3,00
2.50
2.00
1.50
x
O i.oo
0.50
0.00
O
~
0.20 g/bhp-hr vr
Q
o
O
O
0.72
o
p
S
~
0.47
0-5 km/h
~0-3mph
5-40 km/h
' 3-25 mph
40-80 km/h
¦" 25-50 mph
o
6
Q
O Avg bsNOx
Q Avg Time Fraction
~ Avg NOx Fraction
0,24
80-145 km/h
™ 50-90 mph
0,47
0-145 km/h
~ 0-90 mph
1.0
0,9
0,8
0.7 "5
P
0.6 f-
0.5
0.4
0.3
0.2
0.1
0.0
o
E
o
+3
u
m
Figure 2-17: Brake-specific NOx by Vehicle Speed Bins for 26 Vehicles with MHD Diesel Engines and an FEL
of 0.20 g/bhp-hr
3.00
2.50
2.00
1.50
1.00
0.50
0.00
O
~
0.20 g/bhp-hr
O-
o
Q
o
8
S
9*
~
O
©*¦
0.S1
0.55
O Avg bsN Ox
~ Avg Time Fraction
~ Ave NOx Fraction
L-"
0,24
0.43
0-5 km/h
~0-3 mph
5-40 km/h
" 3-25 mph
40-80 km/h
25-50 mph
80-145 km/h
~ 50-90 mph
0-145 knVh
~ 0-90 mph
1.0
0.9
0.8
0.7 15
P
0.5
0.5
0.4
0,3
0.2
0.1
0.0
Q
c
o
ij
Figure 2-18: Brake-specific NOx by Vehicle Speed Bins for 49 Vehicles with LHD Diesel Engines and an FEL
of 0.20 g/bhp-hr
87
-------
2.2.2.5 HDIUT Data by Work-Based Window
Figure 2-19 shows a comparison of brake-specific NOx (g/bhp-hr) calculated using the
standard method ("bsNOx method") and CO2 based method ("NOx/CCh method"). This graph is
based on measurement data from 85 HHD diesel vehicles with NOx FEL < 0.20 g/bhp-hr. The
binning is by average power of the window over engine rated power. The windows are
continuous and non-exclusive - window n+l starts at time t=n+1. The amount of time in each
window is based on the engine power demand during the window. This analysis led to 2.90
million windows. The height of the columns represents the mean of all the windows in that
power bin and the error bars represent standard deviation of the mean. The 95% confidence
interval is not shown since windows are not independent. The CO2 based method is more robust
at very small loads, such as the 0-5% and 5-10% average power windows. In these cases, the
small amount of work done (bhp-hr) leads to higher brake-specific NOx values for the standard
method while also causing very large standard deviation. The CO2 based method addresses the
low-load in the denominator issue while also not penalizing vehicles that have lower CO2
emissions, by normalizing against CO2 over work (the second term in the equation). Another
takeaway is that emissions are much higher at lower loads, lowest at loads near the FTP and SET
cycles, and then creep up at higher loads. This suggests the engines are tuned to perform best at
loads/conditions similar to the certification cycle while less optimized for other real-world
operation.
Q.
J=
x
O
z
l/>
-Q
I bsNOx mean
NOx/C02 mean
bsNOx method =
NOx,
window
bhp — hr
NOx
1
CO 2
-method =
NOx,
window
window
CO 2
testday
HD FTP
|l iKll i/m i. ¦' jl I' r
C02window bhp - hr-testday
Ramped Mode Cycle
rfc .<£ .<& °P
> >$> <$> &
^window-avg/^rated (^)
Figure 2-19: Brake-specific NOx by Window Average Power Bins for 85 Vehicles with HHD Diesel Engines
and an FEL of 0.20 g/bhp-hr
88
-------
2.2.2.6 HDIUT Data by Simulated Cycle
Figure 2-20 and Figure 2-21 show analysis using simulated cycles. Figure 2-20shows the
drive cycles, which are converted to an OpMode time distribution, which is then combined with
the OpMode based emission rates for each vehicle in the HDIUT data set. The MOVES national
run OpMode distribution is also included for comparison in Figure 2-21. The Vehicle FTP and
UDDS drive cycles have similar speed traces and thus their OpMode distributions are also
similar. The transient drive cycle has a considerable amount of low speed transient operation,
which shows up as higher time spent in OpModes 11-14 and particularly OpMode 12. The
MOVES national run for combination-long haul trucks has most of its operation at highway
speeds, and thus most of the time is allocated to OpModes 33 and above.
Vehicle FTP Cycle
0 200 400 600 800
Time (s)
1000 1200
70
60
— 50
.£
| 40
"S 30
a;
a.
" 20
10
0
UDDS Cycle
5.55 miles
1,060 sees
4-
4-
400 600 800
Time (s)
Transient Cycle
100
200
300 400
Time (s)
500
600
700
Each driving cycle was converted to an OpMode based
time distribution based on the MOVES default road-
load coefficients for combination long-haul trucks.
Road grade was set to zero for the UDDS and Transient
Cycles.
Figure 2-20: Vehicle Speed Profile of HD Duty Cycles
89
-------
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
liii
I Sim. Vehicle FTP
¦ Sim. UDDS
E Sim. Transient
¦ MOVES Combi Long-Haul
ililJlHlUhft.il
L
iLU
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES OpMode
Figure 2-21: MOVES OpMode Time Fraction for each Simulated HD Combination Long-Haul Duty Cycle
Figure 2-22 and Figure 2-23 show the cycle average NOx emission rates calculated using the
per vehicle OpMode based emission rate and the drive cycle OpMode time distribution shown in
Figure 2-21. The average rate and spread for Vehicle FTP and UDDS are similar, since the
cycles are similar. The transient cycle produces the highest average rate because of low speed
operation that has lower aftertreatment temperature. The transient cycle also has the largest
spread, similar to the larger spread for the low speed analysis in Figure 2-16. The MOVES cycle
has the lowest average and spread because the operation is predominantly in the high-speed
zone, where emission rates are better controlled.
90
-------
3.50
3.00
_ 2.50
i-
.C
2.00
-Q
1.50
X
o
Z 1.00
0.50
0.00
0.75
0.80
Is
\s
0.20 g/bhp-hr
A Avg of 93 Vehicles
is
1.10
L-/
0.44
Simulated
Vehicle FTP
Simulated
UDDS
Simulated
Transient
MOVES National
Combi Long-Haul
Figure 2-22: Brake-specific NOx emissions by simulated cycle for HHD diesel engines with NOx FEL of 0.20
g/bhp-hr
Simulated
Vehicle FTP
Simulated
UDDS
Simulated
Transient
MOVES National
Combi Long-Haul
Figure 2-23: Distance-specific NOx emissions by simulated cycle for HHD diesel engines with NOx FEL of
0.20 g/bhp-hr
91
-------
2.2.3 Proposed Updates to CI Engine In-Use Test Program and Off-Cycle Standards
The focus of the current in-use NTE compliance testing is operation at relatively high load;
the data analysis procedure thus excludes a wide range of vehicle operations that occur in-use, in
particular operations at lower loads. Importantly, the excluded portion of the data makes up the
bulk of vehicle operation, specifically areas where NOx production is high.
To improve the coverage of the in-use testing program, we propose updates to the program to
include all vehicle operation within the regulated in-use standard. To keep the results
representative of actual engine/aftertreatment performance and minimize issues with temporally
misaligned data, we propose the analysis methodology be based on a series of moving average
windows (MAW).
2.2.3.1 Background on Euro VI MAW
The European Union Euro VI emission standards for heavy-duty engines requires testing for
in-service conformity starting with model year (MY) 2014 engines using portable emission
measurement systems (PEMS).126 127 The intent is to confirm whether heavy-duty engines
continue to comply with the emissions standards while in use, under normal operation, over time,
under real world conditions. Manufacturers must check for "in-service conformity" by operating
their engines over a mix of urban, rural, and freeway driving on prescribed routes using portable
emission measurement system (PEMS) equipment to measure emissions. Compliance is
determined using a work-based windows approach where emissions data are evaluated over
segments or "windows." A window consists of consecutive 1 Hz data points that are summed
until the engine performs an amount of work equivalent to the European transient engine test
cycle (World Harmonized Transient Cycle).
Engines are tested over a mix of urban, rural, and freeway driving. Testing starts at 18
months and a minimum of 25,000 km, and continues every two years thereafter out to seven
years and 700,000 km. There are no carve-outs for engine load or aftertreatment low
temperature operation. There are carve-outs for altitude (> 5577 ft), maximum ambient
temperature (100 °F at sea level), and minimum ambient temperature > 20 °F. There is no cold
start emission measurement requirement. Emission and work integration start when engine
coolant temp >70°C and is stable to within ± 2°C, or 20 minutes after engine start, whichever is
first. Vehicle payload must be 50 to 60 % of maximum.
Compliance is determined using a work-based windows approach where emissions are
evaluated over data segments or windows. A window consists of consecutive 1 Hz data points
that are summed until an amount of work equivalent to the World Harmonized Transient Cycle
(WHTC) is performed. For the window to be considered valid, the average power within a
window must be > 15% of engine maximum power for MY 2014-2018 and > 10% for MY
2019 and later. The vehicle must accumulate at least 5 complete windows over its shift-day and
50% of the windows must be valid. Compliance is demonstrated at 1.5 times the EURO VI
emission limit and there is no separate measurement allowance to account for in-use
measurement uncertainty (it is built into the conformity factor multiplier of 1.5).
EPA and others have compared the performance of US-certified engines and Euro VI-
certified engines and concluded that the European engines' NOx emissions are comparable to US
2010 standards-certified engines under city and highway operation, but lower in light-load
92
-------
conditions.128 This suggests that manufacturers responded to the Euro VI test procedures by
designing their emission controls to perform well over broader operation. EPA intends the
proposed rulemaking, if finalized, to expand our in-use procedures to capture nearly all real-
world operation. We are considering an approach similar to the European in-use program, with
key distinctions that improve upon the Euro VI approach, as discussed below.
2.2.3.2 Proposed Updates
For the updated in-use testing data analysis method, we propose using a MAW methodology
similar to that established in the Euro VI emission standards. However, most carve-outs would
be eliminated. Additionally, in order to adequately capture all vehicle operation, there would be
no minimum power requirement for valid windows. We are not currently proposing prescribed
routes for our in-use compliance test program. Our current program requires data to be collected
in real-world operation and we would consider it an unnecessary step backward to change that
aspect of the procedure. In what we believe to be an improvement to a work-based window, we
are proposing a moving average window (MAW) approach consisting of time-based windows.
Instead of basing window size on an amount of work, the proposed MAW includes a window
size of 300 seconds.LL The time-based windows are intended to equally weight each data point
collected.
We also recognize that it would be difficult to develop a single standard that would be
appropriate to cover the entire range of operation that heavy-duty engines experience. For
example, a numerical standard for CO2 specific NOx that would be technologically feasible
under worst case conditions such as idle, would be higher than the levels that are feasible when
the aftertreatment is functioning optimally.
Thus, we are proposing separate standards for distinct modes of operation. Our proposal is to
group the 300-second windows constructed from the second-by-second in-use data into one of
three bins using the nominal "normalized average CO2 rate" from the certification test cycles to
identify the boundaries. The normalized CO2 rate is defined as the average CO2 rate in the
window divided by the engine's maximum CO2 rate. Our proposal to define the engine's
maximum CO2 rate is to use the engine's rated maximum power multiplied by the engine's
family certification level (FCL) for the FTP certification cycle.
Windows with a normalized average CO2 rate greater than 20 percent (equivalent to the
average power of the current FTP) would be classified as medium/high-load operation and
binned together. Windows with a normalized average CO2 rate of 6 percent or less (6 percent is
equivalent to the average power of the low-load certification cycle) would be classified as idle
and binned together. The remaining windows, bounded by the idle and medium/high-load bins,
would be binned together in the low-load bin.
We are proposing that the emissions performance of the binned data in the medium/high and
low load bins would be determined using the sum of the total mass of emissions divided by the
sum of CO2 mass emissions. Emissions performance for the binned data in the idle bin would be
determined using the mass rate (total mass of emissions divided by total time) of the emissions.
This "sum-over-sum" approach would successfully account for all emissions; however, it
LL Our evaluation includes weighing our current understanding that shorter windows are more sensitive to
measurement error and longer windows make it difficult to distinguish between duty cycles.
93
-------
requires the measurement system (PEMS or a NOx sensor) to be accurate across the complete
range of emission concentrations. We have also considered the advantages or disadvantages of
other statistical approaches that evaluate a high percentile of the data (for example the 90th or
95th percentile) instead of the full data set.
As mentioned previously, we are proposing a separate MAW-based standard for each bin. In
our current NTE-based program, the NTE standards are 1.5 times the certification duty-cycle
standards. Similarly, for the MAW-based standards, we propose that the in-use program
standards correspond to the laboratory-based cycles, with each bin having its own standard.
2.2.3.3 Data Collection and Exclusion
For the HDUIT, emissions data are to be collected from key-on (I = 0) until the end of the
shift day when the engine is turned off. Data are to be collected once every second (i.e., 1-Hz
data).
From the data collected at 1 Hz, some data points are to be excluded from the remaining
process. Data to be excluded are (1) data collected during the PEMS zero drift check, (2) data
collected where the engine is off, (3) data collected during infrequent regeneration events, and
(4) data collected when any approved AECD for emergency vehicle applications is active.
2.2.3.4 Defining Windows
With the extended idle times frequently present in HDIUT samples, a work-based window
approach would include longer period of time for these windows, and the methodology would be
very sensitive to small inaccuracies in power measurement. To ensure the final set of windows
more accurately reflects the operation of the vehicle, we propose adopting a time-based window
approach, where each window contains an equal amount of time rather than an equal amount of
work, as in the Euro VI work-based window approach.
For this proposed methodology, a window will consist of the summation of 300 consecutive
1-Hz data points (i.e., a 300-second window). The windows are continuous and non-exclusive,
with subsequent windows beginning one second after previous windows (i.e., at the next data
point). The first window will begin at initial key-on (t = 0), and the final sequential window will
begin 300 seconds before the last data point taken. To limit the impact of instances where data
exclusions would reduce the weighting of an individual data point, exclusions of < 600 seconds
are removed and the remaining data concatenated. For exclusions > 600 seconds, the final pre-
exclusion window begins 300 seconds before the exclusion, and the next subsequent window
begins immediately after the exclusion.
Except for the data points as the beginning and end of the test and those around long data
exclusions, this methodology equally weights emissions at each data point during the in-use
testing. We believe this is appropriate, as the under-weighted data points consist of a small
percentage of the HDIUT data, which contain a minimum of 10,800 1-Hz data points.
2.2.3.5 Emission bins
The agency recognizes that including operation currently excluded from the standard,
including low-load operation and low aftertreatment temperature, will result in a higher range of
variability in both the vehicle operation represented, and in the data captured during testing.
94
-------
Thus, we are proposing to differentiate the data collected by vehicle operation, and
independently set standards for each operational characteristic.
To differentiate among various types of operation, we are proposing to divide the windows
among three bins that are characterized by the normalized average CO2 rate: an idle bin, a low-
load bin, and a medium/high load bin. The normalized CO2 rate of each window is defined as the
total window CO2 mass divided by the 300-second window length and then divided by the
maximum CO2 rate of the engine. The engine's maximum CO2 rate is defined as the engine's
rated maximum power multiplied by its family certification level (FCL) for the FTP certification
cycle.
The three bins we are proposing are defined as follows:
• Idle bin: window normalized average CO2 rate < 6 %
• Low-load bin: window normalized average CO2 rate > 6 % and < 20 %
• Medium/high-load bin: window normalized average CO2 rate > 20 %
2.2.3.6 Bin Size and Test Validity
We are proposing that, for a test to be considered valid, all bins must contain a minimum of
2400 windows. To ensure there are enough windows in the idle bin you may idle the engine any
time during the shift day. If any bin contains fewer than 2400 windows, the vehicle must be
tested over an additional shift day. The resulting windows from the second or subsequent shift
day are added to the appropriate bin, so that all windows from all shift days are included.
Using data from 168 previous HDUIT tests of one shift day each, the 1-Hz data from these
tests was collected into windows and binned according to the proposed process as seen in Figure
2-24. Of these single shift day tests, a total of 73% contained 2400 windows in each bin. An
additional 16% contained over 1200 windows in each bin (and thus would be likely to contain
over 2400 windows in each bin if data were recorded over two shift days). Note some HDUIT
tests have multiple bins containing fewer than 2400 windows. From this data, we estimate that,
under the proposed process, approximately 73% of tests would be valid with a single shift day's
worth of data, and most of the remainder would have over 2400 windows with a two-shift-day
test.
If, after consideration of comments on this rule and other available data, EPA's technical
assessment is that a minimum amount of data is required to adequately characterize emission
performance over the wide range of vehicle operations that occur in-use, we will consider other
options that require a minimum number of windows in each bin.
95
-------
c
1q
-£Z
o
re
a>
_c
5
o
TJ
C
5
0% 5% 10% 15% 20% 25% 30%
Percentage of total number of windows
Figure 2-24: Number of windows in each bin from 168 IIDUIT shift days, sorted in order (the first 30% are
shown). The black lines indicate 2400 windows and 1200 windows. Note some HDUIT tests have multiple bins
containing fewer than 2400 windows.
2.2.3.7 Defining Bin Emissions
Historically, engine standards have been work-specific. Using this approach, the standards
can apply to a wide range of engine sizes. Where this approach is challenged is when the test
interval that is being is evaluated has very little to no work produced, such that the emissions are
then divided by zero or near zero. Thus, we propose establishing an "equivalent work" standard
based on CO2 emitted from the engine. This methodology also does not rely on estimating or
recording the second-by-second power output of the engine to determine work done. For this
standard, the engine's FTP FCL CO2 emission value (eco2FTLFCL) is used to convert the CO2
specific emissions to equivalent work specific emissions.
For the FEDUIT data processing, emission values are calculated for each bin, using a sum-
over-sum value. For the low-power and medium/high-power bins, CO2 specific emissions are
determined for each bin, which are converted to work specific emissions using the engine's FTP
FCL CO2 emission value (eco2FTLFCL) as follows:
Equation 2-2
/ g \ Total bin emission
e = ~T^TbhU0TX
In the idle bin, under nominal idle conditions the engine produces no work, giving an
incentive to minimize CO2 production in these operation modes (for example, with cylinder
96
6000
-------
deactivation). So as not to artificially increase the stringency of the idle bin standard for engines
with low idle CO2, we propose setting standards as emission rates rather than work specific
values. For the idle bin, the emission rate would also be calculated using a sum-over-sum value
as follows:
We are also considering defining bin emissions for the idle bin on a work specific basis, as in
Equation 2-2.
2.2.3.8 Bin Emissions and Standards
We propose that the standards apply to the sum-over-sum emissions value for the entire bin,
as shown in Equation 2-2 and Equation 2-3. This methodology accounts for all emissions
included in a particular bin equally and reduces the influence of potential errors in data
collection. See Figure 2-25.
If, after consideration of comments to this rule or review of any additional data, EPA's
technical assessment is that emissions cannot be quantified accurately over the entire emission
range of the engine, we will consider options that would minimize the importance of quantifying
low emissions rates. One possibility for doing this is calculating sum-over-sum emission values
for each window independently. Within each bin, the windows are ordered by the magnitude of
emissions in g/hp-hr, from smallest to largest. From these, the emissions values for the windows
at specific percentiles (for example, the 60th and 95th percentiles) are selected and compared to a
standard. If this approach was used the standards for these percentiles in each of the three bins
would replace the single bin standards.
Equation 2-3
g \ Total bin emission
Total bin time
97
-------
6
5
^ 4
CL
J£
CU5
X _
0 3
2
c
1 2
i
o
0 10 20 30 40 50 60 70 80 90 100
Percentile of window values in bin
Figure 2-25: Range of NOx rates in windows in each bin, using data from 93 HDIUT results in the HHD, 0.2
FEL category. The heavy line is the median value from all tests, with error bars represent the 25th and 75th
percentile of the data.
2.3 Spark-Ignition Test Procedures and Standards
Spark-ignition test procedures are a crucial aspect of the heavy-duty criteria pollutant
program. This rulemaking is proposing to establish several new test procedures for spark
ignition engine compliance. This chapter describes the existing test procedures as well as the
development process for the test procedures being proposed. This includes the determination of
emissions from both engines and hybrid powertrains as well as the development of new duty
cycles.
2.3.1 Current SI Test procedures
Heavy-duty spark-ignition engines currently are certified for criteria pollutants using the
Heavy-Duty Otto-Cycle Engine Federal Test Procedure (HDOE FTP). 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
Flydrocarbons, Carbon Monoxide, and Particulate Matter. The HDOE FTP duty cycle is defined
in 40 CFR part 86 Appendix I. All emission measurements and calculations are defined in 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.
Near idle
Low-mid power
High power
—
C II
-r
¦—¦—1—•—1
rf-1
1 !
1 '
—i
98
-------
The measurement method for CO is described in 40 CFR 1065.250. For measurement of
NMHC refer to 40 CFR 1065.260. For measurement of NOx refer to 40 CFR 1065.270. For
measurement of PM, refer to 40 CFR 1065.140, 1065.170, and 1065.290. Table 1 of 40 CFR
1065.205 provides performance specifications that we recommend analyzers meet. Note that 40
CFR 1065.307 provides linearity verifications that the system must meet. For the calculation
method brake specific mass emissions for CO, NMHC, NOx and PM refer to 40 CFR 1065.650.
2.3.1.1 HDOE FTP
The current HDOE 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. This procedure is well-defined, mirrors in-use operating parameters, and
continues to be appropriate also for the continued assessment of criteria pollutant emissions from
heavy duty engines.
A complete HDOE FTP involves three test sequences. First, a 20-minute test is run over the
duty-cycle with the engine at the same ambient temperature as the test cell (between 68°F and
86°F). The engine undergoes a 10-minute hot-soak following the cold-start. A 20-minute hot
start test is run over the same duty-cycle following the hot-soak. The HDOE FTP emission level
for the engine is determined by weighting the cold start emissions by 1/7 (14 percent) and the
hot-start emission results by 6/7 (86 percent).
2.3.1.2 Test Procedure Engine Mapping Improvements
The heavy-duty FTP test cycle is composed of second by second speed and torque targets that
are based on the engine's design operating speeds and the torque levels produced over the full
range of allowed speeds. In order to determine the torque level at any engine speed, a mapping
of the engine is performed prior to the actual FTP testing. The mapping is a sweep across the
mechanically- or electronically-allowed operating speeds to determine the highest possible
torque level the engine can produce at any specific speed. From this "maximum torque" sweep,
the FTP targets are determined for the subsequent transient FTP test and for any additional test
(e.g. SET).
As noted above, the measured torque values are intended to represent the maximum torque the
engine can achieve under fully warmed-up operation with the appropriate design fuel grade (i.e.
regular grade octane fuel) across the allowed range of engine speeds from idle to the typically
electronically limited highest RPM. The intent is to reflect a torque value that would be
maintained if the engine were to stabilize at a specific speed over a longer period of time such as
what might be observed in a "real world" condition when an engine is held by a transmission in a
specific RPM by the selected gear.
Electronic control of all aspects of engine hardware and operation has resulted in some
challenges to performing the mapping of the engine. Variable torque levels have been observed
in engine testing related to such things as electronic control response to fuel octane, anticipated
exhaust thermal conditions, transmission torque limiting models, and other electronic features
incorporated into the engine management strategies. This torque variability has been particularly
evident after the change to the mapping test procedures in 40 CFR part 1065.510(b)(5)(i), which
requires that an engine be mapped by performing a transient sweep from idle to maximum rated
99
-------
speed. Prior to this change in the test procedure, the mapping procedure required the engine to
stabilize at discrete engine speed breakpoints before recording the engine torque value for that
engine speed that would be used for the FTP and other testing.
There are two potential improvements to reduce the torque measurement variability. One
option could be to perform the sweep in both directions (i.e. idle to rated maximum speed and
back to idle) and determine the torque value at any speed to be the highest measured torque level.
Typically, we would expect any torque limiting AECDs (e.g. power enrichment delay) to be
active during the sweep up, allowing the maximum torque to be determined during the sweep
down after any torque limiting AECD has concluded. A second option could require
manufacturers to turn off the torque limiting AECDs during the torque mapping sequence and
allow them to perform the procedure as written today. For example, an AECD that results in
delay of component thermal protection due to a temperature model would not be allowed for the
power mapping sequence since any stabilized operation at a temperature limited speed point
would typically require thermal protection. During the transient FTP test however, the AECD
could be active to simulate real world operation and thermal management strategies implemented
to improve emission performance and conserve fuel.
2.3.2 Proposed updates to SI Test procedures and Standards
We are proposing an overarching update to the location of our highway heavy-duty engine
regulations, moving from the current 40 CFR part 86 to part 1036. As part of this process, we
propose to clarify our nomenclature and no longer refer to "otto-cycle" engines; instead, these
engines are more accurately labeled spark-ignition engines throughout part 1036. This section
provides additional details related to the test procedures we are proposing in Section III.D. of the
preamble to this rulemaking. Refer to the preamble for specific requests for comment related to
these procedures.
2.3.2.1 HD SI FTP
As part of our proposed migration to part 1036, the FTP duty cycle maintains the weighting
factors or the duty cycle speed values from the current HDOE FTP duty cycle that applies to
criteria pollutant regulation in 40 CFR part 86 Appendix I (f)(1). We are proposing a change to
the negative torque values, as noted below. The HD Technical Amendments that were published
in June, 29 2021 finalized the migration of some heavy-duty highway engine standard setting
part test procedures from 40 CFR part 86 to part 1036. This included the migration of the HDOE
FTP drive schedule to Appendix II (b) of part 1036 in order to add vehicle speed and road grade
to the duty-cycle to facilitate powertrain testing of hybrid powertrains for compliance with the
HD Phase 2 GHG standards.
We are proposing in this rule to align the changes made for certification to GHG standards
with criteria pollutant testing. Specifically, the removal of and footnoting of the negative
normalized vehicle torque values over the HDOE FTP duty-cycle. The footnote denotes that
these torque points are controlled using closed throttle motoring, which would then match how
negative torque values have been controlled in the HDDE FTP. This change reflects the way
that engine manufacturers were already controlling to negative torque from spark-ignition
engines and harmonizes the methodology with the HDDE FTP, with no effect on stringency.
100
-------
The spark-ignition engine denormalization equation in 40 CFR 86.1333(a)(l)(ii) contains a
divide by 100 which equates it to the denormalization equation in 40 CFR 1065.610(c)(1)
(equation 1065.610-3), thus the elimination of the 40 CFR part 86 equation from the standard
setting part will have no consequence.
2.3.2.2 Engine mapping
We are proposing to change the procedure for SI engine torque mapping in 40 CFR part
1065.510. In order to determine the torque level at any engine speed, a mapping of the engine is
performed prior to the actual FTP testing. The mapping is a sweep across the mechanically- or
electronically-allowed operating speeds to determine the highest possible torque level the engine
can produce at any specific speed. From this "maximum torque" sweep, the FTP targets are
determined for the subsequent transient FTP test and for any additional test (e.g. SET).
The measured torque values are intended to represent the maximum torque the engine can
achieve under fully warmed-up operation with the appropriate design fuel grade (i.e. regular
grade octane fuel) across the allowed range of engine speeds from idle to the typically
electronically limited highest RPM. The intent is to reflect a torque value that would be
maintained if the engine were to stabilize at a specific speed over a longer period of time such as
what might be observed in a "real world" condition when an engine is held by a transmission to a
specific RPM by the selected gear.
Variable torque levels have been observed in engine testing related to such things as
electronic control response to fuel octane, anticipated exhaust thermal conditions, transmission
torque limiting models, and other electronic features incorporated into the engine management
strategies. This torque variability has been particularly evident with the change to the mapping
test procedures in 40 CFR part 1065.510(b)(5)(i), which allowed an engine to be mapped by
performing a transient sweep from idle to maximum rated speed. Prior to this change in the test
procedure, the mapping procedure required the engine to stabilize at discrete engine speed
breakpoints before recording the engine torque value for that engine speed that would be used for
the FTP and other testing.
We are proposing to require the engine to achieve a stabilized torque reading at the different
speeds prior to recording the final torque values, which would be accomplished by disabling any
controls that limit torque during the engine mapping test.
2.3.2.3 Supplemental Emissions Test for HD SI
As noted in Section 2.1.1.2, the compression-ignition engines currently comply with SET-
based standards that represent high-speed and high-load operation. The SET duty cycle is a
ramped modal cycle in which the engine is tested on an engine dynamometer over a sequence of
steady-state modes. As we show in Section 4.2.3.2, there are opportunities to reduce emissions in
high-load operating conditions where engines often experience enrichment for either catalyst
protection or a power boost. We are proposing to include SET-based standards for HD SI
engines to ensure that emission controls are properly functioning in the high load conditions
covered by that duty cycle. We are proposing the same Cl-based SET procedure, summarized in
Section 2.1.2.2, for HD SI engines, including the existing weighting factors shown in Table 2-2.
101
-------
2.3.2.4 Onboard Refueling Vapor Recovery
The current ORVR test procedure, which can be found in 40 CFR 1066, subpart J, for
measuring emissions from chassis-certified vehicles during a refueling event, requires that the
testing occur in a sealed housing evaporative determination (SHED) enclosure containing the
complete vehicle. This procedure applies to all light-duty and heavy-duty complete vehicles
subject to the ORVR standards, and manufacturers designed and built the SHEDs at their test
facilities for these vehicles.
During a recent test program, EPA discovered that very few SHEDs are available that could
fit vehicles in the over-14,000 lb GVWR class because of a combination of the length and height
of these work vehicles. Additionally, the limited large volume SHEDs that were available at
third-party laboratories proved to have challenges measuring the refueling emissions because of
the very large volume inside the enclosures.129
Large background volumes of ambient air create a challenge for evaporative emissions testing
because a measurement is only considered representative if emissions are able to reach a
homogeneous distribution throughout the cell prior to initiating a measurement by the emission
analyzers. In EPA's test program, we found that the two heavy-duty test vehicles required a
substantially longer mixing time than the current test procedure developed for light-duty
vehicles.129'130 Another challenge in regards to adopting the existing procedures for larger
vehicles is that the calculations must account for the volume displaced in the SHED by the test
vehicle, which can be highly variable for the range of commercial vehicle designs. If a full
SHED test is desired for ORVR certification, we must consider an appropriate duration of the
test to achieve a representative emissions distribution in the larger SHED, as well as an
appropriate means of calculating the displaced volume of these diverse vehicle designs in order
to get an accurate measurement of the refueling emissions.129'130
Figure 2-26 and Figure 2-27 show examples of the estimated extrapolated mixing time for the
two trucks that were tested in the summer of 2018. It is not known how much additional mixing
time is necessary without additional testing, but it is at least three minutes based on the
extrapolated test results from this program. It might make more sense to require stabilization of
the analyzer hydrocarbon readings rather than a specific time length.
102
-------
Ford - Estimated Additional Time Required for Mixing
160
Actual End of Teal
Rtffuflline ends [114 grams)
Th*« addMornal
mi nut 01 -
Tvnj rfdililkiiM! I.
minutes *
J-Wtg/gSj, '*
One additional ^ \
minute .X
04:00 05:00
Time mlnii«es:s«torids
Figure 2-26: Estimated Projected Ford E-450 ORVR Results based on Extrapolation
1U0
Isuzu - Estimated Additional Time Required for Mixing
Refueling ends
\
\
140
\
MDVbb rnndrls
3,Mg/8al.,
One additional
120
£ 100
2
fin
at Jb airtra
araiml
minute
\
Actual End of Test
2.B3e/gal
n > _ _ _ . '<
[72«;ram5j
X
m
**
GHtflD
di :rxi D7:rxi a=t
-------
One potential solution to issues related to testing these commercial vehicles in limited
availability and large volume SHED equipment could be to allow demonstration of compliance
by testing the complete ORVR system with all of the components (fuel tank, filler pipe, canister,
control valves, etc.) independent of the actual vehicle using an existing SHED designed for
smaller vehicles, such as light-duty applications. These existing SHED enclosures are widely
available to test and certify all ORVR-required vehicles. This approach of only testing the
components associated with refueling emissions would remove the challenge of finding a SHED
with sufficient dimensions to contain the vastly larger (i.e. longer and taller) commercial vehicles
that are part of this proposal. Testing the refueling related components independent of the
vehicle also eliminates the challenge of minimizing other hydrocarbon sources not associated
with fuel or the fuel system (i.e. tires, plastics, paints, etc).
Another option would be to allow independent ORVR hardware described above to be tested
in a small enclosure designed for only component specific testing (i.e. mini-SHED) similar to the
methodology for the "rig" test, allowed for other evaporative testing by California and accepted
by EPA. Similar to the ORVR system-based concept described previously, the mini-SHED
approach would provide a simpler test methodology that would capture just the refueling-related
emissions. Furthermore, this smaller scale, component-based test would eliminate much of the
variability encountered when attempting to test a full vehicle or hardware in a large SHED.
For both system- and component-based ORVR testing, the canister would require a specific
conditioning cycle to get it into a representative state of a real-world loading entering a refueling
event. This could be performed on or correlated to an actual vehicle driven over an existing EPA
test cycle or a "real world" drive cycle. The current preparatory cycle used by today's ORVR-
required vehicles is designed for light-duty vehicle driving patterns and vehicles with typically
much smaller fuel tanks and canisters. The current conditioning procedure is designed to
challenge the purge system in scenarios such as heavy traffic, slow speeds and start-stop events.
Heavy-duty vehicles, with larger fuel tanks and canisters, may drive more miles and have greater
power demands that may help purge the larger canisters more easily than allowed for in the
current light duty vehicle test. Commercial vehicles may still drive under heavy traffic scenarios
but the expectation is that they will drive more miles daily and operate under higher loads on a
regular basis which help to purge larger amount of vapors from the system.
A final potential option would be to leverage similar vehicles already required to test as a
surrogate for certification compliance demonstration. The current 2-day and 3-day evaporative
standards program for this class of heavy-duty vehicles acknowledged that vehicles currently
tested in the lighter classes (i.e. under 14,000 lb GVWR) may contain the exact same hardware-
and purge-related calibrations providing a high degree of confidence that the heavier versions
will also comply with the 2-day and 3-day evaporative requirements. Because of this similarity
between the classes and a high degree in consistent performance, the agency allows the data and
testing from the lighter vehicle to be accepted for the certification compliance demonstration of
the larger class of vehicles. The ORVR systems could similarly be of the same design and purge
calibrations between the vehicles in this proposal and vehicles already required to comply with
ORVR requirements. In these cases, it would seem appropriate to allow the use of the results of
the tested lighter vehicle class along with "good engineering judgement" statements as a means
to demonstrate compliance for the larger vehicles.
104
-------
2.3.3 Idle Test Procedures Considered
As noted in the preamble, we are proposing to ensure that the main component of the
emission control system, the catalyst, remains effective during prolonged idle situations. The
current heavy-duty FTP test does not include extended idle conditions indicative of real-world
behaviors when work vehicles are started and allowed to idle while warming-up the engine or
when the vehicle operator requires the interior temperature to stabilize with either heat or air
conditioning. This prolonged idle can also occur when the vehicle is brought to a stop for
unloading or a similar situation.
We considered the addition of a new test procedure to ensure idle emission controls are
maintained. We considered the FTP or SET run as a pre-conditioning cycle to stabilize the
engine and emission control system, followed by 10-30 minutes of idle. Previous idle
provisions, established for in-use inspection and maintenance programs, required a 0.50 percent
by volume CO limit over a 30-minute idle period, but did not set a limit for NMHC or NOx.131
We also considered options that take advantage the existing SET duty cycle to avoid
introducing a new test procedure. One option could be to reevaluate the weighting factors of the
SET to place a greater emphasis on the idle modes, but this option has two drawbacks. First,
providing more weight to the idle mode would have the counter effect of deemphasizing the
high-load modes that we believe are critical to encouraging reduced fuel enrichment. Second, the
existing SET procedure collects the emissions from all modes in a single bag for analysis, which
means idle performance would be masked by the composite result and the additional weighting
at idle would require us to reevaluate the SET standard feasibility. Another option could be to
require two bags for SET such that the idle modes would be collected in the second bag. This
option would isolate the idle results but would require two separate standards. While this option
would reduce the need for a new test procedure, manufacturers would still likely need to make
adjustments to their test cells to accommodate the second bag.
2.4 Compliance Assurance
2.4.1 Improved Engine Control Module Security as a Deterrent to Tampering
Today's highway trucks have many interconnected, computer-based systems which control a
variety of vehicle features and functions necessary for safe and efficient operation. While these
electronically controlled systems greatly expand the capability of vehicles, they also create a
potential pathway for individuals to tamper with or alter the software and or calibrations in ways
which significantly increase emissions. As more and more functions of vehicles are automated
and rely on computer controls, securing these systems from malicious attacks or tampering is a
topic of concern to vehicle manufacturers and owners alike. These attacks or acts of tampering
can be direct, where an individual has access to the engine control module ("ECM") on a single
vehicle, or indirect, where reprogamming or modification of the ECM is initiated via an over-
the-air or wireless connection (a "cyber" attack) on one or many vehicles. While there are safety,
financial, operational, and privacy concerns when these systems are attacked or compromised,
this discussion will focus on the 'operational' aspect, where tampering with the engine and
emissions control computers can have a negative impact on the emissions and/or diagnostic
performance of a vehicle.132
105
-------
A common method of tampering is the practice of modifying a vehicle's ECM to allow the use
of aftermarket defeat devices, engine tuning kits, or full aftertreatment delete kits, where the
removal, bypass, or disabling of DPF, SCR, DOC, and EGR components can have the effect of
bringing the vehicle emissions to a pre-standards condition. In addition, tampering with certified
calibrations in the ECM with an engine tuning kit may exacerbate the emissions increase, as
changing critical combustion control parameters to increase fuel economy or engine performance
can result in increased NOx and PM emissions as well. For example, the fuel maps in the ECM
may be modified to advance the injection timing, resulting in higher cylinder pressures, which
promote the formation of NOx. Tampering may also include the installation of devices which
mimic a valid signal from emissions control sensors or disabling OBD monitors which, in turn,
can override the inducement or engine derate strategies that are necessary to ensure that high-
quality diesel exhaust fluid ("DEF") is added to the vehicle.
To better understand how manufacturers and industry are securing ECMs from tampering,
ensuring that the engine and vehicle will remain emissions compliant throughout its useful life
and beyond, we met with manufacturers to understand the measures they are employing to secure
these devices. We also met with telematics providers and industry technical organizations to
learn about the efforts currently underway to establish ECM and data security standards and
protocols. Given the increased emissions that a tampered truck will produce throughout its
operational life, it was important for EPA to study and understand the methods and procedures
that manufacturers are currently employing to secure the ECM from tampering efforts, as well as
those that are under development by technical associations such as the Society of Automotive
Engineers (SAE) and the Automotive Information Sharing and Analysis Center (Auto-ISAC), as
well as the National Highway Transportation Safety Administration (NHTSA). As we learned in
these meetings and discussions, there are many approaches currently being used to secure vehicle
computer systems from tampering and malicious attacks, but these approaches are ever changing,
as the sophistication of those seeking to compromise these systems is also increasing, prompting
additional countermeasures by manufacturers.
EPA was considering whether a collection of industry best-practices regarding ECM security
could be identified, and whether it would be beneficial and feasible to include those practices as
a prescribed set of regulatory requirements in a future rulemaking proposal. However, as we
learned throughout these discussions, there are many ways to accomplish ECM security, and
they will by nature always be in a state of change, adapting to the latest threats. As a result, EPA
is not proposing a prescribed set of security requirements, and is instead proposing that
manufacturers document the security measures they are employing on their products at time of
application for certification. We considered more prescriptive approaches for addressing ECM
security, but at this time, we are choosing to propose a requirement that will better allow
manufacturers, industry groups, and safety regulators to take the lead on developing security
requirements for these systems in a timely manner, as issues emerge. In the course of our study
of security practices and discussion with industry and technical experts, we determined that the
topic of computer security is complex, varied in methodology, and rapidly changing, and even if
EPA were to promulgate specific requirements for ECM security, those requirements would have
to be updated annually at a minimum, as the pace of change in security technology is rapid.
Manufacturers need to have the ability to quickly implement securing changes, without the
burden of following prescriptive regulatory requirements.
106
-------
However, some assurance is needed when a product is sold that appropriate measures have
been taken to ensure that the ECM is tamper-proof and that it is not vulnerable to attack using
readily available tools. Section 203(a)(3)(A) and (B) of the Clean Air Act not only prohibit
tampering, but also prohibit the manufacture of a product where a person knows or should know
that "the principal effect of the part or component is to bypass, defeat, or render inoperative any
device or element of design installed on or in a motor vehicle in compliance with regulations."
The process of re-programming or re-flashing the ECM can occur by several methods but
most, if not all, of the methods involve an access through the data link connector (DLC or
commonly known as the OBD port). In some cases, the owner of the vehicle purchases a device
that is capable of re-programming the ECM, and the device is plugged in to the DLC, where the
owner can select one of the preloaded calibrations, or "tunes," that are preloaded on the device.
Another method of re-programming involves remote flashing of the ECM. In this case, the
vehicle owner will purchase a wireless CAN adapter that is connected to the DLC, which allows
the manufacturer of this device to remotely access the vehicle's ECM and initiate the re-flash via
an internet or cellular data connection.
To ensure that there are measures in place to prevent ECM tampering, we are proposing that
manufacturers include a document at time of certification which outlines and describes the
process and/or industry technical standards that were used to prevent unauthorized access to the
ECM on the vehicle. This document shall describe the measures that a manufacturer has used to:
prevent unauthorized access to the ECM; ensure that calibration values, software, or diagnostic
features cannot be overwritten or otherwise disabled; and respond to repeated, unauthorized
attempts to reprogram the ECM, if they become aware of such attempts.
107
-------
Chapter 3 Feasibility Analysis for the Proposed Standards
3.1 Compression-Ignition Technology Feasibility
3.1.1 Diesel Technology Demonstration Programs
3.1.1.1 CARB Heavy-duty Low NOx Stage 3 Research Program
In 2016, The California Air Resources Board (CARB) funded the Heavy-duty Low NOx
Research Program at Southwest Research Institute (SwRI) in San Antonio, TX to explore the
feasibility of diesel HDEs achieving 0.02 g/bhp-hrNOx composite emissions over the FTP
transient test cycle. Stage 3 of this research program investigated the use of dual-SCR systems,
cylinder deactivation (CDA), heated urea dosing system, intercooler bypass, turbine bypass, and
EGR cooler bypass via engine dynamometer testing of a developmental engine based on the
MY2017 Cummins XI5 diesel Heavy HDE. Major specifications for the engine are shown in
Table 3-1. Like many other MY2010 and later diesel HDEs, the X15 is equipped with a variable
geometry turbocharger (VGT), high pressure common rail fuel injection and cooled EGR. The
X15's original equipment exhaust aftertreatment system (EAS) consists of a DOC, DPF, SCR,
and ASC in series. The XI5 engine was modified by SwRI and Eaton to incorporate individual
cylinder deactivation. SwRI also developed an engine calibration to aid catalyst warmup using a
combination of later combustion phasing and increased idle speed. Further details of the
specific, fixed CDA modes evaluated and other details regarding CDA development and engine
instrumentation can be found in two papers by SwRI and Eaton.13X134 Details regarding the EAS,
control systems, and calibration are summarized in three additional papers by SwRI.135'136'137
The complete summary of the work completed as part of Stage 3 is included in the final report
from SwRI.138
Table 3-1: Major engine specifications for the MY2017 Cummins X15 engine used for the CARB Low NOx
Stage 3 Research Program
Engine Displacement
14.95 L
Bore X Stroke
137 X 169 mm
Rated Power (a)y Speed
373 kW (a)y 1800 rpm
Rated Torque @ Speed
2500 N-m (ci), 1000 rpm
108
-------
Figure 3-1: Developmental Cummins X15 Engine equipped with individual cylinder deactivation undergoing
engine dynamometer testing as part of the CARB Stage 3 research at SwRI.
A schematic layout of the developmental EAS with light-off SCR is shown in Figure 3-2.
Photos of the EAS showing details of its installation within the engine dynamometer test cell are
shown in Figure 3-3. The heated urea dosing system, mixer, and light-off SCR with ASC zone-
coated onto the rear of the SCR substrate were mounted downstream of the turbocharger. The
remaining components were mounted in an insulated single-box housing to improve heat
retention. Catalyst substrate specifications are summarized in Table 3-2. The total EAS volume
was approximately 4.6 times the engine displacement. The total volume of the SCR, not
including ASC, was approximately 2.8 times engine displacement. The volume of the light-off
SCR was approximately 0.58 times engine displacement and the volume of the downstream SCR
system was approximately 2.2 times engine displacement (both excluding ASC). The volume of
the downstream SCR is comparable to the sales-weighted displacement-specific SCR volume for
MY2019 Heavy FIDE, so the increase in total SCR volume relative to 2019 FTeavy HDE
applications was due to the addition of light-off SCR.
109
-------
13x6
0£
O
Internal "Switchback" Mixing Tube
tc
o
(/>
tc
o
(0
0s
tc
0s
o
o
o
o
(0
(0
(0
(0
<
4-10.5x10
Figure 3-2: Schematic layout (not to scale) of the dual-SCR EAS tested as part of the CARB Stage 3 research
at SwRI.
110
-------
Heated Doser
Mount
Zoned CSF
SCR + SCR/ASC
(Dual Path)
Figure 3-3: Developmental EAS with light-off SCR installed in engine dynamometer test cell at SwRI (upper
left, upper right) and details of the downstream, single "box" unit (lower left, lower right).
Ill
-------
Table 3-2: Summary of catalyst specifications for developmental EAS with light-off SCR
Component
Dimensions,
Dia.X
Length
(inches)
Substrate
Volume (liters)
Cell Density
(cpsi) / wall
thickness (mil)
Notes
Light-off
SCR/ASC
13X6
13.1
400/4
ASC is zone-coated to the
rearmost 2" of the light-off
SCR substrate
Zone-
coated
CDPF
13X7
15.2
300/7
Zone-coated wall-flow
substrate providing both
DOC and CDPF
functionality
SCR
10.5X4
11.4
600/4.5
Two substrates in parallel
SCR
10.5X5
14.2
600/4.5
Two substrates in parallel
SCR/ASC
10.5X5
14.2
600/4.5
Two substrates in parallel,
both with ASC zone-
coated to the rearmost 2"
of the SCR substrates
Emissions were evaluated using engine dynamometer testing over the cold-start and hot-start
FTP transient cycle, the SET, the LLC, and over specific cycles representing in-use operation
that were provided by the Engine Manufacturers Association (EMA). A baseline, original
equipment 2017 Cummins XI5 EAS was tested at a low-hour, degreened condition. The
developmental EAS with light-off SCR was tested in a degreened state and then was subjected
accelerated aging using the Diesel Aftertreatment Accelerated Aging Cycle (DAAAC).139 The
DAAAC incorporated chemical deSOx at 30 hour intervals and DPF ash maintenance at 500
hour intervals. Emissions results over the FTP, SET, and LLC for the baseline and
developmental EAS are summarized in Table 3-3 through Table 3-7. FTP Composite and SET
NOx emissions results are just over 20 mg/bhp-hr after accelerated aging equivalent to
approximately 435,000 miles and the NOx emissions results over the LLC were just under 50
mg/bhp-hr. Emissions of N2O were approximately half of the current 0.10 g/bhp-hr standards.
The infrequent regeneration factor (IRAF) for this engine is 2 mg/hp-hr for the FTP and SET and
5 mg/hp-hr for the LLC.
112
-------
Table 3-3: Baseline (degreened) emissions results for the OE Cummins EAS
Cycle
Results:
FTP
cold
95%
CI
FTP
hot
95%
CI
FTP
composite
95%
CI
SET
95%
CI
SET
(2021)
95%
CI
NOx
(mg/bhp-hr)
271
31
132
99
152
87
140
6
1005
335
PM
(mg/bhp-hr)
2.0
0.2
2.0
1.0
2.0
0.9
1.2
0.3
NA
NA
NMHC
(mg/bhp-hr)
3
7
2
2
3
3
1.7
0.2
12
25
CO
(mg/bhp-hr)
48
37
17
31
22
29
7.9
0.9
30
24
C02
(g/bhp-hr)
530
4
508
8
511
8
452
4
609
7
N20
(mg/bhp-hr)
42
2
63
9
61
11
68
8
64
NA*
* Single LLC test - no repeats.
For results where the 95% CI is greater than the average, the results are not statistically different from zero based
on a 2-sided Student's t-test at a=0.05
Table 3-4: 0-hour (degreened) emissions results for the developmental EAS system with light-off SCR
Cycle
Results:
FTP
cold
95%
CI
FTP
hot
95%
CI
FTP
composite
95%
CI
SET
95%
CI
SET
(2021)
95%
CI
NOx
(mg/bhp-hr)
54.7
0.8
11
2
17
1
8.9
0.8
20
8
PM
(mg/bhp-hr)
2
1
1.4
0.1
1.5
0.2
0.9
0.2
4.0
0.7
NMHC
(mg/bhp-hr)
23
56
12
50
14
51
1
2
47
171
CO
(mg/bhp-hr)
110
41
12
3
26
6
7
1
62
51
C02
(g/bhp-hr)
539
4
499
1
505
2
454
3
600
4
N20
(mg/bhp-hr)
39
3
47
1
46
2
53
9
43
9
For results where the 95% CI is greater than the average, the results are not statistically different from zero based
on a 2-sided Student's t-test at a=0.05
113
-------
Table 3-5: Emissions results for the developmental EAS system with light-off SCR after 334 hours of
accelerated thermal and chemical aging using the DAAAC (equivalent to approximately 145,000 miles of
operation). The SET (2021) results represent updated 40 CFR §1036.505 SET procedures.
Cycle
Results:
FTP
cold
95%
CI
FTP
hot
95%
CI
FTP
composite
95%
CI
SET
95%
CI
SET
(2021)
95%
CI
LLC
95%
CI
NOx
(mg/bhp-hr)
56
5
12
2
18
2
15
2
15
2
22
9
PM
(mg/bhp-hr)
1.3
0.2
1.6
0.5
1.6
0.4
0.7
0.0
0.7
0.4
3
6
NMHC
(mg/bhp-hr)
33
58
18
42
20
44
10
15
3.0
0.6
19
18
CO
(mg/bhp-hr)
186
94
25
10
48
22
7
1
8
2
104
45
C02
(g/bhp-hr)
541
4
506
5
511
5
454
1
450
2
602
2
N20
(mg/bhp-hr)
31
4
37
4
36
4
34
2
34
1
32
17
For results where the 95% CI is greater than the average, the results are not statistically different from zero based
on a 2-sided Student's t-test at a=0.05
Table 3-6: Emissions results for the developmental EAS system with light-off SCR after 667 hours of
accelerated thermal and chemical aging using the DAAAC (equivalent to approximately 290,000 miles of
operation). The SET (2021) results represent updated 40 CFR §1036.505 SET procedures.
Cycle
Results:
FTP
cold
95%
CI
FTP
hot
95%
CI
FTP
composite
95%
CI
SET
95%
CI
SET
(2021)
95%
CI
LLC
95%
CI
NOx
(mg/bhp-hr)
63
6
15
2
22
3
19
2
14.8
0.4
50
6
PM
(mg/bhp-hr)
1.8
0.4
2.0
1.3
2.0
1.0
1
2
0.8
0.3
4.5
0.5
NMHC
(mg/bhp-hr)
29
12
12
5
14
3
3
2
2.1
0.7
43
5
CO
(mg/bhp-hr)
227
40
139
25
151
17
10
2
11
3
305
42
C02
(g/bhp-hr)
538
1
512
2
515
2
461
2
454.1
0.7
616
1
N20
(mg/bhp-hr)
28
9
32
7
31
7
23
3
25
1
21
6
114
-------
Table 3-7: Emissions results for the developmental EAS system with light-off SCR after 1000 hours of
accelerated thermal and chemical aging using the DAAAC (equivalent to approximately 435,000 miles of
operation). The SET (2021) results represent updated 40 CFR §1036.505 SET procedures.
Cycle
Results:
FTP
cold
95%
CI
FTP
hot
95%
CI
FTP
composite
95%
CI
SET
95%
CI
SET
(2021)
95%
CI
LLC
95%
CI
NOx
(mg/bhp-hr)
61
8
17
3
23
3
22
4
20
4
47
4
PM
(mg/bhp-hr)
2
0
2
1
2
1
2
0
1
0
6
1
NMHC
(mg/bhp-hr)
41
10
24
13
26
12
5
5
4
0
28
77
CO
(mg/bhp-hr)
257
44
184
38
194
34
12
1
11
1
371
79
C02
(g/bhp-hr)
535
9
512
1
515
2
461
6
456
2
617
1
N20
(mg/bhp-hr)
37
7
42
21
41
18
23
2
23
1
18
12
For results where the 95% CI is greater than the average, the results are not statistically different from zero based
on a 2-sided Student's t-test at a=0.05
In addition to the evaluating the feasibility of the new criteria pollutant standards, we also
evaluated how CO2 was impacted on the CARB Stage 3 engine for both, the test procedures used
to show compliance with the engine standards in 40 CFR 1036.108 and the vehicles standards in
40 CFR 1037 Subpart B. To do this we evaluated how CO2 emissions changed from the base
engine on the FTP, SET, and LLC, as well as the fuel mapping test procedures defined in 40
CFR 1036.535 and 40 CFR 1036.540. For all three cycles the Stage 3 engine emitted CO2 with
no statistically significant difference at a 95% level of confidence when compared to the base
2017 Cummins XI5 engine. Comparing the CARB Stage 3 engine with the 0-hour (degreened)
aftertreatment which provides the most direct comparison with the 2017 Cummins XI5 engine
(since the 2017 Cummins XI5 aftertreatment was degreened but not chemically aged) the
percent reduction in CO2 for the FTP, SET and LLC, was 1, 0 and 1% respectively. Although,
since SwRI made changes to the thermal management strategies of the CARB Stage 3 engine
(which increased CO2 emissions from the engine), to improve NOx reduction at low SCR
temperatures after this data was taken, there is no direct comparison between the baseline engine
and the CARB Stage 3 engine. The data at equivalent of 435,000 miles that includes these
changes, the percent increase in CO2 for the FTP, SET and LLC, was 0.6, 0.7 and 1.3%
respectively, but since the aftertreatment had been aged to an equivalent of 435,000 miles prior
to these tests, which included the ash exposure from this aging, that increased the back pressure
on the engine as shown in Figure 3-4, this is not a direct comparison with the baseline engine. To
evaluate impacts to CO2 emissions of the CARB Stage 3 engine on the HD GHG Phase 2 test
procedure, the test procedures were executed with both the baseline engine and the CARB Stage
3 engine with development aged aftertreatment. The fuel maps from these tests were run in
GEM. The results from this analysis that is summarized in the SwRI Stage 3 report138, also
showed that the Stage 3 engine emitted CO2 at the same rate as the 2017 Cummins XI5.
115
-------
12
10
'c?
8
6
O,
C5
T3
Hh 4
ex 4
2
0
• Olii" average
• 667hr average
f\
' i •
I I I
1 'J1
"*h i
< L-' ui-•
A
¦ l
m jmjm mmM..
500 1000 1500
time (sec)
2000
2500
Figure 3-4: The average pressure drop across the DPF on the SET for the degreened aftertreatment and the
equivalent of 290,000 miles of operation aftertreatment
Once the CARB Stage 3 demonstration was completed a second phase of the demonstration,
that was led by EPA and is referred to as the EPA Stage 3 engine. In this phase improvements
were made to the aftertreatment by replacing the zone-coated catalyzed soot filter with a separate
DOC and diesel particulate filter (DPF) that were chemically- and hydrothermally-aged to the
equivalent of 800,000 miles and improving the mixing of the DEF with exhaust for the
underfloor SCR. A schematic of the aftertreatment is shown in Figure 3-5. The results of testing
the EPA Stage 3 engine at the equivalent of 435,000 and 600,000 miles are shown in Table 3-8
and Table 3-9. Test results at the equivalent of 800,000 miles of age will be added to the docket
when they become available.
Q = PJQ^SAnaor =GhF Doling - r-JH Sui im.ii [ = Huaiwd DtF Dnilrig J = Temp Senear
HC Uoiing I Insulated "One-baa" System
J7"lni«1nr) , 2 1D-M fl
¦ M*l"> | 13x7 _
*Vii rioiu i^fi
—flaSlU-H+i—^
1 «
Miner | internal "SwftchlwMfc' Mining Tube M". s*KI j
L _ ____ _____ _
Figure 3-5: Schematic layout (not to scale) of the dual-SCR EAS tested as part of the EPA Stage 3 research at
SwRI.
K
y
0£
o
O
(J
ffl
irt
tA
<
K
ce
o
0
o
{fl
tn
<
\a-,
Table 3-8: Emissions results for the developmental EPA Stage 3 EAS system with light-off SCR and separate
DOC and DPF after 1000 hours of accelerated thermal and chemical aging using the DAAAC (equivalent to
116
-------
approximately 435,000 miles of operation). The SET (2021) results represent updated 40 CFR §1036.505 SET
procedures.
Cycle Results:
FTP
cold
95% CI
FTP
hot
95% CI
FTP
composite
95% CI
SET
(2021)
95% CI
LLC
95% CI
NOx
(mg/bhp-hr)
55
1
14
1
20
1
17
1
29
11
PM
(mg/bhp-hr)
2
1
2
1
2
1
1
1
3
1
NMHC
(mg/bhp-hr)
25
7
9
2
12
2
1
1
35
51
CO
(mg/bhp-hr)
221
61
128
77
141
75
30
22
245
438
C02
(g/bhp-hr)
534
1
511
2
514
2
455
4
617
11
N20
(mg/bhp-hr)
84
7
74
9
76
9
24
69
132
45
Yellow highlighting designates values that are not statistically different from zero based on a 2-sided Student's t-
test at a=0.05
Table 3-9: Emissions results for the developmental EPA Stage 3 EAS system with light-off SCR and separate
DOC and DPF after 1379 hours of accelerated thermal and chemical aging using the DAAAC (equivalent to
approximately 600,000 miles of operation). The SET (2021) results represent updated 40 CFR §1036.505 SET
procedures.
Cycle
Results:
FTP
cold
95%
CI
FTP
hot
95%
CI
FTP
composite
95%
CI
SET
(2021)
95%
CI
LLC
95%
CI
NOx
(mg/bhp-hr)
61
5
21
2
27
2
24
1
33
2
PM
(mg/bhp-hr)
2
0
1
2
1
2
1
0
4
1
NMHC
(mg/bhp-hr)
23
11
7
4
9
5
1
0
16
6
CO
(mg/bhp-hr)
245
31
127
134
144
119
15
0
153
20
C02
(g/bhp-hr)
546
3
515
2
519
2
460
1
623
6
N20
(mg/bhp-hr)
69
9
57
4
58
4
30
6
64
22
For results where the 95% CI is greater than the average, the results are not statistically different from
zero based on a 2-sided Student's t-test at a=0.05
3.1.1.1.1 EPA Stage 3 Off-cycle Emissions Performance
In addition to the FTP, SET and LLC, the Stage 3 engine with the DAAAC aged
aftertreatment to an equivalent of 435,000 miles was run on 5 cycles that that cover a range of
off-cycle operation. These cycles are the CARB Southern Route Cycle, Grocery Delivery Truck
Cycle, Drayage Truck Cycle, Euro-VI ISC Cycle and the ACES 4-hour Cycle. The CARB
Southern Route Cycle is dominantly highway operation with elevation changes resulting in
extended motoring sections followed by high power operation. The Grocery Delivery Truck
Cycle represents goods delivery from regional warehouses to downtown and suburban
supermarkets and extended engine-off events characteristic of unloading events at supermarkets.
117
-------
Drayage Truck Cycle includes near dock and local operation of drayage trucks, with extended
idle and creep operation. Euro-VI ISC Cycle is modeled after Euro VI ISC route requirements
with a mix of 30% urban, 25% rural and 45% highway operation. ACES 4-hour Cycle includes 5
mode cycle developed as part of ACES program. Figure 3-6 through Figure 3-10 show the
engine speed, engine torque and vehicle speed of the cycles. The engine speed and torque shown
in the plots are specific to the Stage 3 engine.
118
-------
WVU Grocery Delfvery Cycie
6COO 8000 10000 12000 14000 16000 18000 200Q0 22000 23COO 26000 WHO 30000
Tifflf, vt
"I MJ
t 40
PU tk A
0 2000 flOCC- 6OD0 SOOO 10000 12000 14000 1GOOO lflCCO 20COD 22COO 24000 26000 28000 300M
Time, set
Figure 3-7: Grocery Delivery Truck Cycle
WVU Dray age Route
10 z
3000 fiOtifl 9 LI CO liKC I SOW iSlX'Cl 21000 24CCC- 270Q£ 30X0
70.0
&a.o
=¦ yj.o
if
=- 50.0
i Mj0
10 jQ
J1
kJj
300C tCS 9000 12.0CC- 150DO 18000 2 300G 24-OKI 27000 3COOO
Figure 3-8: Drayage Truck Cycle
119
-------
WVL) EU ISC Routt! ijwilh extended! idle segment .idded)
120
"S 100
~ 3KK) ffiOD SCOQ I2DOO 15ffl» 1BDOO 21 DM 24000 270M
Figure 3-9: Euro-VI ISC Cycle
CARB 5-Mode (Modified ACES Cycle with Extended Creep Modes)
120
0 3000 6000 9000 12000 15000 18000 21000 24000 27000
Figure 3-10: ACES 4-hour Cycle
Even with this very challenging cycle, the NOx emissions from the Stage 3 engine with
aftertreatment aged to the equivalent of 435,000 miles below the proposed Option 1 off-cycle
standards at 800,000 miles as shown in Table 3-10. Table 3-11 and Table 3-12 show that the
NHMC the emissions from this engine are well below the proposed Option 1 HC an CO
standards.
Table 3-10: Off-cycle NOx emissions results for the developmental EAS system with light-off SCR and
separate DOC and DPF after 1000 hours of accelerated thermal and chemical aging using the DAAAC
(equivalent to approximately 435,000 miles of operation)
120
-------
Bin
SNTE
Grocery
Cycle
ACES
EU ISC
Drayage
Idle (g/hr)
0.7
1.0
0.9
0.4
0.3
Low (mg/hp-hr)
41
25
29
25
15
Mid/High (mg/hp-hr)
30
18
16
33
23
Table 3-11: Off-cycle NMHC emissions results for the developmental EAS system with light-off SCR and
separate DOC and DPF after 1000 hours of accelerated thermal and chemical aging using the DAAAC
(equivalent to approximately 435,000 miles of operation)
Bin
SNTE
Grocery
Cycle
ACES
EU ISC
Drayage
Idle (g/hr)
-0.2
-0.6
-0.4
-0.1
0.0
Low (mg/hp-hr)
2
28
1
6
12
Mid/High (mg/hp-hr)
2
4
1
0
7
Table 3-12: Off-cycle CO emissions results for the developmental EAS system with light-off SCR and
separate DOC and DPF after 1000 hours of accelerated thermal and chemical aging using the DAAAC
(equivalent to approximately 435,000 miles of operation)
Bin
SNTE
Grocery
Cycle
ACES
EU ISC
Drayage
Idle (g/hr)
0.3
12.7
1.0
1.9
7.6
Low (mg/hp-hr)
40
265
66
90
461
Mid/High (mg/hp-hr)
12
21
18
18
44
3.1.1.2 EPA Heavy-duty Diesel Low NOx Demonstration Program
EPA will be evaluating two different EAS designs provided to the Agency by the
Manufacturers of Emissions Control Association (MECA). Both EAS designs incorporate light-
off SCR and dual urea injection. One of the systems will use close-coupling of the light-off SCR
(Figure 3-11, Table 3-13). The other EAS design mounts the light-off SCR closer to the other
EAS components in an under-cab position (Figure 3-12, Table 3-14).
Both EAS designs utilize conventional urea dosing systems for the downstream SCR position.
Both EAS will be evaluated using a heated urea dosing system in the upstream SCR position.
The systems will be tested using a developmental version of a MY2018 Cummins XI5 15-liter
Heavy HDE engine (Figure 3-13 and Table 3-15). EPA's developmental X15 engine is equipped
with a prototype cylinder deactivation system capable of deactivating the intake and exhaust
valves for each of the engine's six cylinders. The engine is equipped with a low-thermal inertia
air-gap exhaust manifold design and air-gap construction for exhaust piping located either
immediately downstream of the light-off SCR (close-coupled system) or immediately upstream
of the urea dosing for the light-off SCR (under-cab system).
121
-------
The under-cab EAS will also be installed in a MY2018 Navistar Day cab Class 8 Tractor
equipped with a second EPA developmental XI5 engine. The engine and cab are instrumented
to allow evaluation of NVH characteristics during truck operation when using CDA.
EPA's developmental XI5 is nearly identical to the engine equipped with cylinder
deactivation and tested as part of CARB's Stage 3 Low NOx Research Program at SwRI with the
exception of some instrumentation differences, EGR routing changes, and differences in EAS
designs.133 The EPA developmental XI5 also has identical power and torque ratings to the
engine from the Stage 3 Program. Under contracts with the EPA [contract # 68HERC20D0014],
SwRI is currently assisting us with engine instrumentation and will be providing initial CDA
calibration strategies, model-based urea dosing calibration strategies, chemical deSOx strategies
and other engine and EAS calibration strategies that were originally developed as part of the
CARB Phase 3 development effort. EPA engineering staff will independently evaluate the SwRI
engine and EAS controls using our developmental XI5 engine and will build upon the previous
CARB Stage 3 project via the development of additional engine and EAS strategies. For
example, fixed cylinder deactivation (i.e., 2, 3, or 4 cylinders deactivated) was used for active
thermal management in the CARB Stage 3 work and dynamic, individual cylinder deactivation
will be explored as part of EPA's continuation of this work. EPA's XI5 developmental engine
also incorporates a simpler EGR cooler bypass design that allows 100% bypass instead of the
50% bypass used in the CARB stage 3 work. There are also additional changes that have been
made to EGR system routing to better facilitate simultaneous EGR and dynamic CDA control
and to improve EGR distribution to the firing cylinders when CDA is active.
EPA's developmental XI5 engine is instrumented with a high-speed data acquisition system
to allow collection of cylinder pressure and crank angle position data, which will allow
determination of apparent rate of heat release and characterization of combustion phasing (e.g.,
crank angle for 50% fuel burned or CA50) and other combustion phenomena. The engine is also
instrumented with non-contact, variable reluctance sensors on the valve bridges in order to sense
valve motion during engine operation (Figure 3-14).
122
-------
Figure 3-11: EAS with close-coupled light-off SCR.
Table 3-13: Summary of catalyst specifications for developmental EAS with close-coupled light-off SCR.
Component
Substrate
Dimensions, Dia.
X Length (inches)
Substrate
Volume
(liters)
Cell Density (cpsi) /
wall thickness (mil)
Notes
Light-off
SCR/ASC
13X7
15.2
400/4
Thin wall/low-mass substrate
with ASC zone-coated to the
rearmost 2"
DOC
13X4
8.7
400/4
Thin wall/low-mass substrate
CDPF
13X8
17.4
300/7
SCR
10.5 X 7
19.9
600/2
Two substrates in parallel
SCR
10.5X5
14.2
600/2
Two substrates in parallel
SCR/ASC
10.5X5
14.2
600/2
Two substrates in parallel, both
with ASC zone-coated to the
rearmost 2" of the SCR substrates
123
-------
Figure 3-12: EAS with light-off SCR integrated into an under-cab mounting position. This system is designed
to be installed in a Navistar Daycab which is shown in the upper right.
Table 3-14: Summary of catalyst specifications for developmental EAS with light-off SCR mounted under-
cab.
Component
Substrate
Dimensions,
Dia. X Length
(inches)
Substrate
Volume (liters)
Cell Density
(cpsi) / wall
thickness (mil)
Notes
Light-off SCR/ASC
10.5X8
11.4
400/4
High porosity/low-mass substrate
with ASC zone-coated to the
rearmost 2"
DOC
10.5X6
8.5
400/4
High porosity/low-mass substrate
CDPF
13X7
15.2
300/7
SCR+SCR/ASC
13 X 7
30.5
600/3
Two substrates in series - volume is
for combined total. ASC is zone-
coated to the reannost 2" of SCR #3
124
-------
Figure 3-13: EPA developmental MY2018 Cummins X15 Heavy HDE.
Table 3-15: Major engine specifications for the MY2018 Cummins X15 engine used for EAS and CDA
development by EPA
Engine Displacement
14.95 L
Bore X Stroke
137 X 169 mm
Rated Power (id, Speed
373 kW (a) 1800 rpm
Rated Torque (2> Speed
2500 N-m @ 1000 rpm
Figure 3-14: Variable reluctance sensor for valve position measurement. The final installation will include
valve position measurement at the valve bridges for each of the six cylinders.
Each EAS design will undergo accelerated aging using a "burner aging" version of the
DAAAC.139"140 Burner aging uses a burner system fueled with diesel fuel and additized engine
lubricant to expose the EAS to both accelerated thermal aging and accelerated chemical aging.
The burner is operated over a series of controlled burner exhaust flow rates and burner exhaust
temperature setpoints that match specific engine speed and engine load setpoints during
operation of the targeted engine and EAS application (see Figure 3-15). A higher sulfur diesel
fuel (>100 ppm) is also used during DAAAC burner aging in order to accelerate sulfur exposure
The DAAAC is designed to accelerate thermal and chemical effects by approximately 10 times
normal engine operation (i.e., 1 hour of DA AAC is approximately equivalent to 10 hours of in-
125
-------
use operation). Operation on the DAAAC for 1,000 hours is approximately equivalent to Heavy
HDE operation in a truck application to the end of UL (435,000 miles).
Each EAS design will be tested using EPA's developmental XI5 engine at the accelerated
aging equivalents of 435,000, 600,000, and 800,000 miles of operation. The resulting data will
provide a better understanding of the impacts of catalyst degradation at much longer in-use
operation than captured by today's regulatory useful life. The systems under evaluation by EPA
have also been specifically designed to meet longer useful life requirements.
7200 7600 8000
8400
8800 9200
Time,s
9600 10000 10400 10800
-Speed
-Load
SCR Inlet Temp
Figure 3-15: Example of engine-speed, engine load, and resulting SCR inlet temperature used over the
DAAAC.
One of the design constraints that will be further explored within EPA's demonstration
program is nitrous oxide (N2O) emissions. N2O emissions are affected by the temperature of the
SCR catalyst, SCR catalyst formulation, urea dosing rates, and the makeup of NO and NO2
upstream of the SCR catalyst. Limiting N2O emissions is important because N2O is a strong
greenhouse gas and because highway heavy-duty engines are subject to a 0.10 g/hp-hr N2O
emissions standard set within the HD GHG Phase I rule.
3.1.2 Baseline Technology Effectiveness
The basis for our baseline technology assessment is the data provided by manufacturers as
part of the heavy-duty in-use testing requirements. This data encompasses in-use operation from
nearly 300 LHD, MHD, and HMD vehicles. Chapter 6 of the draft RIA describes how the data
was used to update the MOVES model emissions rates for HD diesel engines. Chapter 3 of the
draft RIA summarizes the in-use emissions performance of these engines.
To assess emissions levels of current production engines on the regulatory cycles we analyzed
the certification data submitted to the agency. For this analysis we focused on MY2019 and
newer engines.
126
-------
Table 3-17 include the data for the certification results of the FTP and SET cycle. The
certification results are the test results adjusted for IRAF and DF.
One observation from the data is the range of margin between the certification results and the
standard. The margin for NOx on the FTP cycle is as small as 0.02 g/hp-hr or 11% and as large
as 0.15 g/hp-hr or 300%. For the SET the average compliance margin is slightly larger than the
average margin for the FTP. For the other criteria pollutants the margin between the certification
results and the applicable standards are much larger than for NOx.
Table 3-16: Summary of certification data for FTP cycle
NOx (g/hp-hr)
PM (g/hp-hr)
NMHC (g/hp-hr)
CO (g/hp-hr)
N20 (g/hp-hr)
Standard
0.2
0.01
0.14
15.5
0.1
Average
0.13
0.00
0.01
0.18
0.07
Minimum
0.05
0.00
0.00
0.00
0.04
Maximum
0.18
0.00
0.04
1.10
0.11
Table 3-17: Summary of certification data for SET cycle
NOx (g/hp-hr)
PM (g/hp-hr)
NMHC (g/hp-hr)
CO (g/hp-hr)
Standard
0.2
0.01
0.14
15.5
Average
0.11
0.00
0.01
0.00
Minimum
0.00
0.00
0.00
0.00
Maximum
0.18
0.00
0.04
0.20
In addition to analyzing the in-use data submitted by manufacturers, we also conducted and
analyzed engine dynamometer data of three modern HD diesel engines. These engines include a
2018 Cummins B6.7, 2018 Detroit DDI5 and 2018 Navistar A26. These engines were tested on
cycles that ranged in power demand and included the LLC and the SET cycle defined in 40 CFR
1036.505. These results are summarized in Table 3-18, Table 3-19, and Table 3-20
For two of these engines both the SET in 40 CFR 1036.505 and 40 CFR 86.1333 were run.
As can be seen from the data, there was not a measurable difference between the results of these
two cycles. Both of these cycles were also run on the Stage 3 engine. These results are
summarized in Chapter 3.1.1.1. The LLC cycle was also run for these engines to understand the
performance of current engines on this cycle. The results from this cycle vary much more than
the SET and FTP. The Cummins B6.7, which is the only non-tractor engine that was tested, had
the lowest NOx at 0.35 g/hp-hr. The other two engines including the Cummins X15 shown in
Table 3-3 had results that were multiple times higher than the current standards for the FTP and
SET.
127
-------
Table 3-18: 2018 Detroit DD15 engine emissions in g/hp-hr
Cold
FTP
Hot FTP
FTP Composite
SET in
40 CFR
86.1333
SET in 40
CFR
1036.505
LLC
C02
573
550
554
481
472
642
CO
0.54
0.04
0.11
0.01
0.01
0.07
THC
0.02
0.01
0.01
0.01
0.00
0.03
NOx
0.43
0.05
0.10
0.01
0.01
0.61
NMHC
0.01
0.01
0.01
0.01
0.00
0.02
Table 3-19: 2018 Cummins B6.7 engine emissions in g/hp-hr
Cold
FTP
Hot FTP
FTP Composite
SET in
40 CFR
86.1333
SET in 40
CFR
1036.505
LLC
C02
621
569
576
486
480
908
CO
0.09
0.04
0.05
0.01
0.02
0.03
THC
0.02
0.04
0.04
0.02
0.04
0.04
NOx
0.48
0.10
0.15
0.05
0.05
0.35
NMHC
0.02
0.04
0.04
0.02
0.04
0.03
PM
0.00
0.00
0.00
0.00
0.00
Table 3-20: 2018 Navistar A26 engine emissions in g/hp-hr
Cold
FTP
Hot FTP
FTP Composite
SET in
40 CFR
86.1333
LLC
C02
546
527
529
459
710
CO
0.03
0.02
0.02
0.01
0.03
THC
0.01
0.00
0.00
0.00
0.05
NOx
0.37
0.10
0.14
0.12
0.81
NMHC
0.01
0.00
0.00
0.00
0.05
PM
0.00
0.00
0.00
0.00
3.1.3 Projected Heavv-Dutv Diesel Technology Effectiveness
Based upon data from the EPA Heavy-duty Low NOx Stage 3 Research Program (see Chapter
3.1.1.1), HDE NOx reductions of approximately 90% from today's standards are technologically
feasible when using CDA or other valvetrain-related air control strategies in combination with
dual/light-off SCR systems. Emissions results for the EPA Heavy-duty Low NOx Stage 3
Research Program after thermal and chemical aging of the EAS using the DAAAC are
summarized in Figure 3-16 through Figure 3-17. The 1000 hours of engine and EAS operation
128
-------
over the DAAAC is equivalent to approximately 435,000 miles. The SET results reflect changes
to test procedures effective for MY2021 and later heavy-duty engines and thus better reflect SET
results for MY2027 and later compliance. One thousand hours of DAAAC operation is
approximately equivalent to the current Heavy HDE full useful life of 435,000 miles. After
aging to the equivalent of 435,000 miles, NOx emissions over the FTP composite and SET were
90% and 91% below today's standards, respectively. The projected NOx emissions over the FTP
composite, SET and LLC were all below the proposed Option 1 2031 full useful life standards.
NOx emissions over the FTP composite, and SET were at or just above the intermediate useful
life standards, indicating that further calibration and development will be needed in order to meet
the intermediate useful life standards. Aging of the EAS over DAAAC to the equivalent of the
proposed Option 1 800,000-mile full useful life for 2031 Heavy HDE has been completed for the
EPA Stage 3 Research Program, however testing with the aged EPA Stage 3 EAS was not
completed in time to be included in this proposal. However, data from testing of the fully aged
EPA Stage 3 engine will be added to the docket as it becomes available.141 As discussed Chapter
3.1.1.2, the testing of the CARB Stage 3 EAS was used to inform the design of a EAS with a full
useful life design target of less than 30 mg/bhp-hr at 800,000 miles, an intermediate useful life
design target of less than 20 mg/bhp-hr at 435,000 miles, and with further improvements in NOx
emissions reduction over the SET. The improved EAS design will be evaluated as part of the
EPA Heavy-duty Diesel Low NOx Demonstration Program.
60
50
j3 40
€
hD o a
H30
X
§20
•A"
10
0
300
400 500 600 700
Equivalent Miles (x 1000)
EPA Stage 3: FTP A EPA Stage 3: SET
¦Proposed Option 1 — Proposed Option 2
800
129
-------
Figure 3-16: FTP composite and SET NOx emissions results including IRAF for the EPA Stage 3
developmental engine and emissions control system versus equivalent miles of operation.
120
100
— — r - - ¦*
80
£ 60
x
O
Z 40
20
0
300
400
500 600 700
Equivalent Miles (x 1000)
800
EPA Stage 3 Data
¦Proposed Option 1 Proposed Option 2
Figure 3-17: LLC NOx emissions results including IRAF for the EPA Stage 3 developmental engine and
emissions control system versus equivalent miles of operation.
EPA will continue to evaluate EAS durability via accelerated aging of advanced emissions
control systems. When taking into consideration the proposed Option 1 longer useful life (UL)
and the anticipated additional degradation in SCR NOx reduction between 600,000 miles and
800,000 miles, when taking lead time into consideration to 2027 and 2031, the proposed Option
1 MY 2031 and later emissions standards of 40 mg/hp-hr for FTP composite and SET, 100
mg/hp-hr for the LLC, and the respective off-cycle standards are feasible at a useful life of
800,000 miles beginning in MY 2031.
3.1.4 GHG Impacts
The combination of active and passive thermal management anticipated for meeting the
proposed standards can be designed and developed in a manner that does not pose an additional
burden for meeting the heavy-duty engine GHG standards in 40 CFR 1036.108 or the heavy-duty
vehicles GHG standards in 40 CFR 1037 Subpart B. As described in section 3.1.1.1, the design
130
-------
and calibration of the CARB Stage 3 system achieved significant NOx reductions that were GHG
neutral. The system design and calibration strategy of that system took advantage of GHG
improvements at light-load conditions from CDA and calibration changes at higher load
conditions (e.g., injection timing) that approximately offset the impact of increased backpressure
from the additional SCR catalyst volume.
3.1.5 Technology Cost
The final program is based on the need to obtain significant emissions reductions from the
heavy-duty transportation sector, and the recognition that there are technically feasible, cost
effective technologies to achieve such reductions in the 2027-2031 timeframe at reasonable cost
per vehicle, with no compromise to vehicle utility or safety. As in many prior mobile source
rulemakings, the decision on what standard to set is largely based on the effectiveness of the
emissions control technology, the cost (both per manufacturer and per vehicle) and other impacts
of implementing the technology, and the lead time needed for manufacturers to employ the
control technology. EPA also considered the need for reductions of greenhouse gases, the degree
of reductions achieved by the standards, and the impacts of the standards in terms of costs,
quantified and unquantified benefits, safety, and other impacts. The availability of technology to
achieve reductions and the cost and other aspects of this technology are therefore a central focus
of this rulemaking.
3.1.5.1 Heavy-duty Diesel Exhaust Aftertreatment System (EAS) Costs
Costs for baseline MY2019 EAS were estimated primarily based upon cost data published by
Dallman et al.142 combined with manufacturer's technical descriptions of EAS components that
were submitted as part of engine certification packages for MY2019. Manufacturer's data was
then combined into projected-sales-weighted averages for diesel engines by primary intended
service class (Light HDE, Medium HDE, and Heavy HDE) in order to account for diesel EAS
system differences by HDE category and to protect any data for individual EAS system
component specifications claimed as confidential business information. Projected-sales-
weighted average engine displacements, EAS component sizing, DOC and CDPF PGM loadings,
and other specifications are shown in Table 3-21 for all HDE categories except urban buses. The
small number of manufacturers producing engines for urban buses raised a possibility that
reverse engineering of sales-weighted-average data could occur and thus derivation of more
detailed information on individual urban bus EAS specifications. For that reason, component
level details for MY2019 urban buses were not included within publicly available data for this
analysis, but such details were included within the cost analysis for MY2019 urban bus EAS.143
In general, average engine displacement, component sizing, DOC and CDPF PGM loading, and
other EAS properties urban bus applications were generally similar to the weighted-average
values found for diesel Medium HDE. Most MY2019 Heavy HDE diesel applications used a
separate ASC downstream of the SCR, while most MY2019 diesel Light HDE, Medium HDE
and urban bus applications used zone-coating of ASC onto the outlet of the SCR substrate. In
MY2019, the majority of DPF substrates were made of extruded cordierite, but there were also a
significant number of CDPFs made from silicon carbide (SiC) and aluminum titanate (AhTiOs).
A sales-weighted-average material cost scaling factor was derived for each engine class to
account for the additional cost of SiC and AhTiOs DPF substrates relative to the cost of
cordierite substrates. The scaling factor was applied as a multiplicative correction to the DPF
131
-------
substrate costs derived from Dallman et al., which had assumed exclusive use of cordierite
material for DPF substrates.
Table 3-21: MY2019 average engine displacement and dicscl EAS specifications for Light HDE, Medium
HDE, and Heavy HDE applications based on projected-sales-weighted-averages from EPA heavy-duty diesel
engine certification data.
EAS Components
Light
HDE
Medium
HDE
Heavy
HDE
DOC
Average Engine Displacement (L)
6.65
7.54
13.23
Displacement-specific Substrate Volume (L/L)
0.56
0.61
0.74
Pt Loading (g/L)
1.01
0.93
0.73
Pd Loading (g/L)
1.16
0.35
0.27
CDPF
Displacement-specific Substrate Volume (L/L)
1.31
1.39
1.49
Material Cost Scaling Factor*
1.10
1.03
1.20
Pt Loading (g/L)
0.10
0.10
0.09
Pd Loading (g/L)
0.03
0.03
0.03
SCR
Displacement-specific Substrate Volume (L/L)
2.40
2.11
2.24
ASC*
Displacement-specific Substrate Volume (L/L)
0.43
0.38
0.40
Pt Loading (g/L)
0.064
0.054
0.077
*ASC was typically zone-coated to the SCR substrate for MY2019 Light HDE and Medium
HDE that reported specific details regarding ASC. Use of a separate ASC substrate was more
typical for MY2019 Heavy HDE applications that reported specific details regarding ASC.
Catalyst substrate sizes were scaled by engine displacement based on certification data. The
scaling of EAS component costs by catalyst size and/or engine displacement and for costs that
were approximately constant across different engine platforms (e.g., NOx and NH3 sensors) were
based upon Dallmen et al. for the following components: DOC, DPF, SCR, and ASC substrate
costs, washcoating costs, and mounting accessories; scaling and cost of DEF dosing system
components including the storage tank, level sensor, heating system, mounting accessories,
pump, injector, tubing, dosing control unit (DCU), exhaust decomposition tube/mixer; and other
components such as temperature sensors, NOx sensors, NFb sensors and associated electronics.
The EPA used costs that differed from Dallman et al. for catalyst substrate canning costs ($10/L-
catalyst-volume) and for SCR combined substrate and wash-coating costs ($28/L-catalyst-
volume). EPA estimates of these costs were based upon direct manufacturing cost data provided
by component manufacturers that were claimed as confidential.143 PGM cost estimates also
differ from Dallman et al. within the EPA analysis. EPA substituted 5-year average commodity
prices for PGM. PGM commodity prices were averaged between August 1, 2014 and August 6,
2019.144 Scaling for DOC, DPF, SCR and ASC by engine displacement was based upon
MY2019 HD engine certification data.
Estimated costs and approximate catalyst scaling and volume data for MY2019 DOCs,
CDPFs, and SCR systems are shown in Table 3-22, Table 3-23, and Table 3-24, respectively.
Total MY2019 EAS costs are summarized in Table 3-25. Note that the costs estimated for
MY2019 systems include the improvements described in draft RIA Chapter 1.1.1 that have
resulted in substrate size reduction and cost reductions between MY2010 and MY2019. These
MY2019 engine and aftertreatment costs estimates are used as the MY2027 baseline cost
presented in draft RIA 7.1.1 after conversion to 2017 dollars.
132
-------
Table 3-22: MY2019 DOC System Costs for Light HDE, Medium HDE, & Heavy HDE Applications.
EPA HD Engine Class
Light HDE
Medium HDE
Heavy HDE
Urban Bus
Sales-weighted average engine displacement (L)
6.65
7.54
13.23
DOC catalyst volume scaling factor by engine
displacement
0.561
0.607
0.744
Catalyst volume (L)
3.7
4.6
9.8
Platinum costs
$ 119
$ 135
$ 226
Palladium costs
$ 123
$ 45
$ 75
Total PGM Costs
$ 241
$ 180
$ 301
Substrate costs
$ 25
$ 31
$ 66
Washcoat
$ 48
$ 59
$ 128
Total PGMs + substrate + washcoat
$ 315
$ 270
$ 495
Canning
$ 37
$ 46
$ 98
Accessories—brackets
$ 13
$ 13
$ 16
Total direct manufacturing cost (2019 $):
$ 365
$ 329
$ 610
$ 327
Table 3-23: MY2019 CDPF System Costs for Engine-dynamometer Certified Light HDE, Medium HDE,
Heavy HDE, and Urban Bus Applications.
EPA HD Engine Class
Light HDE
Medium HDE
Heavy HDE
Urban Bus
Sales-weighted average engine displacement (L)
6.65
7.54
13.23
CDPF volume scaling factor by engine displacement
1.31
1.39
1.49
Catalyst volume (L)
8.7
10.5
19.7
Platinum
$ 28
$ 34
$ 56
Palladium
$ 7
$ 7
$ 19
Total PGMs
$ 35
$ 42
$ 74
Substrate
$ 201
$ 241
$ 454
Adjustment for substrate material cost for alternatives to
cordierite
(SiC, Al-titanate)
$ 221
$ 248
$ 546
Washcoat
$ 87
$ 105
$ 197
Total PGMs + substrate + washcoat
$ 344
$ 395
$ 817
Canning
$ 87
$ 105
$ 197
Accessories—brackets
$ 13
$ 13
$ 16
Active regeneration system
(AP sensor, piping, electronic hardware)
$ 81
$ 81
$ 81
Total direct manufacturing cost (2019 $):
$ 525
$ 594
$ 1,112
$ 591
133
-------
Table 3-24: MY2019 SCR System Costs for Engine-dynamometer Certified Light HDE, Medium HDE,
Heavy HDE, and Urban Bus Applications.
EPA HD Engine Class
Light HDE
Medium HDE
Heavy HDE
Urban Bus
Sales-weighted average engine displacement (L)
6.65
7.54
13.23
SCR catalyst volume scaling factor by engine
displacement
2.40
2.11
2.24
SCR Catalyst volume (L)
15.95
15.90
29.65
SCR substrate and washcoat costs
$ 459
$ 456
$ 830
SCR canning costs
$ 164
$ 163
$ 297
SCR Catalyst Cost
$ 622
$ 619
$ 1,127
$ 641
ASC volume scaling factor by engine displacement
0.43
0.38
0.40
ASC Catalyst Volume (L)
2.84
2.89
5.30
ASC Substrate Canning Costs
N/A*
N/A*
$ 26.51
ASC Substrate Costs
N/A*
N/A*
$ 35.52
ASC Washcoating Costs
N/A*
N/A*
$ 68.92
Pt Costs
$ 5.72
$ 4.92
$ 12.88
ASC Cost
$ 5.72
$ 4.92
$ 143.84
$ 4.66
Urea tank volume (L)
39.9
45.2
79.4
Urea tank cost
$ 210
$ 228
$ 332
Urea level sensor
$ 48
$ 48
$ 48
Urea tank accessories (brackets, bolts, spacers)
$ 37
$ 38
$ 42
Urea pump
$ 78
$ 81
$ 98
Urea injector
$ 52
$ 54
$ 65
Tubing, stainless steel
$ 109
$ 124
$ 217
Urea injection mounting parts
(brackets, bolts, gaskets, spacers, tubing connectors)
$ 30
$ 32
$ 42
Urea heating system, 200 W, 12 V DC
$ 60
$ 62
$ 75
Temperature sensor (4 per system)
$ 84
$ 84
$ 84
Decomp/mixer
$ 39
$ 39
$ 53
Dosing control unit
$ 200
$ 200
$ 200
NOx sensor cost (2 per system)
$ 340
$ 340
$ 340
Total urea system
$ 1,287
$ 1,330
$ 1,595
Total direct manufacturing cost - short term (2019 $):
$ 1,915
$ 1,954
$ 2,866
$ 2,011
*ASC assumed to be zone-coated onto SCR substrate for Light HDE & Medium HDE. Use of a separate ASC
substrate assumed for Heavy HDE applications.
Table 3-25: Summary of MY2019 EAS Costs for Engine-dynamometer Certified Light HDE, Medium HDE,
Heavy HDE, and Urban Bus Applications.
EPA HD Engine Class
Light
HDE
Medium
HDE
Heavy
HDE
Urban
Bus
Sales-weighted average engine displacement (L)
6.65
7.54
13.23
8.26
DOC - total direct manufacturing cost - short term
(2019 $):
$ 365
$ 329
$ 610
$ 327
CDPF - total direct manufacturing cost - short term
(2019 $):
$ 525
$ 594
$ 1,112
$ 591
SCR - Total direct manufacturing cost - short term
(2019 $):
$ 1,915
$ 1,954
$ 2,866
$2,011
Total EAS costs (2019 $)
$ 2,804
$ 2,877
$ 4,587
$ 2,929
134
-------
The analytical approach developed for estimating MY2019 costs was also used for estimating
the costs of proposed Option 1 MY2027-2030, MY2031 and later, and the proposed Option 2
EAS (see Table 3-26 and Table 3-27). For the proposed Option 1 and 2 standards we assumed
the same aftertreatment technology. For analyzing the costs, both SCR catalyst volume and
DOC, DPF and SCR canning costs were increased in part to account for greater useful life
requirements. For example, Heavy HDE total SCR+ASC catalyst substrate volume increased
from 2.64 times engine displacement for MY2019 systems to 3.0 times engine displacement and
4.0 times engine displacement for the analysis of the proposed Option 1 and 2 standardsMM. The
SCR catalyst volume and urea injection systems also reflect implementation of dual-SCR in
order to meet more stringent NOx emissions standards both in-use and over regulatory engine-
dynamometer test cycles. The proposed Option 1 and 2 standards include addition of an NH3
sensor, one additional NOx sensor, and the addition of a fuel dosing system located downstream
of the light-off SCR and upstream of the DOC. The system costs for the proposed Option 1 and
2 standards assume the use of zone-coated ASC for all HD diesel engine classes. We would also
anticipate the use of either high-porosity or thin-wall substrates for both the light-off SCR and
DOC for the proposed Option 1 and 2 standards. The change in bulk density would reduce
substrate mass and improve catalyst warmup. The scaling of individual EAS components for the
proposed Option 1 and 2 standards represent the upper bounds of component sizing for
developmental EAS systems designed for the EPA Heavy-duty Diesel Low NOx Demonstration
Program (see Chapter 3.1.1.2). Total SCR volume increases by approximately 79% relative to
MY2019 systems. We assume the use of light-off SCR within the cost estimates for either the
proposed Option 1 and 2 standards, with total SCR catalyst volume split between the light-off
SCR and the downstream SCR. The light-off SCR is estimated to account for approximately
25% of total SCR volume, with the downstream SCR accounting for the remaining 75% of SCR
volume. We assume the use of a single pump for both urea dosing systems within EAS using
light-off SCR. The cost of substrate canning for the DOC and DPF was increased within the cost
analyses for the proposed Option 1 and 2 standards in order to account for improvements to
substrate matting and other materials necessary for the longer useful life requirements.
1414 Note that MY2019 Heavy HDE system costs assume the use of separate SCR and ASC substrates, while the cost
analysis for the proposed Option 1 and 2 standards are based upon ASC zone-coated to the final 2" of SCR substrate
volume. Other engine categories assumed the use of zone-coated ASC for MY2019 and the proposed Option 1 and
2 standards.
135
-------
Table 3-26: Proposed Option 1 and 2 SCR System Costs for Engine-dynamometer Certified Light HDE,
Medium HDE, Heavy HDE, and Urban Bus Applications
EPA HD Engine Class
Light HDE
Medium HDE
Heavy HDE
Urban Bus
Sales-weighted average engine displacement (L)
6.65
7.54
13.23
SCR1 catalyst volume scaling factor by engine
displacement
1.07
0.940
1.00
SCR2 catalyst volume scaling factor by engine
displacement
3.21
2.83
3.00
SCR1 catalyst volume (L)*
7.1
7.1
13.2
SCR2 catalyst volume (L)*
21.4
21.3
39.7
Total SCR catalyst volume scaling factor by engine
displacement
4.3
3.8
4.0
Total SCR catalyst volume (L)*
28.5
28.4
53.0
SCR1 Substrate and washcoat cost
$ 199
$ 199
$ 370
SCR1 canning cost
$ 89
$ 89
$ 165
SCR2 substrate and washcoat cost
$ 610
$ 607
$ 1,124
SCR2 canning cost
$ 272
$ 271
$ 502
Cost for canning improvements for longer FUL
$ 90
$ 90
$ 167
SCR Catalyst Cost
$ 1,261
$ 1,255
$ 2,328
$ 1,303
ASCI volume scaling factor by engine displacement
0.190
0.171
0.179
ASCI Catalyst Volume (zone-coated) (L)
1.27
1.29
2.36
ASC2 volume scaling factor by engine displacement
0.427
0.384
0.401
ASC2 Catalyst Volume (zone-coated) (L)
2.84
2.89
5.30
ASCI Pt Costs
$ 2.55
$ 2.20
$ 5.75
ASC2 Pt Costs
$ 5.72
$ 4.92
$ 12.88
ASC Cost
$ 8.28
$ 7.12
$ 18.63
$6.73
Urea tank volume
39.9
45.2
79.4
Urea tank cost
$ 210
$ 228
$ 332
Urea level sensor
$ 48
$ 48
$ 48
Urea tank accessories (brackets, bolts, spacers)
$ 37
$ 38
$ 42
Urea pump
$ 78
$ 81
$ 98
Urea-injector-1 (heated dosing)
$ 78
$ 81
$ 98
Urea-injector-2
$ 52
$ 54
$ 65
Tubing, stainless steel
$ 153
$ 174
$ 304
Urea injection mounting parts
(brackets, bolts, gaskets, spacers, tubing connectors)
$ 59
$ 63
$ 83
Urea heating system, 200 W, 12 V DC (SCRl-only)
$ 60
$ 62
$ 75
Temperature sensor (4 per system)
$ 105
$ 105
$ 105
Mixers
$ 89
$ 89
$ 121
Dosing control unit
$ 220
$ 220
$ 220
NOx sensors (3 per system)
$ 510
$ 510
$ 510
NH3 sensor
$ 170
$ 170
$ 170
Total urea system
$ 1,869
$ 1,924
$ 2,272
$ 1,968
Total direct manufacturing cost - short term (2019 $):
$ 3,139
$ 3,187
$4,618
$ 3,278
Incremental SCR Cost from 2019 to 2031 (2019 $):
$ 1,224
$ 1,233
$ 1,752
$ 1,267
*Total SCR volume includes 2" of zone-coated ASC onto the outlet of SCR1 and SCR2
f ASC volume is included within total SCR volume
136
-------
Table 3-27: Summary of proposed Option 1 and 2 EAS Costs* for Engine-dynamometer Certified Light HDE,
Medium HDE, Heavy HDE, and Urban Bus Applications.
EPA HD Engine Class
Light HDE
Medium HDE
Heavy HDE
Urban Bus
Sales-weighted average engine displacement (L)
6.65
7.54
13.23
DOC - total direct manufacturing cost (2019 $):
$ 365
$ 329
$ 610
CDPF - total direct manufacturing cost (2019 $):
$ 525
$ 594
$ 1,112
Increased DOC+CDPF Canning Costs for longer FUL:
$ 31
$ 38
$ 74
HC dosing system (2019 $):
$ 225
$ 225
$ 225
SCR - Total direct manufacturing cost (2019 $):
$3,139
$ 3,187
$4,618
Total EAS costs (2019 $)
$ 4,285
$ 4,373
$ 6,638
$ 4,460
Total EAS Incremental Cost from 2019 to 2027 (2019
$):
$ 1,481
$ 1,495
$2,051
$ 1,531
Table Notes:
* Proposed Option 1 and 2 standards were both assumed to use comparable EAS system designs, which includes
increased DOC/DPF costs to account for increased useful life requirements and use of dual SCR systems with
heated urea injection for the light-off SCR position.
3.1.5.2 Cost Teardown Studies
Publicly available information regarding the engineering cost of new engine and vehicle
technologies is a subject of considerable interest. A number of cost analyses in the past few
years have utilized supplier price quotes on designated bills of materials as a methodology for
estimating the increased cost of vehicle improvements. In general, the actual price quotes
provided by suppliers were claimed as confidential business information and have not been
released. In addition, supplier price quotes are typically provided for near-term (e.g., 3-5 years)
estimation, as these are how contracts between OEMs and suppliers are typically written.
This methodology for estimating technology costs to the consumer has several deficiencies.
The lack of transparency regarding the data provided by suppliers does not provide an
opportunity for a full public evaluation of the information. In addition, these near-term price
quotes may not be appropriate for estimating the long-term costs of a regulatory program
implemented in the future. A large fraction of the near-term fixed costs may be recovered and no
longer part of the costs to consumers. EPA is required to evaluate the near- and long-term costs
to consumers that may result from proposed regulatory requirements. To effectively estimate and
communicate those costs, EPA requires a transparent engineering analysis that separates direct
and indirect costs for each major component in the technologies it projects will be implemented
to meet the new requirements.
EPA previously directed contractual work to develop an analytical methodology that is based
on technical knowledge of the engineering, design, and development of advanced vehicle
technology components, systems, and subsystems. In addition, the previous contractual work
performed pilot studies to demonstrate the methodology on representative vehicle categories.
A key objective for these studies was transparency— methodologies, assumptions, and inputs
well-documented, clearly explained, and releasable to the public, except to the extent that those
essential inputs included information claimed as confidential.
137
-------
3.1.5.2.1 EPA HDE Cylinder Deactivation and Variable Geometry Valvetrain Teardown Cost
Study
The cost of CDA for Light HDE, Medium HDE, Heavy HDE, and Urban Buses was
estimated based upon a detailed, tear-down study of heavy-duty diesel valvetrains, the Heavy -
Duty Engine Valvetrain Technology Cost Assessment (or "FEV Valvetrain Study").145 The
study was conducted by FEV North America, Inc. under a contract with EPANN and was
submitted to an independent peer review.146,147 The FEV Valvetrain Study investigated design
modifications to a production engine cylinder head from the Cummins XI5 engine. These
design modifications allowed the addition of a variable-geometry valvetrain system in one of two
different configurations. One configuration was implementation of individual CDA with an
integrated exhaust brake. The other configuration was implementation of late intake valve
closing (LIVC). The final objective was to evaluate t the incremental cost of CDA and LIVC
hardware as two distinct technology packages.
The cost of the CDA and LIVC technology packages were evaluated relative to a baseline
valvetrain technology represented by a 2019 Cummins XI5 engine. FEV also investigated other
valvetrain designs used by diesel HDE in order to develop fleet average per-cylinder costs for
these valvetrain technologies. The baseline and new technology packages were be required to
have similar overall performance with respect to service life and other functional objectives. For
this study, estimates for direct, indirect, and operating costs other than fuel costs were
considered. Table 3-1 shows estimated costs for application of CDA to diesel HDE based on
costs derived from the FEV study. For purposes of EPA's cost analysis of CDA applied to diesel
HDE, Light HDE costs were based on application of CDA hardware to 8-cylinder engines while
Medium HDE, Heavy HDE, and Urban Bus costs were based on application of CDA hardware to
6-cylinder engines. Both costs, airflow control, and thermal management appeared to roughly
comparable for both CDA and LIVC within the analysis, with some advantages with respect to
higher BMEP CO2 reduction for LIVC at slightly higher cost. For more details regarding the
FEV Valvetrain Study, please refer to the final report for the study within the docket for this
rule. The costs for CDA were assumed as the cost for active thermal management via valvetrain
system improvements within EPA's costs for diesel HDE for the proposed Option 1 and 2
standards.
Table 3-28: Summary of CDA Costs from Teardown Study
EPA HD Engine Class
Light HDE
Medium HDE
Heavy HDE
Urban Bus
CDA Valvetrain Hardware - Tier 1
Supplier Cost to Manufacturer (2019 $):
204.16
153.12
214.56
153.12
3.1.5.2.2 EPA Advanced EAS Teardown Cost Study
EPA also sponsored an additional study to examine in detail an advanced heavy-duty diesel
EAS technology package utilizing a dual-SCR system with heated dosing for the light-off SCR
and capable of approximately 90% NOx reduction relative to post-2010 heavy-duty diesel
emission control systems, the Heavy-Duty Vehicles Aftertreatment Systems Cost Assessment
(FEV EAS Study)148. As with the valvetrain study, this was also conducted by FEV North
U.S. EPA Contract No. 68HERC19D0008, Task Order No. 68HERH20F0041.
138
-------
America, Inc. The FEV EAS Study was not completed in time for inclusion within the cost
analysis for the proposed standards but our intention is to update EAS costs for the final rule
based upon this study. The costs associated with the advanced EAS technology packages were
evaluated relative to a baseline OE (MY2018) EAS technology representative of the current state
of design. The costs from the FEV EAS Study are summarized within Table 3-29. The
incremental costs for the advanced system were approximately $108 to 316 higher for the FEV
Study relative to the costs for systems meeting the proposed standards from the EPA analysis
presented Chapter 3.1.1.3. For details regarding the FEV EAS Study, please refer to the final
report for the study within the docket for this rule.148
Table 3-29: Summary of Costs from the FEV Exhaust Aftertreatment System Study
EPA HD Engine Class
Light HDE
Medium
HDE
Heavy HDE
Urban Bus
Total MY2019 EAS costs estimated by the FEV EAS
Study (2019 $)
$2,698.69
$2,649.28
$3,934.77
$2,729.61
Total MY2027 and later EAS costs estimated by the
FEV EAS Study (2019 $)
$4,494.79
$4,350.83
$6,093.82
$4,464.79
Total EAS Incremental Cost from 2019 to 2027 and
later costs estimated by FEV EAS Study (2019 $):
$1,796.10
$1,701.55
$2,159.05
$1,735.18
Difference in Incremental Cost for FEV EAS Study
relative to the EPA Analysis in Chapter 3.1.1.3, Table
3-27
$315.56
$206.06
$107.89
$204.39
Note: Costs from the FEV Study were originally calculated on a 2020$ basis and were converted to a 2019$ basis
for this table to allow direct comparison with other costs within this draft RIA. Costs for the advanced system
studied by FEV would be equally applicable to the costs of systems designed to meet the proposed Option 1 and 2
standards.
3.1.5.3 Closed Crankcase Systems Technology Costs
We project that this proposed requirement to close the crankcase on turbocharged engines that do
not have closed crankcase systems already will force manufacturers to rely on engineered closed
crankcase ventilation systems that filter oil from the blow-by gases prior to routing them into
either the engine intake or the exhaust system downstream of the turbocharger but upstream of
the exhaust aftertreatment system. We have estimated the initial cost of these systems to be
approximately $41 (2002$).149 To estimate the baseline cost, we multiplied $41 (2002$) by the
percentage of engines that already have closed crankcase systems, which resulted in a baseline
cost of $13 (2002$). We estimated the percentage of engines that already have closed crankcase
systems at 32.5%, based on the certification data. For our cost analysis, we converted these
estimates to 2017 dollars, which resulted in $18 (2017$) for the baseline cost and the same cost
of $37 to implement closed crankcases in the remaining CI engines for our proposed Option 1
and 2 standards.
139
-------
3.2 Spark-Ignition Technology Feasibility
3.2.1 Baseline Technology Effectiveness
3.2.1.1 EPA Baseline Real World Test Program for Exhaust Emissions
In 2018 EPA evaluated heavy-duty gasoline Class 3 and 4 vehicles from three different
manufacturers to better understand the state of criteria pollutant control technology incorporated
on gasoline engines used in these applications150. Evaluations were conducted using laboratory
chassis dynamometer testing and real-world Portable Emissions Measurement System (PEMS)
testing.
Most chassis-certified heavy-duty vehicles are subject to EPA's light-duty Tier 3 program and
these vehicles have adopted many of the emissions technologies from their light-duty
counterparts (79 FR 23414, April 28, 2014). To meet these Tier 3 emission standards,
manufacturers have been required to reduce the time needed for the catalyst to reach operational
temperature by implementing cold-start calibration strategies to reduce light-off time, and they
have also moved the catalyst closer to the engine. Manufacturers have not widely adopted the
same strategies for their heavy-duty engine-certified products, and purpose of this test program
was to observe differences between emissions performance for technologies that are available in
the market today and establish a baseline to evaluate the performance of advanced technologies
to further reduce criteria emissions.
3.2.1.1.1 Baseline Vehicles Tested
Three vehicles were chosen for evaluation based on majority market share. Two vehicles,
Class 4 (GVWR 14,001 - 16,000 pounds), with powertrains produce by General Motors (GM)
and Ford respectively, utilized engines that that were dyno certified. The third vehicle, produced
by Fiat Chrysler Automobiles (FCA), was a Class 3 (GVWR 10,001 - 14,000 pounds) chassis
certified truck that meets the Federal HDV2 Tier 3 Bin 570 standards, and has the same
powertrain (engine and transmission) that FCA uses in their Class 4 engine certified trucks. The
FCA test article had comparable gross combined vehicle weight (GCWR) as the Class 4 vehicles
tested but employed aftertreatment technology tailored for Tier 3 chassis certification. Table
3-30 lists the major specifications of the three vehicle/powertrain combinations considered in this
study.
140
-------
Table 3-30: Heavy-Duty Gasoline Vehicle Emissions Investigation Vehicle Specifications
G.M.
Ford
FCA
Configuration
Box Truck
Box Truck
Pickup
Certification
HDGE
HDGE
Tier 3 Bin 570
Model Year
2015
2016
2017
Odometer
48,000
37,000
43,000
Eng. System
NA/PFI/TWC
NA/PFI/TWC
NA/PFI/EGR/TWC
Displacement
6.0L V8
6.8L V10
6.4L V8
Power
297 hp @ 4,300 rpm
305 hp (a> 4,250 rpm
410 hp @ 4,600 rpm
Torque
372 lbft @ 4,000
rpm
420 lbft @ 3,250
rpm
429 lbft @ 4,000 rpm
Transmission
6 spd Auto
6 spd Auto
6 spd Auto
GVWR (lbs)
14,500
14,500
13,300
GCWR (lbs)
20,500
22,000
19,900
3.2.1.1.2 Baseline Tests Performed
As previously stated, two of the vehicles tested were equipped with dyno certified engines for
use in vehicles with a GVWR over 14,000 lbs while the third vehicle was a chassis certified
HDV2. These vehicles were chosen as the engines represent the bulk of the HD SI vehicle
market. The lighter HDV2 vehicle was chosen because of its chassis certification resulting in its
aftertreatment system more closely resembling what is commonly found on Tier III light-duty
vehicle.
The purpose of this particular program was to investigate the current state of emissions
performance of HD SI engine criteria emissions performance. Because cold start emissions are
not strongly emphasized in HD SI engine test, manufacturers generally locate three-way catalysts
for exhaust aftertreatment significantly downstream from the exhaust manifold. Because chassis
certification places a higher weighting on cold start results, the HDV2 vehicle we tested was
designed to reach light-off temperature sooner, and its catalyst were significantly closer to the
exhaust manifold that they would be in an engine certified configuration. Table 3-31 shows the
average distance, in meters, from the outlet of the exhaust manifold to the front face of the
catalyst substrate. Where two catalyst are used, one for each bank of cylinders on a V8 or VI0
engine, the value represents the average of the two distances.
Table 3-31: Average distance from exhaust manifold to catalyst.
Manufacturer
Average Exhaust Manifold
to Catalyst Distance
(meters)
G.M./Isuzu
2
Ford
1.6
FCA
0.9
Location of the catalyst relative to the exhaust manifold has a significant impact on overall
tailpipe emissions, as a shorter distance will enable more rapid heating and catalytic reduction
after cold start, and a longer distance reduces the maximum catalyst temperature during high
141
-------
load operation, protecting the washcoat from thermal degradation. The assumptions investigated
in this test program were as follows:
1. Gasoline stoichiometric operation and advanced three-way catalyst can provide a high
level of efficiency and nearly zero warmed-up emissions rates.
2. Vehicle weights and loads can drive high exhaust gas temperatures.
3. High exhaust gas temperatures can lead to need for fuel enrichment, to protect engine
components and the catalyst.
4. Location of the catalyst is partially dictated by exhaust gas temperature.
5. Rearward catalyst locations can hinder catalyst light-off as well as performance under
extended low-load operation.
3.2.1.1.3 On Road PI-MS testing
Each vehicle was subject to real world emissions testing and driven during the workday on
several routes EPA uses to collect real world drive emissions. During these drive evaluations,
each vehicle was equipped with a portable emissions measurement system (PEMS). These
PEMS unit are CFR 1065 compliant, with independent emissions measurement for CO, CO2,
NO, NO2 and NMHC emissions. Additionally, data is collected for exhaust flow, GPS location,
environmental conditions and selected CAN signals via the vehicles OBD connector.
Temperatures within the exhaust system were also recorded during these drive cycles: exhaust
gas temperature at the exhaust manifold outlet and catalyst inlet cone, and catalyst substrate
temperature 1-inch reward of the front face of the catalyst.
Where possible, each vehicle was tested across a range of test weights: curb weight plus
instrumentation; 90% GVRW (gross vehicle weight rating); and where possible, 90% GCWR
(gross combined weight rating, which represents truck and trailer weight.) Each real-world drive
schedule was on public roads and subject to varying traffic loads dependent on the time of day as
well as all variation as a result of traffic control. When possible, each vehicle was driven on each
route on three different days. Table 3-32 describes the routes used for collection of real-world
drive emissions.
Table 3-32: Description of Real-world PEMS Testing Routes
Route
Distance
(mi)
Avg. Speed
(mph)
Description
A
7
21
Low speed, light load
B
12
24
Medium speed, short duration high load
C
32
45
High speed, short duration high load
D
84
63
High speed, sustained high load
E
30
40
Medium speed, high load
3.2.1.1.4 Laboratory Chassis Testing
The vehicles described previously were tested at EPA's Ann Arbor, MI National Vehicle and
Fuel Emissions Laboratory (NVFEL). Laboratory testing was conducted to remove the
142
-------
variability inherent in real world PEMS testing such as engine coolant start temperature, load and
traffic conditions. Both the Ford and the GM/Isuzu vehicles engines were, as previously noted,
certified to HD engine standards on an engine dynamometer, while the FCA vehicle was chassis
certified as a complete vehicle to HDV2 standards. By testing engine certified vehicles on the
chassis rolls, it was possible to highlight the emphasis each type of certification places on the
vehicle, engine, and catalyst system design considerations. For testing purposes, each vehicle
was subjected to both cold start and hot start testing using Tier 3 (10% ethanol) certification fuel.
Each vehicle was tested at an estimated test weight (ETW) condition that represented the
vehicle's curb weight plus the weight of instrumentation, as well as 90% GVWR condition.
Because the FCA vehicle was certified as a Tier 3 HDV2 it was tested at curb weight plus 1/2
payload. Table 3-33 are the test weight and dynamometer coefficients used for this testing.
Table 3-33: Test weights and dynamometer coefficients used for NVFEL HD gasoline testing.
Test #1
Test #2
Test #3 (FCA
only)
ETW (lbs)
9,320
14,000
11,000
Target A (lbf)
123.23
87.54
87.54
Target B (lbf/mph)
0
1.399
1.399
Target C (lbf/mphA2)
0.0917
0.1215
0.1215
Each vehicle-engine combination was subject to four distinct test cycles, two of which were cold starts, with the
remaining two warm starts. Each test cycle is described below:
Cold Start, Federal Test Procedure (FTP-75). The FTP-75 is used for emissions certification
and fuel economy testing of light-duty vehicles in the United States. The FTP-75 consists of
three phases; a cold transient phase (ambient temperature 20-30 deg. C), a stabilized phase and a
hot start transient phase. The FTP-75 was used to understand the differences in cold start
strategies and catalyst architectures resulting from the differences in certification testing between
HD dynamometer and HD chassis. Figure 3-18 illustrates the FTP-75 three-phase test cycle.
143
-------
70
60
50
CL
E
-40
9
4i
a.
un
30
20
10
EPA Federal lest Procedure
Lentgh=1874 seconds, Distance=ll,04 mi, Avg. Speed 21.2 mph
Cold Start Phase
0 -L
0
A
Stabilized Phase
I
llot Start Phase
LL
200 400 600 800 1000 1200 MOO 16O0 1S00 2000
Time (sec)
Figure 3-18: FTP-75, Cold start three phase test cycle
Highway Fuel Economy Driving Schedule (HWFE). This cycle was run as double HWFE
cycles, where the first cycle is used as a warmup, or prep, and there is no emissions sampling or
recording during. This cycle was chosen to compare emissions performance under simulated
urban driving conditions, and Figure 3-19 illustrates the HFE test cycle that was used.
144
-------
70
60
-50
Q.
E
-o 40
ai
aj
Q.
aj 30
u
20
10
0 J
c
EPA Highway Fuel Economy Cycle
Length=765 seconds, Distance=10.26 mi, Avg. Speed=48.3 mph
) 100 200 300 400 500 600 700
Time (sec)
800 900
Figure 3-19: EPA Highway Fuel Economy Cycle
Phase LA92 drive cycle. The LA92 drive cycle is a CARB-developed dynamometer schedule.
The LA92 was originally developed as an inventory improvement tool, and compared to the
FTP, it has a higher top speed, a higher average speed, less idle time, fewer stops per mile, and
higher rates of acceleration. For the purposes of this comparison testing, two back-to-back LA 92
cycles were utilized, doubling the distance and creating a 4-phase test. Phase 1, a warm start, was
followed by a warmed-up Phase 2, followed by a 30-minute engine off soak. Phase 3 is a warm
start after the engine off soak and is a repeat of the earlier Phase 1 drive cycle, and Phase 4 is a
repeat of Phase 2.
145
-------
4 Phase LA-92 Drive Schedule
Length=2870 seconds, Distance=19.64 mi, Avg. Speed=24.6 mph
80
70
60
_c
Q_
50
~o
CD
01
Q_
40
00
Q)
(J
30
d)
>
20
10
0
Phase 1
500
Phase 2
h
1000
Phase 3
1500
Time (sec)
2000
Phase 4
h
t,
2500
Figure 3-20: EPA 4 phase LA92 test cycle.
HD GEM Cycle (i.e. Super Cycle). This cycle is a composite of the many cycles used in the
process of certifying trucks to HD GHG vehicle standards. The super cycle drive cycle consists
of a combination of low speed, low load cycles followed by a 10-minute idle, and a return-to-
service portion consisting of 55 and 65 mph cruise conditions. Phases 1 and 2 are consecutive
ARB Heavy Heavy-Duty Diesel Truck transient modes from the ARB HHDDT schedule, are
14.3 miles in combined length, and represent an average speed 15.4 mph. Phases 1 and 2 are
followed immediately by Phase 3, a 10-minute idle. Phase 4 is a return-to-service cycle
consisting of an acceleration from idle to a 55-mph cruise, followed by another acceleration to a
65 mph cruise, and a return to idle. Phase 4 for has an average speed of 55.8 mph, is 29.2 miles
in length, and both acceleration and deceleration rates are 0.5 mph/sec. The purpose of this cycle
was to investigate how the lower exhaust gas temperature resulting from low-load operation
affect catalyst activity and the emissions generated during a high-load, return-to-service event.
146
-------
Heavy-duty GHG Cycle (Super Cycle)
Lerigth=58Q0 seconds, Dtstance=43.4 mi, Avg. Speed=30 mph
90
80
70
60
a.
~o
v
50
0J
a.
i/i
40
u
u
JZ
30
>
20
10
0
Phase 1
0n
n
Phase 2
idle
I
iooa
M
2000
111
Phase 4
3000
Time (sec)
4000
5000
Figure 3-21: Super cycle, GEM greenhouse gas cycle
3.2.1.1.5 Baseline Results
Due to the inherent variability of real-world driving conditions, as well as the absence of
defined test cycles or in-use emission standards for HD gasoline engines, direct comparisons of
the onroad PEMS and chassis testing results cannot be used to classify the emissions
performance of any one vehicle as above or below an applicable HD standard. However,
comparisons of emission rates observed under similar conditions, time to catalyst light-off, and
overall performance of each configuration relative to the others, can be made.
Most illustrative are the test results and data acquired in the laboratory employing the FTP-75
cold start test procedure. Figure 3-22 and Figure 3-23 show cumulative hydrocarbons and NOx
respectively for each vehicle at each test weight. Each figure clearly shows that once catalyst
light-off is achieved, the sharp knee in the curve between 20 seconds and 140 seconds, emissions
rates decline significantly and remain so for the remainder of the test. The top lines in the figures
also illustrate how quickly the emissions can accumulate if catalyst light-off is delayed, allowing
most of the emissions totals to be achieved in the first minutes of operation.
147
-------
100
5
> 10
sao aoo
Ti rrir (sec)
:ddo
1?QD
V.S. (km/tl) l"urd 6.3! (9.170 lb;,) Ftitd S.BI (14,000 lb.4>
-GM/lsuiu G.OL (9320 lbs} GM/lsum G.OL (14,000 lbs> FCA 0.4L (9320 lb?)
FCA 6.4L (14,000 lbs)
Figure 3-22: FTP-75 Cold start cumulative HC comparison
3 EJ
z
r>
3
L 0
60O atXJ lOUO 1200
Time (sec)
-V.S. (km/h( Ford 6.8L (9320 lbs) Ford 6.81 (14,000 lbs>
-GM/l-sutu G.OL (9320 Ibs} GM/lsuzu G.OL (14,000 lbs) FCA C..4L (9320 lbs)
-FCA 6.41 (14,000 lbs)
Figure 3-23: FTP-75 Cold start cumulative NOx comparison
148
-------
Figure 3-24 sharpens the focus on catalyst architecture as well as possible calibration
techniques driven by the particular certification tests. Figure 3-24 shows the effect that engine
load has on the exhaust temperature of the Ford 6.8L and the GM/Isuzu 6.0L, both of which are
dyno certified with little emphasis on cold start emissions. In contrast the FCA 6.4L which is a
Tier 3 Bin 570 chassis certified vehicle shows no impact on catalyst light-off time, exhaust gas
temperature, due to increased load. The FCA 6.4L was certified to the FTP-75 which
emphasizes cold start emissions. This emphasis results in catalyst architecture, shorter exhaust
manifold to catalyst distance, as well as cold start controls and calibrations more closely related
to light-duty trucks and passenger cars.
h i...
Ford 6.8L GM/Isuzu 6.0L FCA 6.4L
Dyno Cert Dyno Cert Chassis Cert
¦ 9320# ¦ 14,000#
Figure 3-24: FTP-75 Catalyst light-off time comparison
Catalyst location cannot only affect light-off but also effects catalyst temperature during
extended periods of idle. If idle conditions are long enough catalyst temperatures may fall below
350 C, a temperature associated with reduced conversion efficiency and emissions spikes when
load is applied, and the vehicle returns to service. Figure 3-25 illustrates this condition during the
10-minute idle portion of the Super Cycle (Figure 3-21). Each vehicle enters the idle with a
catalyst temperature of approximately 500 C. Over the course of the idle, catalyst temperatures
decline and those vehicle with the largest distances from manifold to catalyst (Table 3-31) falling
below 300 C.
160
140
__ 120
O
(D
— 100
o
$ 80
o
+-»
*2 60
i—
40
20
149
-------
Empty Test Weight (9320 lbs)
10 Min. Idle
Figure 3-25: Extended idle catalyst cool down comparison
3.2.1.2 Baseline Technology for Evaporative Emissions
As mentioned in Chapter 3.3.2.3 of this draft RIA, these vehicles are subject to evaporative
emission standards, but have no refueling requirements. We are unaware of any HD SI engines
certified for incomplete vehicles that implement ORVR technologies today. For our feasibility
analysis, we believe these HD SI engines would not implement ORVR without a regulatory
driver and assumed zero adoption of ORVR technology in our baseline.
3.2.2 Projected Technology Effectiveness
The emissions performance of the advanced catalyst technologies were evaluated in EPA's
HD SI demonstration program. We also evaluated a combination of additional data sources
including MY 2019 compliance data and engine mapping data to project the effectiveness of
these technologies and inform the level of stringency in our proposed standards. A description of
these data and our analysis of them is presented in this section.
We project the effectiveness of implementing ORVR for incomplete HD SI vehicles based on
the performance of complete vehicles subject to the Tier 3 evaporative and refueling
requirements applying assumptions to account for increased fuel tank sizes.
3.2.2.1 MY 2019 IID SI Compliance Data for FTP Emission Performance
Four engine manufacturers certified HD SI engines in MY 2019. These manufacturers
certified six engine families ranging in displacement from 6.0 to 8.8 liters.131 Table 3-34 presents
the MY 2019 FTP-based emission levels reported for the three pollutants addressed by a TWC:
NOx, NMHC and CO. We labeled the engines by descending NOx level. One engine, labeled
"Cert Engine 6", is below the proposed Option 1 NOx standard for MY 2027-2030 while
150
-------
maintaining relatively lowNMHC and CO emissions. While this high performing engine, which
is available today, demonstrates that it is possible to meet the proposed Option 1 MY 2031 and
later NOx standard, we acknowledge that these certification results are representative of a shorter
useful life period than we are proposing. PM emissions for most of these engines were
undetectable and reported as zero for certification, suggesting the 5 mg/hp-hr standard proposed
for CI engines is feasible for HD SI as well.00
Table 3-34: Family Emission Limits Reported for the Six Certified HD SI Engines in MY 2019; NOx and
NMHC values are converted from g/hp-hr to mg/hp-hr to match the units of our proposed standards
Cert Engine
1
Cert Engine
2
Cert Engine
3
Cert Engine
4
Cert Engine
5
Cert Engine
6
NOx (mg/hp-hr)
160
120
104
89
70
29
NMHC (mg/hp-hr)
50
60
80
42
80
42
CO (g/hp-hr)
3.7
6.6
8.6
1.5
12.7
2.3
Fraction of
MY 2019 HD SI
Sales
2%
20%
4%
20%
48%
5%
In order to evaluate the NMHC and CO emissions, we calculated an overall average emission
rate for each pollutant that includes all engines, and separately averaged a smaller subset of three
engines with the lowest NOx levels. Table 3-35 compares these two averages with the EPA 2010
standards and results from the engine family with the best NOx emission performance of the MY
2019 compliance data.
Table 3-35: Average emission performance for Certified HD SI Engines in MY 2019
Pollutant
EPA 2010
Standard
Overall Average
Subset
Average
Best NOx
Performance
NOx (mg/hp-hr)
200
95
63
29
NMHC (mg/hp-hr)
140
59
55
42
CO (g/hp-hr)
14.4
5.9
5.5
2.3
Figure 3-26 compares the NOx, NMHC, and CO emission performance of the six engines and
displays the current EPA 2010 standard, overall average of the six engines, and average of the
subset of three engines. When calibrating their engines, SI manufacturers experience a tradeoff
in emissions performance for the three pollutants in their TWCs and each manufacturer will
optimize their emission controls differently. As expected, is the table shows no clear trend in
NMHC and CO emissions related to the reduced NOx. However, the overall average NMHC and
CO levels are both met by three engine families today and likely achievable by two additional
engine families with minor calibration changes, such as incorporation of cold start catalyst light-
off strategies and refinement of the catalyst protection fuel enrichment levels. These results
suggest FTP standards of 60 mg NMHC/hp-hr and 6 g CO/hp-hr, consistent with the overall
average NMHC and CO levels achieved today, are feasible. These emission levels would be a
low cost first step that could establish consistent emission performance across all certified HD SI
00 One engine reported a 0.005 g/hp-hr PM FEL.
151
-------
engines, and also serve as anti-backsliding standards, as manufacturers optimize their TWCs to
reduce NOx and improve component durability to meet the increased useful life periods we are
proposing.
As we consider NMHC and CO FTP standards for MY 2031, we note that one MY 2019
engine family was certified below the proposed Option 1 MY 2031 NOx standard while NMHC
and CO emissions for that family remained below the overall average across the range of
families certified. Using this low NOx engine's performance as a basis, an NMHC standard of 40
mg/hp-hr and CO standard of 4 g/hp-hr are feasible in the 2031 timeframe. While this engine's
CO performance is below 4 g/hp-hr, a balance is needed to account for uncertainty in durability
over longer useful life periods. We expect advanced catalyst formulations would provide
substantial emission reductions beyond the performance demonstrated by technologies on
engines certified today. Additionally, manufacturers have a range of other strategies from their
chassis-certified products that they can employ from improve emissions performance (see
Chapter 1.2). Consequently, it is reasonable to project all HD SI engine families should be
capable of achieving emission levels comparable to today's best performing engines.
152
-------
250
-EPA 2010
Ove ra i I Average S u bset Av erage
150
l/>
&
CO
(T
100
o
UY
£
HI
O 50
x
0 -U
20
a. 15
55
cr>
B
-------
3.2.2.2 EPA Engine Mapping Test Program for SET Emissions Estimation
To assess the potential for emission reductions in HD gasoline engines over sustained loads,
EPA evaluated engine fuel mapping data from a testing program previously performed by the
agency as part the of the HD GHG Phase 2 rule, EPA contracted SwRI to test a production MY
2015 Ford 6.8L V10 gasoline engine to assess CO2 emissions and to evaluate the new fuel
mapping test procedures developed for that rulemaking. As part of that work, the engine was run
on an early version of 40 CFR 1036.535, which is the steady state fuel mapping procedure that
requires the engine to be run at nearly 100 speed and torque points for 90 seconds. The first 60
seconds is for the engine and fuel consumption to reach stability and the last 30 seconds are
averaged to create the fuel map.
Since continuous dilute criteria emissions were also collected for the test, we recently directed
SwRI to reevaluate those results and create three versions of the data that summarized fuel
consumption and emissions (NOx, CO, NMHC and CO2) versus engine speed and torque. The
first version analyzed conditions where the engine went into power enrichment, consistent with
strategy used in the production application of the engine. The second version analyzed the
conditions where the engine controller activated a catalyst protection fuel enrichment strategy
but did so before a power enrichment strategy was activated (this is due to a programmed delay
for power enrichment of approximately one minute in the production engine controller.) The
third version analyzed only conditions where the engine maintained stoichiometric air-fuel ratio,
achieved by limiting engine load to keep exhaust temperatures slightly below the level that
would activate the thermal protection strategy programmed into the production software.
These three analyses of the data differed only in the peak torque portions of the map, as in
other portions, the engines maintained stoichiometric air-fuel ratio control for a majority of the
points (below about 90 percent throttle). For each of the maps the peak torque points were used
to calculate the A, B and C speeds as well as the torque values, so there were three unique sets of
surrogate SET test points. Emission mass rates for CO, NMHC, NOx, and CO2 and fuel
consumption were calculated from each map by interpolation of the maps at each of the SET test
points. Finally, the results were weighted according to the existing Cl-based weighting factors
outlined in 40 CFR 1036.510. The engine and emission control components were not aged to the
useful life requirements in this proposal.
The data analysis below includes operation at three distinct engine speeds described above
and at several different loads, consistent with the approximate test points that would be required
to perform the SET test procedure and then calculate a composite emissions level for the engine.
The data presented includes the emission levels, fuel consumption rates, and engine power
observed at the required SET test points, while operating in three distinct modes as allowed by
production software controls: power enrichment mode, catalyst protection enrichment mode, and
stoichiometric operation.
While not typically observed during the transient FTP test or torque mapping procedure, the
engine controller activated a power enrichment mode after approximately one minute when
throttle openings above 90%, where the extra fuel resulted in a slight increase in power. Power
enrichment is sometimes used on gasoline engines to produce approximately 5% additional
power beyond what is made when the air to fuel ratio is maintained at stoichiometry.
Stoichiometric operation is the fundamental operating mode needed for three-way catalyst
154
-------
systems to simultaneously reduce HC, CO and NOx emissions, but as described above, it is not
the mode that produces peak power.
Another operating mode observed in the data is catalyst protection fuel enrichment. When the
catalyst or other critical engine components are exposed to high exhaust gas temperatures,
damage can occur that affects the durability of these components, and manufacturers typically
implement control strategies that use a limited amount of fuel enrichment to cool the exhaust gas
and protect critical components. The fuel enrichment reduces the amount of excess oxygen that
supports the exothermic (heat releasing) reaction in the catalyst and also reduces the temperature
of the combustion gases exiting the engine. The combination of these two temperature-reducing
strategies effectively provides control of exhaust gas temperatures and protects critical exhaust
components from irreversible damage. Other strategies that maintain effective emission control,
expand the area of stoichiometric operation, and still provide protection of critical engine and
catalyst components are discussed in Chapter 2 of the RIA.
As observed in the composite SET test data below, any enrichment mode, whether for power
or catalyst protection purposes, can result in substantial emission increases and higher fuel
consumption. As seen in Table 3-36, when the engine is commanded into power enrichment
mode and is no longer maintaining stoichiometric operation, the NMHC and CO increase
substantially, and the engine consumes more fuel. The NMHC emissions are more than 10 times
higher while the CO emissions are almost 50 times higher than the stoichiometric operating
mode. NOx emissions are reduced about 60% in power enrichment mode as expected because of
the rich operation, however stoichiometric emissions NOx emissions can be improved with
catalyst design and calibration. Since this is a MY 2015 production engine, it was not designed
or calibrated for optimum emissions for sustained high load operation at mid operating speeds
such as demonstrated over the SET cycle. Improved NOx emission control required over the
FTP test cycle with this proposal is expected to also result in improvements in the NOx levels
over the SET cycle. It is important to note that this power enrichment mode is not typically
observed during the transient FTP test due to the short periods of time spent at full power loads,
which limits any power related enrichment features from activating like observed in the
sustained full power test points in the SET testing described above.
Table 3-36: Comparison of Simulated 6.8L V10 SET Composite Emissions to Proposed Standards
NOx
NMHC
CO
CO2
BSFC
(mg/hp-hr)
(mg/hp-hr)
(g/hp-hr)
(g/hp-hr)
(lb/hp-hr)
Power Enrichment Allowed
11
110
45.2
587.3
0.479
Catalyst Protection with
No Power Enrichment
19
30
11.4
617.7
0.463
Stoichiometric Operation
28
10
0.97
626.6
0.457
Proposed Option 1 Spark-
Ignition Exhaust Emission
Standards for SET Duty
35
60
6
Cycle
MY2027 - MY2030
Proposed Option 1 Spark-
Ignition Exhaust Emission
Standards for SET Duty
20
40
6
Cycle
MY2031 and later
155
-------
As discussed above and illustrated in Figure 3-27, NOx emissions remain reasonably
controlled under all operating modes; however, NMHC and CO emissions increases are closely
tied to enrichment events. The MY 2027 and MY 2031 NMHC and CO standards we are
proposing for the FTP cycle are achieved in stoichiometric operation, but CO begins to approach
today's FTP standard when catalyst protection is enabled. Power enrichment causes drastic
spikes in both NMHC and CO. We are proposing to include the SET duty cycle to incentivize
manufacturers to expand the stoichiometric operation of their HD SI engines and maintain the
maximum TWC effectiveness. The FTP standards considered in Chapter 3.1.1.8 for NMHC and
CO would require manufacturers to significantly reduce the frequency of fuel enrichment events,
yet would allow for some necessary catalyst protection and power enrichment operation. We are
proposing to apply the same numeric values for FTP and SET duty cycles for NMHC and CO
standards. As with FTP, we are proposing to maintain fuel neutral standards, such that the
NMHC and CO standards developed based on HD SI engine performance would apply to CI
engines and the NOx standards proposed for HD CI engines would apply to HD SI. These
proposed SET standards are summarized in Table 3-37.
120 -i
100
Cl
-C
Ol
>_
o
CD
E,
in
E
(9
ct:
c
o
tn
-------
temperature-tolerant catalyst washcoats, and the design and calibration strategies available to
ensure rapid catalyst light-off and reduce HC and CO emissions under high load. As indicated in
the results of the SET composite analysis in Table 3-36, where emission results for three distinct
modes of fuel control under high load operation were simulated, the emission levels for each can
vary significantly.
First, the current power enrichment mode, which is allowed solely for the purpose of
providing a modest increase in power, can produce emission results that exceed two of the
proposed Option 1 composite SET standards (see Table 3-38 below.) This fuel enrichment
approach does indeed increase power, but produces higher CO and NMHC emissions as a result,
in addition to increasing fuel consumption. Reducing the amount to time spent in this enrichment
mode, or eliminating it entirely, could result in a significant reduction in emissions. Second, the
catalyst thermal protection mode, where fuel enrichment is used solely for catalyst thermal
protection (to limit temperatures inside the catalyst to a value specified by manufacturers). This
enrichment mode, controlled by software-based temperature models in the engine control module
(ECM), also results in increased emissions, but is necessary to prevent irreversible damage to the
catalyst. As indicated in Chapter 2 technology discussion, catalyst washcoats and other related
exhaust components have progressed in recent light-duty applications and are able to tolerate
significantly higher exhaust gas temperatures while still achieving acceptable component
durability and catalytic deterioration targets. The use of these improved materials, along with the
more robust temperature models or temperature measurement devices discussed in Chapter
2.2.1.5 should result in significant reductions in CO emissions and allow engines to meet the
proposed emission standards. Also discussed in Chapter 2.2.1.7 is the use of engine down
speeding, which can avoid the high speed, high exhaust gas temperature conditions that typically
result in fuel enrichment du to engine component durability and catalyst thermal concerns. With
the integration of modern multi-speed electronically controlled transmissions, this down
speeding approach is extremely feasible and likely to also reduce engine wear and improve fuel
consumption with little perceived effect on performance under commercial and vocational
operation. Note that in order to meet GHG and fuel consumption goals, this engine has already
implemented some degree of down speeding as evident in the reduced maximum test speeds
reported by one manufacturer. The agency believes that the more recent introduction of 10-
speed transmissions provides additional opportunities for down speeding that have not yet been
explored.
Finally, the third mode of operation, where the ECM maintains a stoichiometric fuel-to-air
ratio throughout all of the SET cycle test points, and potentially, under high load in-use operation
as well, results in the greatest degree of emission control. Under stoichiometric operation,
NMHC, CO, and NOx emissions are simultaneously reduced, and the three-way catalyst be
optimized to reduce all three pollutants. This strategy is discussed in chapter 2.2.1.6, and our
analysis of MY 2019 certification data indicates thatNMCH and CO emissions are well below
the proposed Option 1 SET standards, and NOx emissions meet the proposed Option 1 MY
2027-2030 standards, and are slightly higher than the proposed Option 1 MY 2031 and later
standards. This level of NOx control was achieved without any improvements or refinements to
the calibration and control strategies that we believe manufacturers will utilize to meet the
proposed FTP standards. As observed in the analysis in Table 3-38, a slight drop in power is
observed at the three SET test points as fuel enrichment is decrease, however, this slight power
loss is also accompanied by a noticeable decrease in fuel consumption, which can be a
157
-------
potentially important operational cost benefit in a commercial vehicle applications. Similar to the
previous discussion, the agency believes that several engine hardware and control technologies,
in addition to the additional gear ratios in current transmission designs, will provide the
opportunity for maintaining stoichiometric fuel-air control under all load and speed conditions.
Table 3-38: SET Operation Mode Power Comparison
Power (kW)
Torque (Nm)
SET Set Points
SET Set Points
A
B
C
A
B
C
Power Enrichment Allowed
211
187
145
546
572
547
Catalyst Protection with
No Power Enrichment
211
182
141
542
554
524
Stoichiometric Operation
201
179
137
522
551
526
3.2.2.3 Spark-Ignition Technology Demonstration Program
EPA initiated a program with Southwest Research Institute to better understand the emissions
performance limitations of current heavy-duty spark-ignition (SI) engines as well as investigate
the feasibility of advanced three-way catalyst aftertreatment and technologies and strategies to
meet our proposed exhaust emission standards. In addition to investigating emission performance
on the FTP duty cycle, the test program evaluated the proposed SET duty cycle that are not
currently required for certification. This section describes the results of the SI demonstration
program at the time of this proposal. See Chapter 1.2 for an expanded description of these and
other technologies and strategies to address exhaust emissions for HD SI engines.
A MY 2019-certified heavy-duty gasoline engine was used for this evaluation. This particular
engine was chosen because it represented the newest design among the three most-common
engines in the market and includes technologies not normally found in HD SI engines, such as
variable valve timing (VVT) and cooled EGR. Additional considerations for selecting this engine
were the availability of chassis-certified trucks with the options and driveline configuration
desired, as well as the ability to install and operate the engine in a dynamometer test cell. Table
3-39 describes the HD SI engine that was used for this evaluation.
158
-------
Table 3-39: Major engine specifications of the MY2019 HD SI gasoline engine used for the EPA
demonstration program
Engine Component
Specification
Engine Displacement (L)
6.4
Configuration / Type
90° PushrodV-8
Bore (mm)
103.9
Stroke (mm)
94.8
Aspiration
Naturally aspirated
Injection
Sequential multi-port fuel injected
Compression Ratio
10.0:1
Engine Block Material
Cast Iron
Cylinder Head Material
Cast Aluminum, Hemispherical combustion chamber
Valve Train
2 valve per cylinder, Cam-in-block, WT, Hydraulic roller lifters
Ignition
8 individual coils, 16 spark plugs, 2 per cylinder
Exhaust Gas Recirculation
Cooled EGR
Fuel Requirement
89 Octane recommended
Peak Horsepower
360 HP @4715 rpm
Peak Torque
408 lb-ft (ai 4000 rpm
This program includes a baseline evaluation of emissions performance as well as a
demonstration of the reductions possible through the application of advanced catalyst designs
that included decreased substrate wall thickness and increased cell density, a washcoat
formulation that is more tolerant of high exhaust gas temperatures, and forward placement of the
catalyst substrate (i.e. moving a portion of the total catalyst volume closer to the engine). The
catalysts were artificially aged to represent performance equivalent of 250,000 miles of real-
world operation in a manner approved by the engine manufacturer.
We also investigated the impact of engine down-speeding and calibration changes to
demonstrate further emission reduction potential of both the baseline and advanced catalyst
configurations on the FTP and SET. As noted in Chapter 1.2.1.7, this engine down-speeding
strategy is currently used by at least one HD gasoline engine manufacturer and this lower speed
is made possible by transmission strategies preventing over-speeding, which allows the emission
controls to operate in a much more desirable and lower emitting area of engine operation. For the
down-speed testing in our demonstration program, the maximum test speed (MTS) was lowered
from the manufacturer's stated MTS of 4715 rpm to 4000 rpm.
Finally, engine calibration parameters affecting air-fuel enrichments and biasing (lambda <
1.0) were manipulated to further reduce CO emissions on the SET. Because of the limited
abilities of aftermarket vehicle engine control module programmers and the complexity of OEM
engine calibrations and control strategies, this effort was met with limited success. We do
however believe that manufacturers with access to the latest tools for calibration and complete
access to engine control strategies and calibrations will be able to optimize lambda biasing as
well as any necessary air-fuel enrichments for catalyst and engine protection. With the
capabilities previously mentioned we believe that the that through a combination of engine
down-speeding and calibration optimization that the emissions goals proposed are achievable.
Installation of the engine in the test cell included instrumenting the engine's aftertreatment
with thermocouples at exhaust manifold exit, catalyst inlet, and 1-inch rearward of the catalyst
159
-------
front face. For all engine tests, CAN data from the engine control module, including, but not
limited to engine speed, short and long-term fuel correction, and spark advance were recorded.
We evaluated the following test procedures, performing three repeats of each cycle:
• HD SI FTP cycle (40 CFR 1036.510(a)(1))
• Engine mapping (40 CFR 1036.535 and 1036.540)
• HD SET (40 CFR 1036.505)
In all tests, we measured NOx, CO, PM, and NMHC, as well as the GHG-related parameters
of brake-specific fuel consumption (BSFC), CFU' and CO2. Emissions were measured from two
locations throughout each test cycle: before the catalytic aftertreatment and at the tailpipe.
Table 3-40 and Table 3-41 present results representing application of advanced catalyst
technology and engine down-speeding and a combination of engine down-speeding and
calibration for the FTP and SET duty cycles. In all test cases, FTP and SET, the results show
NOx and NMHC to be at or below the proposed Option 1 standards for both MY 2027-2030 and
MY 2031 and later. For the FTP results in Table 3-40, with regards to the 4715 rpm MTS, CO is
below the proposed Option 1 MY 2027 standards but slightly above the 2031 proposed limit. For
the 4000 rpm MTS, CO is significantly below each proposed standard. Figure 3-28 illustrates the
CO breakthrough associated with the 4715 MTS. The lambda excursions seen in Figure 3-28 are
a direct result of catalyst protection lambda enrichment. Please see the discussion in Chapter 1.2
regarding engine operating modes and possible calibration philosophy to address excess CO
emissions. Calibration changes to address the borderline NOx and SET CO levels, richer lambda
bias and no throttle-based enrichment, resulted in slight increases in NOx and CO.
Table 3-40: Spark-Ignition Demonstration Program Preliminary FTP Results
NOx
(mg/hp-hr)
CO
(g/hp-hr)
NMHC
(mg/hp-hr)
PM
(mg/hp-hr)
BSFC
(lb/hp-hr)
Proposed Option 1
MY 2027-2030
35
6.0
60
5
Proposed Option 1
MY 2031 and later
20
4.0
40
5
250k Catalysts
4715 RPM MTS
19
4.9
32
4.8
0.456
250k Catalysts
4000 RPM MTS
18
0.25
35
4.5
0.448
250k Catalysts
4000 RPM MTS
Modified Cal
21
0.99
1
4.4
0.448
160
-------
4715 rpm Vs. 4000 rpm MTS Comparison
FTP CO Reduction
1.5
¦a
-Q
0.5
Mill
i.
A
30000
00
20000
200 400 600 800
Time (seconds)
1000
1200
oc
10000 <3
O
U
•4715 rpm Lambda 4000 rpm Lambda
¦4175 rpm CO 4000 rpm CO
Figure 3-28: Engine RPM Down-Speeding FTP CO Comparison
Table 3-41 compares the emissions results for the SET cycle at MTSs of 4715 rpm and 4000
rpm. Like the FTP results NMHC and NOx are well below the proposed standards. Unlike the
FTP results both the 4715 rpm and 4000 rpm MTS operation show CO above the proposed
standards. Calibration changes, described above, were then applied to attempt further reducing
the CO. In the first two cases there was no calibration attempted to decrease the amount of air-
fuel enrichment during operation, Figure 3-29. Because no calibration was attempted the
decrease in all emissions constituents is due to the lower operating loads and the need for air-fuel
enrichments commensurate with the lower MTS. When calibration changes to richer lambda
biasing and removal of throttle-based enrichments were applied NOx did decrease while CO
increased even higher above goal.
Table 3-41: Spark-Ignition Demonstration Program Preliminary SET Results
NOx
CO
NMHC
PM
BSFC
(mg/hp-hr)
(g/hp-hr)
(mg/hp-hr)
(mg/hp-hr)
(lb/hp-hr)
Proposed Option 1
MY 2027-2030
50
6.0
60
Proposed Option 1
MY 2031 and later
20
4.0
40
5
250k Catalysts
4715 RPM MTS
8
36.7
6
7
0.462
250k Catalysts
4000 RPM MTS
5
7.21
1
3
0.437
250k Catalysts
4000 RPM MTS
1
9.65
1
3
0.438
Modified Cal
161
-------
1.4
1.2
1
¦a
-Q
0.8
E
ro
0.6
i
0.4
0.2
0
4715 rpm Vs. 4000 rpm MTS Comparison
SET CO Reduction
-ji*r
¦rw
—[-
n [1
n
g
—b
.n
I
n n
[ r
J h
100000
80000
60000 iB
40000
20000
1000 2000 3000 4000
Time (seconds)
5000
6000
00
ro
oc
10
10
ro
O
U
¦4715 rpm Lambda 4000 rpm Lambda
¦4175 rpm CO 4000 rpm CO
Figure 3-29: Engine RPM Down-Speeding SET CO Comparison
As mentioned previously, SwRI undertook a calibration effort to address the level of CO
witnessed, 7.21 g/hp-hr, during SET testing and lower NOx for both the FTP and SET. Because
of the limited abilities of aftermarket vehicle engine control module programmers and the
complexity of OEM engine calibrations and control strategies, this effort was met with limited
success. We do however believe that manufacturers with access to the latest tools for calibration
and complete access to engine control strategies and calibrations will be able to optimize lambda
biasing as well as any necessary air-fuel enrichments for catalyst and engine protection. With the
capabilities previously mentioned we believe that the that through a combination of engine
down-speeding and calibration optimization that the emissions goals proposed are achievable.
3.2.2.4 Refueling Emissions Technology Effectiveness
As described Chapter 2.3.2 of this draft RIA, HD SI engines certified as incomplete heavy-
duty vehicles are not currently required to meet ORVR. The technology package we considered
for these engines is based on the technologies implemented by chassis-certified complete
vehicles to meet the evaporative and refueling requirements of Tier 3. The technology package
includes four main equipment components and strategies that incomplete heavy-duty vehicles
would need to update to implement ORVR: increased working capacity of the carbon canister to
handle additional vapors volumes, flow control valves to manage vapor flow pathway during
refueling, filler pipe and seal to prevent vapors from escaping, and the purge system and
management of the additional stored fuel vapors. Chapter 1.2.4 includes descriptions of these
technologies. The assumptions we applied to account for the larger fuel tanks and other
considerations for larger incomplete vehicles are summarized in Section 3.1.1.13 where we
present our projected costs.
162
-------
The proposed refueling controls would result in 29.2% lower VOC and Benzene by 2031,
80.2% lower by 2040 and 88.5% lower by 2045 for heavy duty gasoline vehicles over 14,000
lbs. See the discussion and table in Chapter 5, Section 5.3.3.
3.2.3 Technology Cost
For this analysis of the aftertreatment costs, heavy-duty spark-ignition (HD SI) engines are
categorized by the type of fuel they use: liquid fuels (i.e., gasoline, gasoline-ethanol blends, and
ethanol) or gaseous fuels (i.e., compressed natural gas and liquified petroleum gas). The gaseous-
fueled category includes engines derived from SI platforms and engines converted to SI from
heavy-duty CI engines. The heavy-duty SI engine category is further divided into heavy heavy-
duty (HHD) and urban bus. We projected the costs of achieving the proposed HD SI engine
exhaust emission standards based on the technologies we evaluated in our demonstration
program (see Chapter 3.1.1.10).
3.2.3.1 Spark Ignition Exhaust Aftertreatment System Cost Analysis
Manufacturers will optimize the design of their aftertreatment systems specific to their
different vehicles. Primary considerations include cost, light-off performance, warmed-up
conversion efficiency, and the exhaust temperatures encountered by the vehicle during high-load
operation. Vehicles having low power-to-weight ratios will tend to have higher exhaust gas
temperatures and exhaust gas flow which will result in a different design when compared to
vehicles having higher power-to-weight ratios.
Manufacturers and catalyst suppliers perform detailed studies evaluating the cost and
emission performance of aftertreatment systems. It is anticipated that manufacturers will
optimize their aftertreatment systems to achieve the proposed heavy-duty emission standards and
meet the durability criteria for all vehicle classes.
Similar to the CI engine cost analysis, costs for baseline HD SI engine aftertreatment systems
were estimated using cost data published by Dallman et al.152, Pasoda er al.153 as well as data
from manufacturer's technical descriptions of aftertreatment catalyst components submitted as
part of engine certification packages for MY2019. Manufacturer's data was then combined into
projected sales-weighted averages by type of fuel (liquid and gaseous fuels), including two
distinct categories for gaseous-fueled engines identified as heavy heavy-duty and urban bus that
have distinctly different aftertreatment demands.
Baseline projected sales-weighted average engine displacements, catalyst volumes, PGM
loadings and costs are shown in Table 3-42 for both liquid and gaseous fueled SI engines. As
mentioned previously, these are based on certification data from MY 2019. These MY2019
engine and aftertreatment costs estimates are used as the MY2027 baseline cost presented in
draft RIA Chapter 7.1.1 after conversion to 2017 dollars.
163
-------
Table 3-42: 2019 MY Sales-Weighted Baseline SI Engine Technology Costs (2019$)
Liquid Fueled
SI Engine
Gaseous Fueled
SI Engine
Engine Displacement (L)
6.6
7.2
Total TWC Volume (L)
4.3
5.1
CATv/ENGd Ratio (L/L)
0.65
0.72
Total Pt ($)
$0
$0
Total Pd ($)
$53
$68
Total Rh ($)
$172
$179
Substrate cost ($)
$56
$68
Washcoat cost ($)
$26
$31
Canning cost ($)
$15
$19
Total cost ($2019)
$322
$365
We separately evaluated two distinct categories for the three gaseous-fueled HD SI engines
that certified to California's optional and more stringent 0.02 g/hp-hr NOx standard in MY 2019:
HHD and urban bus. One engine is derived from a traditional SI gasoline-fueled engine and two
are converted from CI diesel-fueled engines. Given the small number of engines in each of these
categories, we are not publicly releasing the component-level details for MY 2019 HD SI, HHD
and urban bus engines, but summarize the total costs for these categories. Table 3-43 shows the
baseline aftertreatment cost for HHD and urban bus gaseous-fueled engines.
Table 3-43: 2019 MY HHD and Urban Bus Gaseous-fueled Technology Baseline Costs (2019$)
Gaseous Fueled
SI Engine
Gaseous Fueled
HHD Engine
Gaseous Fueled
Urban Bus Engine
Engine Displacement (L)
6.8
11.9
8.9
Total cost ($2019)
$641
$3,348
$2,511
As mentioned previously, the three MY2019 gaseous-fueled HD SI, HHD and urban bus
engines currently meet a 0.02 g/hp-hr NOx standard, and we assumed that additional technology
would not be needed for these engines to meet the proposed standards in future model years.
However, it is reasonable to believe that improvements in the materials and design of the catalyst
substrate support structure (e.g., can material, mat, seals, etc.) will be needed to achieve
durability over the longer useful life and we estimated a nominal addition per-engine cost for
these engine categories.
For the other gaseous fueled engines category, we assumed the same technologies would be
used to meet the proposed Option 1 MY 2027-2030 and MY 2031 and later standards and that
manufacturers would update calibration to achieve lower emissions in the final step of the rule.
Table 3-44 shows the MY 2027 and later gaseous-fueled engine technology costs adjusted for
improved catalyst and component durability.
164
-------
Table 3-44: Projected Gaseous Fueled Engine Technology Cost based on Proposed Standards (2019$)
Gaseous Fueled
SI Engine
Gaseous Fueled
HHD Engine
Gaseous Fueled
Urban Bus Engine
MY 2027 and later
Total cost ($2019)
$646
$3,376
$2,531
The MY2027 and MY2031 technology cost for the liquid fueled SI engines are based on the
demonstration engine described in Chapter 3.1.1.10. Costs were estimated using the same
Dallman et al.154 and Pasoda et al.155 data as our baseline estimates and data from the specific
aftertreatment catalyst components used for the HD SI demonstration program. We did not make
any specific cost adjustments to account for the lengthened useful life, since the aftertreatment
system used in the demonstration program represented catalysts aged to 250,000 miles. Table
3-45 contains the details of this analysis.
Table 3-45: Projected Liquid Fueled SI Engine Technology Cost to meet Proposed Standards ($2019$)
Technology Description
MY 2027 and later
Liquid Fueled SI Engine
Total TWC Volume (L)
5.8
CATv/ENGd Ratio (L/L)
0.91
Light-off Catalyst
Number of Catalysts
2
L.O. Catalyst Volume (L)
0.82
Total Pt ($)
$0.0
Total Pd ($)
$268
Total Rh ($)
$40
Substrate Cost ($)
$21
Washcoat Cost ($)
$10
Canning Cost ($)
$6
Underfloor Catalyst
Number of Catalyst
2
U.F. Catalyst Volume (L)
2.1
Total Pt ($)
$0
Total Pd ($)
$187
Total Rh ($)
$102
Substrate Cost ($)
$55
Washcoat Cost ($)
$25
Canning Cost ($)
$18
Total Demonstration TWC Cost ($2019)
$732
Table 3-46 summarizes the costs for each of the HD SI engine categories evaluated in this
analysis. These technologies costs are used in the analysis to determine the overall costs of the
proposed program, as detailed in Chapter 7 of this draft RIA.
165
-------
Table 3-46: Summary of HD SI Engine Technology Cost Comparison
Cost Packages (2019$)
Liquid Fueled
SI Engine
Gaseous Fueled
SI Engine
SI HHD
SI Urban Bus
Baseline Technology
$322
$365
$3,348
$2,511
MY 2027 Technology
$732
$646
$3,376
$2,531
MY 2027 Incremental
$410
$281
$28
$20
3.2.3.2 Onboard Refueling Vapor Recovery Anticipated Costs
As described Chapter 2.3.2 of this draft RIA, HD SI engines certified as incomplete heavy-
duty vehicles are not currently required to meet ORVR. There are four main equipment
components and strategies incomplete heavy-duty vehicles would need to update to implement
ORVR: increased working capacity of the carbon canister to handle additional vapors volumes,
flow control valves to manage vapor flow pathway during refueling, filler pipe and seal to
prevent vapors from escaping, and the purge system and management of the additional stored
fuel vapors. Chapter 1.2.4 includes more information on these technologies. The associated costs
for these updates are summarized below. We are proposing to have ORVR requirements extend
to heavy-duty gasoline engines in incomplete vehicles starting in model year 2027. For our cost
analysis, we assumed all heavy-duty gasoline engines that are identified as LHD, MHD and
HHD in MOVES will have an average of a 70-gallon fuel tank.
Capturing the increased vapor volume from the vapor displaced during a refueling event will
require canisters to increase vapor or "working" capacity approximately 15%-40% depending on
the individual vehicle systems (i.e., fuel tank size). This can be achieved by increasing the
canister volume using conventional carbon, the fundamental material used to store fuel vapors. A
typical Tier 3 canister has approximately 5.1 liters of conventional carbon to capture overnight
diurnal evaporative emissions for a 70-gallon fuel tank. An increase in required capacity to allow
refueling vapors to be captured would result in the need for an additional 1.9 liters of
conventional carbon. A change in canister volume to accommodate additional carbon could
include increased costs for retooling and additional canister plastic material, as well as design
considerations to fit the larger canister on the vehicle.
An alternative to retooling for a larger single canister could be to add a second canister for the
extra canister volume to avoid the re-tooling costs. Several smaller volume canisters are
available on the market today. Another approach based on discussions with canister and carbon
manufacturers, can be achieved by using a higher adsorption carbon along with modifications to
compartmentalization within the existing canister plastic shell that will increase the canister
working capacity without requiring a larger canister size.
Additionally, there are two primary technologies used to prevent vapors from escaping into
the atmosphere through the filler neck and around the fuel nozzle area when the vehicle is
refueling that can affect the canister vapor capacity design requirements: a mechanical seal
which makes direct physical contact with the refueling nozzle to create a nozzle to filler neck
seal; or a liquid seal further down in the filler pipe which uses the liquid fuel mass flowing down
the filler pipe and entering the tank to hydraulically prevent vapors from migrating back up the
fill pipe. There is approximately a 20% reduction in carbon volume required if a mechanical seal
is used at the filler neck versus a liquid seal approach. Typically, mechanical seals have not been
a preferred approach because it introduces another wearable part that can deteriorate and would
166
-------
potentially need to be monitored under OBD requirements, as well as requiring replacement
during the useful life of the vehicle if it wears prematurely. Facing the choices available for the
larger volume fuel tanks and the need for a larger matching carbon containing canister to handle
these large quantities of fuel vapors, it may be more economical to choose a mechanical seal
design to avoid excess canister carbon requirements and possible retooling charges, but we share
our assumptions and cost estimates for both seal options in Table 3-47 and Table 3-48. A
mechanical seal approach costs approximately $10.00 per seal. A dual tank may require two
seals if dual filler necks are used instead of a single filler neck and transfer pump to move fuel
between the two tanks.
The second required equipment update would be to install flow control valves, which may be
integrated into existing roll-over/vapor lines. The flow control valves are needed to manage the
vapors during the refueling event by providing a low restriction pathway for vapors to enter the
canister for adsorption and storage on the carbon materials. We anticipate vehicles would require
on average one valve per vehicle which would be approximately $6.50 per valve. A dual tank
system may require a flow control valve system per tank depending on the design approach.
Thirdly, as mentioned above, a filler pipe and seal system would be needed for each filler
nozzle to keep the vapors contained during refueling. Manufacturers have the option of a
mechanical seal that costs approximately $10.00 per seal, or a liquid seal which costs effectively
nothing for the seal, but would require approximately $15 of new hardware modifications to
provide enough back pressure to stop the refueling nozzle fuel flow when tank reaches full
capacity to avoid spitback of the liquid fuel.
Lastly, the engine control of the canister purge rates may need to be addressed. This update
would include calibration improvements and potentially additional hardware to ensure adequate
purge volumes are achieved as required to maintain an appropriate canister state to manage
vapors generated during diurnal and subsequent refueling events. If required for a dual tank
system, an extra purge valve may be needed if the two tank system maintains independent
canisters instead of a single common cannister as observed in dual tank single canister light-duty
applications.
Table 3-47 shows our calculations estimating the amount of extra canister size for
conventional carbon for a 70-gallon tank, using Tier 3 requirements as a baseline. Currently
under Tier 3 requirements the canister and purge strategy is sized for the diurnal test and
designed to meet the Bleed Emissions Test Procedure (BETP) requirements. During the diurnal
test, the canister is loaded with hydrocarbons over two or three days, allowing the hydrocarbons
to load a conventional carbon canister (1500 GWC, gasoline working capacity) at a 70%
efficiency. During a refueling event, which takes place over a few minutes, the vapor from the
gas tank is quickly loaded onto the carbon in the canister with an ORVR system, causing the
efficiency of the canister loading to drop to 50% efficiency mainly because the high volume of
fuel vapors required to be adsorbed in the short period of a refueling event. Typically, a design
safety margin adds an extra 10% carbon to ensure adequate performance over the life of the
system. Therefore, even though there is typically less fuel vapor mass generated and managed
during a refueling event than is generated over a three-day diurnal time period, the amount of
carbon that is necessary to contain the vapor is higher for a refueling event.
In order for carbon in the canisters to be effective at managing vapors for diurnals and
refueling events, the vehicle engine must sufficiently purge the canister during engine operation
167
-------
in preparation for future events that will require vapor adsorbing capacity. The purge
requirements are shown in Table 3-47. The diurnal drive cycle is only 30 minutes and targets 200
bed volumes of purge to clean the canister before the evaporative emissions test. When the bed
volumes of purge are multiplied by the canister volume, the total purge volume can be
calculated. The total purge volume divided by the number of minutes driving gives us the
average purge rate. An ORVR test requires proper conditioning for a very clean canister in order
to pass the ORVR test. To clean out the canister over the 97 minutes of driving cycles for the
ORVR prep, there is a much higher amount of bed volumes necessary, therefore the purge rate
required is also higher. Table 3-48 shows cost estimations for the different approaches. For our
cost analysis described in Chapter 7, we used $25, which is the average of all approaches
considered, as the cost estimate for the additional canister capacity and hardware for our
proposed ORVR requirements.
168
-------
Table 3-47: Assumptions for gasoline-fueled heavy-duty spark-ignition vehicles for conventional carbon
requirements to meet the proposed ORVR
Tier 3 Baseline
Proposed Requirements
Mechanical Seal
Liquid Seal
Diurnal
ORVR
Diurnal Heat Build
72-96°F
80°F
RVP
9 psi
Nominal Tank Volume
70 gallons
Fill Volume
40%
10% to 100%
Air Ingestion Rate
0%
13.50%
Mass Vented per heat build, g/day
120
Mass Vented per refueling event
255
315
Hot Soak Vapor Load
5
Mass Vented over 48-hour test
227.2
Mass Vented over 72-hour test
323.3
1500 GWC, g/L (Efficiency)3
70
50
50
Excess Capacity
10%
10%
10%
Canister Volume, liters'3
4 8-hour
3.6
72-hour
5.1
ORVR0
5.6
6.9
Limiting Drive Cycle, minutes
30
97
97
Bed Volumes Purge
200
646
646
Total Purge Volume, liters'1
1020
3618
4457
Average Purge Rate, LPMe
34
37
46
BETP Purge
37
46
a Efficiency of conventional carbon
b Canister Volume = l.l(mass vented)/ 1500 GWC (Efficiency)
0 ORVR adds .5 liters and 1.8 liters for Mechanical Seal and Liquid Seal respectively
d Total Purge Volume, liters = canister volume, liters * Bed Volumes Purge represent the potential volume of
purge for the 97 minute drive cycle used for the ORVR test procedure. Required purge volume to clean out the
canister of fuel vapors for the larger ORVR canisters is likely much lower, approximately the ratio of new canister
volume to the previous canister volume multiplied by the target bed volumes (220 bed volumes for a mechanical
seal and 271 bed volumes for a liquid seal)
e Average Purge Rate, LPM = Total Purge Volume, liters / Limiting Drive Cycle, minutes however as noted in (d),
this is not necessarily the required purge volumes or rates
169
-------
Table 3-48: Estimated Costs for ORVR Over Tier 3 as Baseline
Liquid Seal
Mechanical Seal
New
Canister
Dual Existing Canisters in
Series
New
Canister
Dual Existing Canisters in
Series
Additional
Canister Costs
$20
$15
$8
$8
Additional
Tooling3
$0.50
$0.50
Flow Control
Valves
$6.50
$6.50
Seal
$0
$0
$10
Totalb
$27
$22
$25
" Assumes the retooling costs will be spread over a five-year period
b Possible additional hardware for spitback requirements
170
-------
Chapter 4 Health and Environmental Impacts
4.1 Health Effects Associated with Exposure to Pollutants
Heavy duty vehicles emit pollutants that contribute to ambient concentrations of ozone, PM,
NO2, CO, and air toxics. A discussion of the health effects associated with exposure to these
pollutants is presented in this section of the draft RIA. The following discussion of health
impacts is mainly focused on describing the effects of air pollution on the population in general.
Additionally, children are recognized to have increased vulnerability and susceptibility related to
air pollution and other environmental exposures; this and effects for other vulnerable and
susceptible groups are discussed below.
4.1.1 Ozone
4.1.1.1 Background on Ozone
Ground4evel ozone pollution forms in areas with high concentrations of ambient NOx and
VOCs when solar radiation is high. Major U.S. sources of NOx are highway and nonroad motor
vehicles and engines, power plants, and other industrial sources, with natural sources, such as
soil, vegetation, and lightning, serving as smaller sources. Vegetation is the dominant source of
VOCs in the U.S. Volatile consumer and commercial products, such as propellants and solvents,
highway and nonroad vehicles, engines, fires, and industrial sources also contribute to the
atmospheric burden of VOCs at ground-level.
The processes underlying ozone formation, transport, and accumulation are complex.
Ground-level ozone is produced and destroyed by an interwoven network of free radical
reactions involving the hydroxyl radical (OH), NO, NO2, and complex reaction intermediates
derived from VOCs. Many of these reactions are sensitive to temperature and available sunlight.
High ozone events most often occur when ambient temperatures and sunlight intensities remain
high for several days under stagnant conditions. Ozone and its precursors can also be transported
hundreds of miles downwind, which can lead to elevated ozone levels in areas with otherwise
low VOC or NOx emissions. As an air mass moves and is exposed to changing ambient
concentrations of NOx and VOCs, the ozone photochemical regime (relative sensitivity of ozone
formation to NOx and VOC emissions) can change.
When ambient VOC concentrations are high, comparatively small amounts of NOx catalyze
rapid ozone formation. Without available NOx, ground-level ozone production is severely
limited, and VOC reductions would have little impact on ozone concentrations. Photochemistry
under these conditions is said to be "NOx-limited." When NOx levels are sufficiently high,
faster NO2 oxidation consumes more radicals, dampening ozone production. Under these "VOC-
limited" conditions (also referred to as "NOx-saturated" conditions), VOC reductions are
effective in reducing ozone, and NOx can react directly with ozone resulting in suppressed ozone
concentrations near NOx emission sources. Under these NOx-saturated conditions, NOx
reductions can actually increase local ozone under certain circumstances, but overall ozone
production (considering downwind formation) decreases and even in VOC-limited areas, NOx
reductions are not expected to increase ozone levels if the NOx reductions are sufficiently large -
large enough to become NOx-limited.
171
-------
4.1.1.2 Health Effects Associated with Exposure to Ozone
This section provides a summary of the health effects associated with exposure to ambient
concentrations of ozone.pp The information in this section is based on the information and
conclusions in the April 2020 Integrated Science Assessment for Ozone (Ozone ISA).156 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. QQ 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 metabolic effects, including metabolic syndrome (i.e., changes in insulin or
glucose levels, cholesterol levels, obesity and blood pressure) and complications due to diabetes
are likely to be causally associated with short-term exposure to ozone, and that evidence is
suggestive of a causal relationship between cardiovascular effects, central nervous system effects
and total mortality 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,
metabolic effects, reproductive and developmental effects, central nervous system effects and
total mortality. The evidence is inadequate to infer a causal relationship between chronic ozone
exposure and increased risk of 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
pp 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.
QQ 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.
172
-------
greater fraction of air through the mouth.RR Children also have a higher asthma prevalence
compared to adults. Recent epidemiologic studies provide generally consistent evidence that
long-term ozone exposure is associated with the development of asthma in children. Studies
comparing age groups reported higher magnitude associations for short-term ozone exposure and
respiratory hospital admissions and emergency room visits among children than for adults. Panel
studies also provide support for experimental studies with consistent associations between short-
term ozone exposure and lung function and pulmonary inflammation in healthy children.
Additional children's vulnerability and susceptibility factors are listed in Section XII of the
Preamble.
4.1.2 Particulate Matter
4.1.2.1 Background on Particulate Matter
Particulate matter (PM) is a complex mixture of solid particles and liquid droplets distributed
among numerous atmospheric gases which interact with solid and liquid phases. Particles in the
atmosphere range in size from less than 0.01 to more than 10 micrometers (|im) in diameter.157
Atmospheric particles can be grouped into several classes according to their aerodynamic
diameter and physical sizes. Generally, the three broad classes of particles include ultrafine
particles (UFPs, generally considered as particles 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" (PM10-2.5, particles with a nominal mean aerodynamic diameter
greater than 2.5 |im and less than or equal to 10 |im). EPA currently has standards that regulate
PM2.5 and PMio.ss
Most particles are found in the lower troposphere, where they can have residence times
ranging from a few hours to weeks. Particles are removed from the atmosphere by wet
deposition, such as when they are carried by rain or snow, or by dry deposition, when particles
settle out of suspension due to gravity. Atmospheric lifetimes are generally longest for PM2.5,
which often remains in the atmosphere for days to weeks before being removed by wet or dry
deposition.158 In contrast, atmospheric lifetimes for UFP and PM10-2.5 are shorter. Within hours,
UFP can undergo coagulation and condensation that lead to formation of larger particles in the
accumulation mode, or can be removed from the atmosphere by evaporation, deposition, or
1111 Children are more susceptible than adults to many air pollutants because of differences in physiology, higher per
body weight breathing rates and consumption, rapid development of the brain and bodily systems, and behaviors
that increase chances for exposure. Even before birth, the developing fetus may be exposed to air pollutants through
the mother that affect development and permanently harm the individual.
Infants and children breathe at much higher rates per body weight than adults, with infants under one year of age
having a breathing rate up to five times that of adults. In addition, children breathe through their mouths more than
adults and their nasal passages are less effective at removing pollutants, which leads to a higher deposition fraction
in their lungs.
ss 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).
173
-------
reactions with other atmospheric components. PM10-2.5 are also generally removed from the
atmosphere within hours, through wet or dry deposition.159
Particulate matter consists of both primary and secondary particles. Primary particles are
emitted directly from sources, such as combustion-related activities (e.g., industrial activities,
motor vehicle operation, biomass burning), while secondary particles are formed through
atmospheric chemical reactions of gaseous precursors (e.g., sulfur oxides (SOx), nitrogen oxides
(NOx) and volatile organic compounds (VOCs)). From 2000 to 2017, national annual average
ambient PM2.5 concentrations have declined by over 40 percent,TT largely reflecting reductions in
emissions of precursor gases.
4.1.2.2 Health Effects Associated with Exposure to Particulate Matter
Scientific evidence spanning animal toxicological, controlled human exposure, and
epidemiologic studies shows that 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 2019.160 The PM ISA
characterizes the causal nature of relationships between PM exposure and broad health categories
(e.g., cardiovascular effects, respiratory effects, etc.) using a weight-of-evidence approach.uu
Within this characterization, the PM ISA summarizes the health effects evidence for short- and
long-term exposures to PM2.5, PM10-2.5, and ultrafine particles, and concludes that human
exposures to ambient PM2.5 are associated with a number of adverse health effects. The
discussion below highlights the PM ISA's conclusions pertaining to the health effects evidence
for both short- and long-term PM exposures. Further discussion of PM-related health effects can
also be found in the 2020 Policy Assessment for the review of the PM NAAQS.161
EPA has concluded that recent evidence in combination with evidence evaluated in the 2009
PM ISA supports a "causal relationship" between both long- and short-term exposures to PM2.5
and mortality and cardiovascular effects and a "likely to be causal relationship" between long-
and short-term PM2.5 exposures and respiratory effects.162 Additionally, recent experimental and
epidemiologic studies provide evidence supporting a "likely to be causal relationship" between
long-term PM2.5 exposure and nervous system effects, and long-term PM2.5 exposure and cancer.
In addition, EPA noted that there was more limited and uncertain evidence for long-term PM2.5
exposure and reproductive and developmental effects (i.e., male/female reproduction and
fertility; pregnancy and birth outcomes), long- and short-term exposures and metabolic effects,
TT See https://www.epa.gov/air-trends/particulate-matter-pm25-trends and https://www.epa.gov/air-
trends/particulate-matter-pm25-trends#pmnat for more information.
1,11 The causal framework draws upon the assessment and integration of evidence from across scientific disciplines,
spanning atmospheric chemistry, exposure, dosimetry and health effects studies (i.e., epidemiologic, controlled
human exposure, and animal toxicological studies), and assess the related uncertainties and limitations that
ultimately influence our understanding of the evidence. This framework employs a five-level hierarchy that
classifies the overall weight-of-evidence with respect to the causal nature of relationships between criteria pollutant
exposures and health and welfare effects using the following categorizations: causal relationship; likely to be causal
relationship; suggestive of, but not sufficient to infer, a causal relationship; inadequate to infer the presence or
absence of a causal relationship; and not likely to be a causal relationship (U.S. EPA. (2019). Integrated Science
Assessment for Particulate Matter (Final Report). U.S. Environmental Protection Agency, Washington, DC,
EPA/600/R-19/188, Section P. 3.2.3).
174
-------
and short-term exposure and nervous system effects resulting in the ISA concluding "suggestive
of, but not sufficient to infer, a causal relationship".
As discussed extensively in the 2019 PM ISA, recent studies continue to support and extend
the evidence base linking short- and long-term PM2.5 exposures and mortality.163 For short-term
PM2.5 exposure, recent multi-city studies, in combination with single- and multi-city studies
evaluated in the 2009 PM ISA, provide evidence of consistent, positive associations across
studies conducted in different geographic locations, populations with different demographic
characteristics, and studies using different exposure assignment techniques. Additionally, the
consistent and coherent evidence across scientific disciplines for cardiovascular morbidity,
particularly ischemic events and heart failure, and to a lesser degree for respiratory morbidity,
with the strongest evidence for exacerbations of chronic obstructive pulmonary disease (COPD)
and asthma, provide biological plausibility for cause-specific mortality and ultimately total
mortality.
In addition to reanalyses and extensions of the American Cancer Society (ACS) and Harvard
Six Cities (HSC) cohorts, multiple new cohort studies conducted in the U.S. and Canada,
consisting of people employed in a specific job (e.g., teacher, nurse) and that apply different
exposure assignment techniques, provide evidence of positive associations between long-term
PM2.5 exposure and mortality. Biological plausibility for mortality due to long-term PM2.5
exposure is provided by the coherence of effects across scientific disciplines for cardiovascular
morbidity, particularly for coronary heart disease (CHD), stroke and atherosclerosis, and for
respiratory morbidity, particularly for the development of COPD. Additionally, recent studies
provide evidence indicating that as long-term PM2.5 concentrations decrease there is an increase
in life expectancy.
A large body of recent studies examining both short- and long-term PM2.5 exposure and
cardiovascular effects supports and extends the evidence base evaluated in the 2009 PM ISA.
Some of the strongest evidence from both experimental and epidemiologic studies examining
short-term PM2.5 exposures are for ischemic heart disease (IHD) and heart failure. The evidence
for cardiovascular effects is coherent across studies of short-term PM2.5 exposure 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 emergency department visits
and hospital admissions due to cardiovascular disease and cardiovascular mortality. For long-
term PM2.5 exposure, there is strong and consistent epidemiologic evidence of a relationship with
cardiovascular mortality. This evidence is supported by epidemiologic and animal toxicological
studies demonstrating a range of cardiovascular effects including coronary heart disease, stroke,
impaired heart function, and subclinical markers (e.g., coronary artery calcification,
atherosclerotic plaque progression), which collectively provide coherence and biological
plausibility.
Recent studies continue to provide evidence of a relationship between both short- and long-
term PM2.5 exposure and respiratory effects. Epidemiologic and animal toxicological studies
examining short-term PM2.5 exposure provide consistent evidence of asthma and COPD
exacerbations, in children and adults, respectively. This evidence is supported by epidemiologic
studies examining asthma and COPD emergency department visits and hospital admissions, as
well as respiratory mortality. However, there is inconsistent evidence of respiratory effects,
specifically lung function declines and pulmonary inflammation, in controlled human exposure
175
-------
studies. Epidemiologic studies conducted in the U.S. and abroad provide evidence of a
relationship between long-term PM2.5 exposure and respiratory effects, including consistent
changes in lung function and lung function growth rate, increased asthma incidence, asthma
prevalence, and wheeze in children; acceleration of lung function decline in adults; and
respiratory mortality. The epidemiologic evidence is supported by animal toxicological studies,
which provide coherence and biological plausibility for a range of effects including impaired
lung development, decrements in lung function growth, and asthma development.
Since the 2009 PM ISA, a growing body of scientific evidence examined the relationship
between long-term PM2.5 exposure and nervous system effects, resulting for the first time in a
causality determination for this health effects category. The strongest evidence for effects on the
nervous system come from epidemiologic studies that consistently report cognitive decrements
and reductions in brain volume in adults. The effects observed in epidemiologic studies are
supported by animal toxicological studies demonstrating effects on the brain of adult animals
including inflammation, morphologic changes, and neurodegeneration of specific regions of the
brain. There is more limited evidence for neurodevelopmental effects in children with some
studies reporting positive associations with autism spectrum disorder (ASD) and others
providing limited evidence of an association with cognitive function. While there is some
evidence from animal toxicological studies indicating effects on the brain (i.e., inflammatory and
morphological changes) to support a biologically plausible pathway, epidemiologic studies of
neurodevelopmental effects are limited due to their lack of control for potential confounding by
copollutants, the small number of studies conducted, and uncertainty regarding critical exposure
windows.
Building off the decades of research demonstrating mutagenicity, DNA damage, and
endpoints related to genotoxicity due to whole PM exposures, recent experimental and
epidemiologic studies focusing specifically on PM2.5 provide evidence of a relationship between
long-term PM2.5 exposure and cancer. Epidemiologic studies examining long-term PM2.5
exposure and lung cancer incidence and mortality provide evidence of generally positive
associations in cohort studies spanning different populations, locations, and exposure assignment
techniques. Additionally, there is evidence of positive associations in analyses limited to never
smokers. The epidemiologic evidence is supported by both experimental and epidemiologic
evidence of genotoxicity, epigenetic effects, carcinogenic potential, and that PM2.5 exhibits
several characteristics of carcinogens, which collectively provides biological plausibility for
cancer development.
For the additional health effects categories evaluated for PM2.5 in the 2019 PM ISA,
experimental and epidemiologic studies provide limited and/or inconsistent evidence of a
relationship with PM2.5 exposure. As a result, the 2019 PM ISA concluded that the evidence is
"suggestive of, but not sufficient to infer a causal relationship" for short-term PM2.5 exposure
and metabolic effects and nervous system effects, and long-term PM2.5 exposures and metabolic
effects as well as reproductive and developmental effects.
In addition to evaluating the health effects attributed to short- and long-term exposure to
PM2.5, the 2019 PM ISA also conducted an extensive evaluation as to whether specific
components or sources of PM2.5 are more strongly related with specific health effects than PM2.5
mass. An evaluation of those studies resulted in the 2019 PM ISA concluding that "many PM2.5
components and sources are associated with many health effects, and the evidence does not
176
-------
indicate that any one source or component is consistently more strongly related to health effects
than PM2.5 mass."164
For both PM10-2.5 and UFPs, for all health effects categories evaluated, the 2019 PM ISA
concluded that the evidence was "suggestive of, but not sufficient to infer, a causal relationship"
or "inadequate to determine the presence or absence of a causal relationship." For PM10-2.5,
although a Federal Reference Method (FRM) was instituted in 2011 to measure PM10-2.5
concentrations nationally, the causality determinations reflect that the same uncertainty identified
in the 2009 PM ISA with respect to the method used to estimate PM10-2.5 concentrations in
epidemiologic studies persists. Specifically, across epidemiologic studies, different approaches
are used to estimate PM10-2.5 concentrations (e.g., direct measurement of PM10-2.5, difference
between PM10 and PM2.5 concentrations), and it remains unclear how well correlated PM10-2.5
concentrations are both spatially and temporally across the different methods used.
For UFPs, the uncertainty in the evidence for the health effect categories evaluated across
experimental and epidemiologic studies reflects the inconsistency in the exposure metric used
(i.e., particle number concentration, surface area concentration, mass concentration) as well as
the size fractions examined. In epidemiologic studies the size fraction can vary depending on the
monitor used and exposure metric, with some studies examining number count over the entire
particle size range, while experimental studies that use a particle concentrator often examine
particles up to 0.3 |im. Additionally, due to the lack of a monitoring network, there is limited
information on the spatial and temporal variability of UFPs within the U.S., as well as population
exposures to UFPs, which adds uncertainty to epidemiologic study results.
The 2019 PM ISA cites extensive evidence indicating that "both the general population as
well as specific populations and lifestages are at risk for PM2.5-related health effects"165 166 For
example, in support of its "causal" and "likely to be causal" determinations, the ISA cites
substantial evidence for (1) PM-related mortality and cardiovascular effects in older adults; (2)
PM-related cardiovascular effects in people with pre-existing cardiovascular disease; (3) PM-
related respiratory effects in people with pre-existing respiratory disease, particularly asthma
exacerbations in children; and (4) PM-related impairments in lung function growth and asthma
development in children. The ISA additionally notes that stratified analyses (i.e., analyses that
directly compare PM-related health effects across groups) provide strong evidence for racial and
ethnic differences in PM2.5 exposures and in the risk of PM2.5-related health effects, specifically
within Hispanic and non-Hispanic Black populations. Additionally, evidence spanning
epidemiologic studies that conducted stratified analyses, experimental studies focusing on animal
models of disease or individuals with pre-existing disease, dosimetry studies, as well as studies
focusing on differential exposure suggest that populations with pre-existing cardiovascular or
respiratory disease, populations that are overweight or obese, populations that have particular
genetic variants, populations that are of low socioeconomic status, and current/former smokers
could be at increased risk for adverse PM2.5-related health effects.
4.1.3 Nitrogen Oxides
4.1.3.1 Background on Nitrogen Oxides
Oxides of nitrogen (NOx) refers to nitric oxide (NO) and nitrogen dioxide (NO2). Most NO2
is formed in the air through the oxidation of nitric oxide (NO) that is emitted when fuel is burned
at a high temperature. NOx is a major contributor to secondary PM2.5 formation, and NOx along
177
-------
with VOCs are the two major precursors of ozone. The health effects of PM and ozone are
discussed in Sections 4.1.1 and 4.1.1 respectively.
4.1.3.2 Health Effects Associated with Exposure to 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).167 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 consists of
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
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 that 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 copollutant 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.
178
-------
4.1.4 Carbon Monoxide
4.1.4.1 Background on Carbon Monoxide
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.
4.1.4.2 Health Effects Associated with Exposure to 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).168 The CO ISA presents
conclusions regarding the presence of causal relationships between CO exposure and categories
of adverse health effects.vv This section provides a summary of the health effects associated
with exposure to ambient concentrations of CO, along with the CO ISA conclusions.ww
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 that 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 that the evidence is
suggestive of a causal relationship between long-term exposures to CO and developmental
effects and birth outcomes.
vv 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.
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.
179
-------
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
copollutants such as ozone, SO2, and PM in two-pollutant models and found that CO risk
estimates were generally robust, although this limited evidence makes it difficult to disentangle
effects attributed to CO itself from those of the larger complex air pollution mixture. Controlled
human exposure studies have not extensively evaluated the effect of CO on respiratory
morbidity. Animal studies at levels of 50-100 ppm CO show preliminary evidence of altered
pulmonary vascular remodeling and oxidative injury. The CO ISA concludes that the evidence
is suggestive of a causal relationship between short-term CO exposure and respiratory morbidity,
and inadequate to conclude that a causal relationship exists between long-term exposure and
respiratory morbidity.
Finally, the 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 copollutant 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.
4.1.5 Diesel Exhaust
4.1.5.1 Background on Diesel Exhaust
Diesel exhaust is 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 of the components present in
diesel exhaust ranges from seconds to days.
4.1.5.2 Health Effects Associated with Exposure to 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
180
-------
exposures, in accordance with the revised draft 1996/1999 EPA cancer guidelines.169'170 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) 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 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 notes 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 and this was retained in 2020 and is being
reconsidered as of June 10, 2021.xx 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.
XX https://www.epa.gov/pm-pollution/national-ambient-air-quality-standards-naaqs-pm
181
-------
Since 2002, several new studies have been published which continue to report increased lung
cancer risk associated 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, including, truck drivers, underground nonmetal miners and other
diesel motor-related occupations. These studies reported increased risk of lung cancer related to
exposure to diesel exhaust, with evidence of positive exposure-response relationships to varying
degrees.171 172 173 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 reinforce 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 (i.e., heavy-duty
highway engines from 2007 and later model years) 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."174 This designation was an update from its 1988 evaluation that considered the
evidence to be indicative of a "probable human carcinogen."
4.1.6 Air Toxics
Heavy-duty engine 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."175 These compounds include, but are not
limited to, benzene, formaldehyde, acetaldehyde, and naphthalene. These compounds were
identified as national or regional risk drivers or contributors in the 2014 National-scale Air
Toxics Assessment and have significant inventory contributions from mobile sources.176177
4.1.6.1 Health Effects Associated with Exposure to Benzene
EPA's Integrated Risk Information System (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.178 179 180 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.^^ 181 The 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.182183
YY A unit risk estimate is defined as the increase in the lifetime risk of an individual who is exposed for a lifetime to
1 ng/m3 benzene in air.
182
-------
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.184'185 The
most sensitive noncancer effect observed in humans, based on current data, is the depression of
the absolute lymphocyte count in blood. 186>187 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, studies sponsored by the Health
Effects Institute (HEI) provide evidence that biochemical responses occur at lower levels of
benzene exposure than previously known.188'189'190'191 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.192,zz
4.1.6.2 Health Effects Associated with Exposure to Formaldehyde
In 1991, EPA concluded that formaldehyde is a Class B1 probable human carcinogen based
on limited evidence in humans and sufficient evidence in animals.193 An Inhalation URE for
cancer and a Reference Dose for oral noncancer effects were developed by EPA and posted on
the IRIS database. Since that time, the NTP and IARC have concluded that formaldehyde is a
known human carcinogen.194'195'196
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.197'198'199 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.200 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.201 Finally, a study of embalmers
reported formaldehyde exposures to be associated with an increased risk of myeloid leukemia
but not brain cancer.202
Health effects of formaldehyde in addition to cancer were reviewed by the Agency for Toxics
Substances and Disease Registry in 1999, supplemented in 2010, and by the World Health
Organization.203-204-205 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.
In June 2010, 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.206 That draft assessment reviewed more recent research from animal and
zz 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.
183
-------
human studies on cancer and other health effects. The NRC released their review report in April
2011.207 EPA's draft assessment, which addresses NRC recommendations, was suspended in
20 1 8.208 The draft assessment was unsuspended in March 2021.
4.1.6.3 Health Effects Associated with Exposure to 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.209
The URE in IRIS for acetaldehyde is 2.2 x 10"6 per |ig/m3.210 Acetaldehyde is reasonably
anticipated to be a human carcinogen by the NTP in the 14th Report on Carcinogens and is
classified as possibly carcinogenic to humans (Group 2B) by the IARC.211'212
The primary noncancer effects of exposure to acetaldehyde vapors include irritation of the
eyes, skin, and respiratory tract.213 In short-term (4 week) rat studies, degeneration of olfactory
epithelium was observed at various concentration levels of acetaldehyde exposure.214'215 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.216 Children, especially those with diagnosed asthma, may be more likely to show
impaired pulmonary function and symptoms of asthma than are adults following exposure to
acetaldehyde.217
4.1.6.4 Health Effects Associated with Exposure to 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.218
Chronic (long term) exposure of workers and rodents to naphthalene has been reported to cause
cataracts and retinal damage.219 EPA released an external review draft of a reassessment of the
inhalation carcinogenicity of naphthalene based on a number of recent animal carcinogenicity
studies.220 The draft reassessment completed external peer review.221 Based on external peer
review comments received, EPA was developing a revised draft assessment that considers all
routes of exposure, as well as cancer and noncancer effects,; this reassessment was suspended in
20 1 8.222 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 NTP 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.223
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.224
Naphthalene also causes a number of chronic non-cancer effects in animals, including
abnormal cell changes and growth in respiratory and nasal tissues.225 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/m3.226 The ATSDR MRL for acute exposure to naphthalene
is 0.6 mg/kg/day.
184
-------
4.1.6.5 Health Effects Associated with Exposure to Other Air Toxics
In addition to the compounds described above, other compounds found in gaseous
hydrocarbon and PM emissions from engines will be affected by this proposed rule. Mobile
source air toxic compounds that would potentially be affected include acrolein, ethylbenzene,
propionaldehyde, toluene, and xylene. Information regarding the health effects of these
compounds can be found in EPA's IRIS database.227
4.1.7 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
employ 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 draft RIA, information on health effects associated with air quality near
major roads or traffic in general is summarized. The information presented in this section relies
on a systematic literature review performed by the Health Effects Institute (HEI) Panel on the
Health Effects of Traffic-Related Air Pollution, Traffic-Related Air Pollution: A Critical Review
of the Literature on Emissions, Exposure, and Health Effects published in 2010.AAA-228 Other
systematic reviews of relevant literature are also cited where appropriate.
It should be noted that HEI has another expert review underway of the literature on health
studies associated with near-road air pollution. The publication of that review is currently
planned for early 2022.
4.1.7.1 Sociodemographic Characteristics of Populations near Major Roads
A number of studies and data sets address the size and composition of the population living in
close proximity of major roads.229 230-231 232 The studies vary in how they describe what
constitutes a major road, and in the distance considered to be near enough to report.
Consistently, the studies demonstrate that there are tens of millions of Americans living near
major roads, key examples follow.
Rowangould (2013) used population data from the 2010 and 2000 decennial U.S. Census
along with traffic data from 2008 to estimate the size and sociodemographic composition of
populations living near roads.233 Considering as "high volume" roads with great than 25,000
annual average daily traffic (AADT), an estimated 59.5 million people lived within 500 meters.
Fourteen counties had more than 50 percent of the population within 500 meters of roads with at
least 25,000 AADT, and counties with the largest share of population living near high volume
roads were mostly in urban areas. Considering a distance of 100 meters from roads with >25,000
AADT, about 3 percent of the population or about 10 million people lived there. Considering the
highest-volume roads (AADT over 250,000), about 0.1 percent of the population, or about 0.4
A A A 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 existing 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.
185
-------
million people, lived within 100 meters of these largest roads, while about 1 percent of the
population lived within 500 meters. According to Rowangould, being a member of a "racial
minority group" (as defined in the study) or living in a location with lower household income
was associated with a higher likelihood of living near a high-volume road. For example, while
19.3 percent of the population lived within 500 meters of roads with >25,000 AADT, 23.7
percent and 29.4 percent of Black and Latino populations (as defined in the study) lived in such
locations. These demographic and economic differences were somewhat attenuated when
controlling for population density.
EPA recently conducted a study to estimate the number of people living near truck freight
routes in the United States.234 Based on a population analysis using the U.S. Department of
Transportation's (USDOT) Freight Analysis Framework 4 (FAF4), an estimated 41 million
people live within 100 meters of these freight routes, 72 million within 200 meters, 148 million
within 500 meters, and 215 million within 1000 meters.BBB In addition, relative to the rest of the
population, people living near FAF4 truck routes are more likely to live in areas with higher
population density (e.g., cities) and with higher percentages of people of color and lower income
residents. Every two years from 1997 to 2009 and in 2011, the U.S. Census Bureau's American
Housing Survey (AHS) conducted a survey that includes data on whether housing units are
within 300 feet of an "airport, railroad, or highway with four or more lanes."ccc>235 The 2013
AHS also included that question, but that was the last AHS that included that question. The
2013 survey reports that 17.3 million housing units or 13 percent of all housing units in the U.S.,
were located in such areas. Assuming that populations are in the same locations as housing
units, this corresponds to a population of more than 41 million U.S. residents in close proximity
to high-traffic roadways or other transportation sources. According to the Central Intelligence
Agency's World Factbook, based on data collected between 2012-2014, the United States had
6,586,610 km of roadways, 293,564 km of railways, and 13,513 airports.236 As such, highways
represent the overwhelming majority of transportation facilities described by this factor in the
AHS. Chapter 6.3.9 of the DRIA also includes a discussion about how human exposure to future
air quality varies with sociodemographic characteristics relevant to potential environmental
justice concerns in scenarios with and without the proposed rule in place.
4.1.7.2 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.
BBB FAF4 is a model from the USDOT's Bureau of Transportation Statistics (BTS) and Federal Highway
Administration (FHWA), which provides data associated with freight movement in the U.S. It includes data from
the 2012 Commodity Flow Survey (CFS), the Census Bureau on international trade, as well as data associated with
construction, agriculture, utilities, warehouses, and other industries. FAF4 estimates the modal choices for moving
goods by trucks, trains, boats, and other types of freight modes. It includes traffic assignments, including truck
flows on a network of truck routes, https://ops.fhwa.dot.gov/freight/freight_analysis/faf/
ccc The variable was known as "ETRANS" in the questions about the neighborhood.
186
-------
4.1.7.3 Cardiovascular Effects
4.1.7.3.1 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.DDD 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 HRV changes from
traffic-related air pollution result in changes to heart rhythms, which can lead to
arrhythmia.237 238
4.1.7.3.2 Heart Attack and Atherosclerosis
The HEI panel report 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 cardiovascular studies of human volunteers exposed to
real-world traffic mixture, which were not entirely consistent. The panel notes that, taken
together, the studies reviewed provide consistent evidence for exposure to PM and impaired
cardiovascular responses. In addition to the HEI panel report, several other reviews of available
evidence conclude that there is evidence supporting a causal association between traffic-related
air pollution and cardiovascular disease, including one from the American Heart Association
(AHA).239
A number of mechanisms for cardiovascular disease are highlighted in the HEI panel report,
including modified blood vessel endothelial function (e.g., the ability to dilate), atherosclerosis,
and oxidative stress. The HEI panel report 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.
4.1.7.4 Respiratory Effects
4.1.7.4.1 Asthma
Pediatric asthma and asthma symptoms are the effects that have been evaluated by the largest
number of studies in the epidemiologic literature on near-roadway exposure. 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
DDD 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.
187
-------
long-term exposure metrics, and a range of different respiratory measures. The HEI panel report
concluded that there is sufficient evidence for a causal association between exposure to traffic-
related air pollution and exacerbation of asthma symptoms in children.
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.240 The study reported
excess main risk estimates for different pollutants ranging from 7-16 percent per 10 |ig/m3 of
long-term exposure (random effects models). Other qualitative reviews (Salam et al., 2008;
Braback and Forsberg, 2009) concluded that available evidence is consistent with the hypothesis
that traffic-associated air pollutants are associated with incident asthma.241242
4.1.7.4.2 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.
4.1.7.4.3 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.243-244-245 However, in its review of 16 epidemiologic studies that address traffic-
related air pollution's effect on allergies, the HEI panel 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 they provide mechanistic support for
observations and inferences derived from epidemiology.
4.1.7.4.4 Luns Function
Lung function is a term describing how well the lungs help a person breath. To detect
conditions associated with reduced lung function, there are numerous measurements of breathing
(spirometry) that indicate the presence or degree of airway diseases 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 FEV1. FEV1 and PEF reflect
188
-------
the function of the large airways. FVC and FEV1, along with their ratio (FVC/FEV1) 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.EEE
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.
4.1.7.5 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 they had limited geographic coverage. One study provided evidence of small but
consistently increased risks using multiple exposure metrics. No studies that were available at
that time 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 concentration levels. Because there was no
overlap in the effects studied by epidemiology and toxicology studies, no synthesis review of the
combined literature was undertaken by the HEI panel.
Since the HEI panel's publication, a systematic review and meta-analysis of air pollution and
congenital abnormalities was published.246 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.
EEE 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 existing 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.
189
-------
4.1.7.6 Cancer
4.1.7.6.1 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.247 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.248-249 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.
4.1.7.6.2 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.
4.1.7.7 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.
4.2 Environmental Effects Associated with Exposure to Pollutants
This section discusses the environmental effects associated with pollutants affected by this
proposed rule, specifically particulate matter, ozone, NOx and air toxics.
4.2.1 Visibility Degradation
Visibility can be defined as the degree to which the atmosphere is transparent to visible
light.250 Visibility impairment is caused by light scattering and absorption by suspended
particles and gases. It is dominated by contributions from suspended particles except under
pristine conditions. Fine particles with significant light-extinction efficiencies include sulfates,
nitrates, organic carbon, elemental carbon, sea salt, and soil.251-252 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 2019 PM ISA.253
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
190
-------
a similar object at a greater distance. See Figure 4-1 for an illustration of the important factors
affecting visibility.234
Optical Characteristics of Illumination
• Detection Thresholds
• Psychological Response to
Incoming Light
• Value Judgements ^PP
• Sunlight ISun Angle)
• Cloud Cover (Overcast, Puffy, etc.)
• Sky
Optical Characteristics of
Intervening Atmosphere
Optical Characteristics of
Viewed Target
• Light Added to Sight Path by
Particles and Cases
• Image-Forming Light Subtracted
from Sight Path by Scattering
and Absorption
• Color
• Contrast Detail (Texture)
• Form
• Brightness
Image-forming
light scattered
out of sight path
Sunlight ^
scattered
Light reflected
from ground
scattered into
sight path
Image-forming
light absorbed
Figure 4-1: Important Factors Involved in Seeing a Scenic Vista (Malm, 2016)
EPA is working to address visibility impairment. Reductions in air pollution from
implementation of various programs called for in the Clean Air Act Amendments of 1990
(CAAA) 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
emissions 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.2''3
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.256 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 that were in existence on August 7, 1977. Figure 4-2 shows the location of
the 156 Mandatory Class I Federal areas.
191
-------
Figure 4-2: 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 NAAQS in 2012, retained it in 2020,
and established a target level of protection that is expected to be met through attainment of the
existing secondary standards for PM2.5.FFF
4.2.1.1 Visibility Monitoring
In conjunction with the U.S. National Park Service, the U.S. Forest Service, other Federal
land managers, and State organizations in the U.S., 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 152 sites that represent all but one of the
156 Mandatory Federal Class I areas across the country (see Figure 4-2). This long-term
visibility monitoring network is known as IMPROVE (Interagency Monitoring of Protected
Visual Environments).
IMPROVE provides direct measurement of 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), and other elements that can be used to estimate soil dust and sea salt contributions.
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
f,|;|; On June 10, 2021. EPA announced that it will reconsider the previous administration's decision to retain the PM
NAAQS. https://www.epa.gov/pm-pollution/national-ambient-air-quality-standards-naaqs-pm.
192
-------
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; and 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
aerosol-related methods. Directly measured light extinction is used under the IMPROVE
protocol to cross check that total light extinction calculated from the IMPROVE reconstruction
formula are consistent with directly measured extinction. Aerosol-derived light extinction from
the IMPROVE equation 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. Figures 13-1 through 13-14 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 month.257
4.2.2 Plant and Ecosystem Effects of Ozone
The welfare effects of ozone include effects on ecosystems, which 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 higher and higher levels of biological organization. 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 plant species depending on the
concentration level and the duration of the exposure.258 In those sensitive speciesGGG, effects
from repeated exposure to ozone throughout the growing season of the plant can tend to
accumulate, so that even relatively low concentrations experienced for a longer duration have the
potential to create chronic stress on vegetation.259 11™ Ozone damage to sensitive plant 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 lead to a reduction in root growth and carbohydrate storage
GGG Only a small percentage of all the plant species growing within the U.S. (over 43,000 species have been
catalogued in the USDA PLANTS database) have been studied with respect to ozone sensitivity.
111111 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.
193
-------
below ground, resulting in other, more subtle plant and ecosystems impacts.260 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
ecosystems111, resulting in a loss or reduction in associated ecosystem goods and services.261
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.262 In addition to ozone effects on vegetation, newer evidence
suggests that ozone affects interactions between plants and insects by altering chemical signals
(e.g., floral scents) that plants use to communicate to other community members, such as
attraction of pollinators.
The Ozone ISA presents more detailed information on how ozone affects vegetation and
ecosystems.263 The Ozone ISA reports causal and likely causal relationships between ozone
exposure and a number of welfare effects and characterizes the weight of evidence for different
effects associated with ozone.JJJ The Ozone ISA concludes that visible foliar injury effects on
vegetation, reduced vegetation growth, reduced plant reproduction, reduced productivity in
terrestrial ecosystems, reduced yield and quality of agricultural crops, alteration of below-ground
biogeochemical cycles, and altered terrestrial community composition are causally associated
with exposure to ozone. It also concludes that increased tree mortality, altered herbivore growth
and reproduction, altered plant-insect signaling, reduced carbon sequestration in terrestrial
ecosystems, and alteration of terrestrial ecosystem water cycling are likely to be causally
associated with exposure to ozone.
4.2.3 Deposition
Deposited airborne pollutants contribute to adverse effects on ecosystems, and to soiling and
materials damage. These welfare effects result mainly from exposure to excess amounts of
specific chemical species, regardless of their source or predominant form (particle, gas or liquid).
Nitrogen and sulfur tend to comprise a large portion of PM in many locations; however, gas-
phase forms of oxidized nitrogen and sulfur also cause adverse ecological effects. The following
characterizations of the nature of these environmental effects are based on information contained
in the 2019 PM ISA, and the 2020 Integrated Science Assessment for Oxides of Nitrogen,
Oxides of Sulfur, and Particulate Matter - Ecological Criteria.264-265 This proposed rule would
reduce emissions of nitrogen and PM but would not change emissions of sulfur.
4.2.3.1 Deposition of Nitrogen and Sulfur
Nitrogen and sulfur interactions in the environment are highly complex, as shown in Figure
4-3.266 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
111 Per footnote above, ozone impacts could be occurring in areas where plant species sensitive to ozone have not yet
been studied or identified.
111 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.
194
-------
ecosystems, excesses of nitrogen or sulfur can lead to acidification and nutrient enrichment.264
In addition, in aquatic ecosystems, sulfur deposition can increase mercury methylation.
Sunlight
Oxidation Dissolution
S02 *¦ H2S04 *> 2H* +S04*"
NO, ~ HNOj p. H*+N03
Dry deposition SO2
NO,, NH„ SO, NO,
Wet Deposition
H*. NH4*. NO3-, S042'
NHj
A
Deposition
N:ONO, . . [
iLikW ii 1
Acidification of water ~ Eutrophication
Ecological
Effect
Figure 4-3: Nitrogen and Sulfur Cycling, and Interactions in the Environment
4.2.3.1.1 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.264 Biological effects of
acidification in terrestrial ecosystems are generally linked to aluminum toxicity and decreased
ability of plant roots to take up base cations.264 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.264
Geology (particularly surficial geology) is the principal factor governing the sensitivity of
terrestrial and aquatic ecosystems to acidification from nitrogen and sulfur deposition.264
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.264
4.2.3.1.1.1 Aquatic Acidification
195
-------
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 acidification at
virtually all levels 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 factor, 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.267-268
4.2.3.1.1.2 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.264 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.269 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.264
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.270
4.2.3.1.2 Ecological Effects from Nitrogen Enrichment
4.2.3.1.2.1 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
(SAY), and low water clarity. Low DO disrupts aquatic habitats, causing stress to fish and
196
-------
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. SAV 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 nitrogen (N) and
phosphorus.271 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.272 Estuaries
in the eastern United States are an important source of food production, in particular for 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.
4.2.3.1.2.2 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 many 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. Nutrient enrichment of the CSS and MCF also affects the
regulating service of fire, by encouraging the growth of more flammable grasses and thus
increasing fuel loads and altering the fire cycle.
4.2.3.1.3 Vegetation Effects Associated with Gaseous Sulfur Dioxide, Nitric Oxide, Nitrogen
Dioxide, Peroxyacetyl Nitrate, and Nitric Acid
Uptake of gaseous pollutants in a plant canopy is a complex process involving adsorption to
surfaces (leaves, stems, and soil) and absorption into leaves. These pollutants penetrate into
197
-------
leaves through the stomata, although there is evidence for limited pathways via the cuticle.264
Pollutants must be transported from the bulk air to the leaf boundary layer in order to reach 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.264
Acute foliar injury from SO2 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.264 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.273
Similarly, in sufficient concentrations, nitric oxide (NO), nitrogen dioxide (NO2),
peroxyacetyl nitrate (PAN), and nitric acid (HNO3) can have phytotoxic effects on plants such as
decreasing photosynthesis and inducing visible foliar injury. It is also known that these gases can
alter the N cycle in some ecosystems, especially in the western U.S., and contribute to N
saturation. Further, there are several lines of evidence that past and current HNO3 concentrations
may be contributing to the decline in lichen species in the Los Angeles basin.274
4.2.3.1.4 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 higher, typically on the order of one
million times, than 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.264 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.264
4.2.3.2 Deposition of Metallic and Organic Constituents of PM
Several significant ecological effects are associated with the deposition of chemical
constituents of ambient PM such as metals and organics.264 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). Direct effect exposures to PM occur via deposition (e.g., wet, dry or
198
-------
occult) to vegetation surfaces, while indirect effects occur via deposition to ecosystem soils or
surface waters where the deposited constituents of PM then interact (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.275 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.264 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.264 Plants growing on roadsides exhibit impact damage from near-road PM
deposition, having higher levels of organics and heavy metals, and accumulating salt from road
de-icing during winter months.264 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.264 Decreases in crop yields (a provisioning service) due to reductions in
solar radiation have been attributed to regional scale air pollution in counties with especially
severe regional haze.276
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.264 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 organic air toxic pollutants depends
on the plant species, site of deposition, physical and chemical properties of the organic
compound and prevailing environmental conditions.264 Different species can have different
uptake rates of PAHs. 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.277'278 Atmospheric deposition of particles is thought to be the major
source of PAHs in the sediments of Lake Michigan, Chesapeake Bay, Tampa Bay and other
coastal areas of the U.S.279
Contamination of plant leaves by heavy metals can lead to elevated concentrations in the soil.
Trace metals absorbed into the plant, frequently by binding to the leaf tissue, and then are shed
when the leaf drops. As the fallen leaves decompose, the heavy metals are transferred into the
soil.280'281 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
199
-------
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.264 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.264
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 runoff264 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.264 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 in receiving waters during spring snowmelt. 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.282 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.
4.2.3.3 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, degrading paints and deteriorating building materials such as stone, concrete
and marble.283 The effects of PM are exacerbated by the presence of acidic gases and can be
additive or synergistic depending on 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).284 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. In addition to aesthetic and functional effects on metals, stone and glass, altered energy
200
-------
efficiency of photovoltaic panels by PM deposition is also becoming an important consideration
for impacts of air pollutants on materials.
4.2.4 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.285 In laboratory experiments, a wide range of tolerance to VOCs has been observed.286
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.287
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.288 289 290 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.
4.3. Climate-Related Effects from GHG Emissions
Elevated concentrations of carbon dioxide (CO2), methane (CH4), and other greenhouse gases
(GHGs) in the atmosphere have been warming the planet, leading to changes in the Earth's
climate including changes in the frequency and intensity of heat waves, precipitation, and
extreme weather events, rising seas, and retreating snow and ice. The well-documented
atmospheric changes due to anthropogenic GHG emissions are changing the climate at a pace
and in a way that threatens human health, society, and the natural environment.
Extensive information on climate change is available in the scientific assessments and EPA
documents that are briefly described in this section, as well as in the technical and scientific
information supporting them. One of those documents is EPA's 2009 Endangerment and Cause
or Contribute Findings for Greenhouse Gases Under section 202(a) of the CAA (74 FR 66496,
December 15, 2009). In the 2009 Endangerment Finding, the Administrator found under section
202(a) of the CAA that elevated atmospheric concentrations of six key well-mixed GHGs - CO2,
CH4, nitrous oxide (N2O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), and sulfur
hexafluoride (SF6) - "may reasonably be anticipated to endanger the public health and welfare of
current and future generations" (74 FR 66523). The 2009 Endangerment Finding, together with
the extensive scientific and technical evidence in the supporting record, documented that climate
change caused by human emissions of GHGs threatens the public health of the U.S. population.
It explained that by raising average temperatures, climate change increases the likelihood of heat
waves, which are associated with increased deaths and illnesses (74 FR 66497). While climate
change also increases the likelihood of reductions in cold-related mortality, evidence indicates
that the increases in heat mortality will be larger than the decreases in cold mortality in the U.S.
(74 FR 66525). The 2009 Endangerment Finding further explained that compared with a future
201
-------
without climate change, climate change is expected to increase tropospheric ozone pollution over
broad areas of the U.S., including in the largest metropolitan areas with the worst tropospheric
ozone problems, and thereby increase the risk of adverse effects on public health (74 FR 66525).
Climate change is also expected to cause more intense hurricanes and more frequent and intense
storms of other types and heavy precipitation, with impacts on other areas of public health, such
as the potential for increased deaths, injuries, infectious and waterborne diseases, and stress-
related disorders (74 FR 66525). Children, the elderly, and the poor are among the most
vulnerable to these climate-related health effects (74 FR 66498).
The 2009 Endangerment Finding also documented, together with the extensive scientific and
technical evidence in the supporting record, that climate change touches nearly every aspect of
public welfare in the U.S. with resulting economic costs, including: changes in water supply and
quality due to changes in drought and extreme rainfall events; increased risk of storm surge and
flooding in coastal areas and land loss due to inundation; increases in peak electricity demand
and risks to electricity infrastructure; and the potential for significant agricultural disruptions and
crop failures (though offset to some extent by carbon fertilization). These impacts are also global
and the effects of climate change occurring outside the U.S. are reasonably expected to impact
the U.S. population. (74 FR 66530).
In 2016, the Administrator issued a similar finding for GHG emissions from aircraft under
section 231(a)(2)(A) of the CAA. In the 2016 Endangerment Finding, the Administrator found
that the body of scientific evidence amassed in the record for the 2009 Endangerment Finding
compellingly supported a similar endangerment finding under CAA section 231(a)(2)(A), and
also found that the science assessments released between the 2009 and the 2016 Findings
"strengthen and further support the judgment that GHGs in the atmosphere may reasonably be
anticipated to endanger the public health and welfare of current and future generations" (81 FR
54424).
Since the 2016 Endangerment Finding, the climate change impacts have continued to intensify,
with new observational records being set for several climate indicators such as global average
surface temperatures, GHG concentrations, and sea level rise. Moreover, heavy precipitation
events have increased in the eastern United States while agricultural and ecological drought has
increased in the western United States along with more intense and larger wildfires.KKK Recent
assessment reports discuss how these observed trends are increasingly attributed to human-induced
climate changeLLL and are expected to continue and worsen over the coming century, with stronger
trends under higher warming scenarios.291'292'293 Climate impacts that occur outside U.S. borders
also increasingly impact the welfare of individuals and firms that reside in the United States
because of their connection to the global economy. This will occur through the effect of climate
change on international markets, trade, tourism, and other activities. For example, supply chain
disruptions are a prominent pathway through which U.S. business and consumers are, and will
KKK See EPA's November 2021 Proposed Standards of Performance for New, Reconstructed, and Modified Sources
and Emissions Guidelines for Existing Sources: Oil and Natural Gas Sector Climate Review
{https://www.govinfo.gov/content/pkg/FR-2021-l 1-15/pdf/202l-24202.pdf) for more discussion of specific
examples. An additional resource for indicators can be found at https://www.epa.gov/climate-indicators.
LLL For example, "[f]ield evidence shows that anthropogenic climate change has increased the area burned by
wildfire above natural levels in western North America from 1984-2017 by double for the Western USA... (high
confidence)" (AR6 WGII, p. 2-5).
202
-------
continue to be, affected by climate change impacts abroad.291'294 Additional climate change
induced international spillovers can occur through pathways such as damages across
transboundary resources, economic and political destabilization, and global migration that can
lead to adverse impacts on U.S. national security, public health, and humanitarian concerns.295'296
These and other trends highlight the increased risk already being experienced due to climate
change as detailed in the 2009 and 2016 Endangerment Findings. Additionally, new major
scientific assessments continue to advance our understanding of the climate system and the
impacts that GHGs have on public health and welfare both for current and future generations.
These assessments include:
• U.S. Global Change Research Program's (USGCRP) 2016 Climate and Health
Assessment and 2017-2018 Fourth National Climate Assessment (NCA4).297'298'291
• IPCC's 2018 Global Warming of 1.5 °C, 2019 Climate Change and Land, and the 2019
Ocean and Cryosphere in a Changing Climate assessments, as well as the 2021 IPCC
Sixth Assessment Report (AR6).299'300'301'302
• The National Academies of Sciences, Engineering, and Medicine's 2016 Attribution of
Extreme Weather Events in the Context of Climate Change, 2017 Valuing Climate
Damages: Updating Estimation of the Social Cost of Carbon Dioxide, and 2019 Climate
Change and Ecosystems assessments.303'304'305
• National Oceanic and Atmospheric Administration's (NOAA) annual State of the
Climate reports published by the Bulletin of the American Meteorological Society, most
recently in August of 2020.306
• EPA Climate Change and Social Vulnerability in the United States: A Focus on Six
Impacts (2021).307
203
-------
Chapter 5 Emissions Inventory
5.1 Introduction
This chapter presents our analysis of the national emissions impacts of the proposed and
alternative standards for calendar years 2027 through 2045. In addition to describing the national
emission inventories, this chapter describes the methods used to estimate the spatially and
temporally-resolved emission inventories in 2016 and 2045 that supported the air quality
modeling analysis documented in Chapter 6.
The onroad national inventories were estimated using an updated version of EPA's Motor
Vehicle Emission Simulator (MOVES) model, as described in detail in Chapter 5.2. The onroad
national emission inventories were developed using a single national modeling domain, referred
to as national-scale in MOVES. Using MOVES national-scale mode facilitates the estimation of
national emission inventories for multiple scenarios and multiple calendar years within our
available resource constraints. Inputs developed to model the national emission inventories for
the proposed and alternative scenarios are discussed in Chapter 5.2.2. The national emissions
inventory impacts on criteria and toxic air pollutants for calendar years 2030, 2040, and 2045 for
proposed Option 1 and 2 and an alternative are presented in Chapter 5.3. In addition, the national
emissions results for calendar years 2027 through 2045 are presented in Chapter 5.5.5 for criteria
pollutants.
As described in Chapter 5.4, MOVES is also used to estimate the emission inventories for air
quality modeling; however, the process is resource-intensive, requiring many sets of MOVES
inputs and many MOVES runs to account for spatial and temporal differences in local vehicle
activity and meteorology data. For this reason, we only used this method to estimate emissions
for one scenario (proposed Option 1) in the two calendar years necessary for air quality analyses
(2016 and 2045). The control scenario analyzed for air quality analysis differed from the
proposed Option 1. Section 5.4 also compares the onroad emission inventories estimated from
the national-scale runs to the emissions inventories estimated for the air quality analysis
aggregated to the national-scale for calendar year 2045, in order to understand the impact of
using different methods and control scenarios to estimate emissions inventories.
5.2 Model and Data Updates
To quantify the impacts of the proposed Options 1 and 2 and alternative on emissions, EPA
developed an updated version of MOVES, referred to as "MOVES CTINPRM", that includes
significant updates to heavy-duty vehicle emission rates and activity from MOVES2014b, the
public version of the model at the time of the analysis (see Table 5-1). Detailed descriptions of
the underlying data and analyses that informed the model updates are documented in peer-
reviewed technical reports referenced in Table 5-1. In addition, the MOVES CTI NPRM version
used to generate the emissions inventories can be found in the docket.308
204
-------
Table 5-1: Updates to MOVES CTINPRM from MOVES2014b
MOVES updates
Description
Heavy-duty Greenhouse
Gas Phase 2
Rulemaking3"931"
Data updates and algorithm modifications to model the emission control technologies
from the HD GHG Phase 2 Rulemaking. These include improvements in engine and
vehicle efficiency, reductions in tire rolling resistance and aerodynamic drag,
increased use of idle reduction technologies, and updated main engine and auxiliary
power unit emission rates
Off-network idle3"9
Update MOVES algorithms to incorporate off-network idling, which includes truck
idle activity off the road network, such as at the curbside, parking lots, distribution
centers, truck depots, ports, and construction sites. Incorporate idle data from
instrumented truck studies
Heavy-duty start activity3"9
Update MOVES algorithms and data to incorporate start data collected from
instrumented truck studies
Hotelling activity3"9
Update MOVES algorithms for modeling hotelling activity (rest periods for long-haul
combination trucks)
Heavy-duty vehicle
weights3"9
Update heavy-duty vehicle weights by vehicle vocation and weight class
Heavy-duty running
exhaust emission rates31"
Update heavy-duty diesel running exhaust emission rates for 2010 and later model
year vehicles using data from the Heavy-Duty In-Use Testing (HDIUT) Program
Heavy-duty start emission
rates31"
Update heavy-duty start emission rates using engine certification data submitted by
engine manufacturers
Glider trucks3"931"
Add glider trucks to MOVES. Incorporate emission rates based on NVFEL test
program of glider trucks
Heavy-duty compressed
natural gas (CNG) vehicles
and emissions3"931"
Update heavy-duty CNG exhaust emission rates based on HDIUT program, and
update MOVES to include CNG fuel use in additional heavy-duty vehicle types
Heavy-duty gasoline
vehicles31"311
Update heavy-duty gasoline tailpipe emission rates based on test program conducted
at NVFEL, and update data regarding onboard refueling vapor recovery systems used
in heavy-duty gasoline vehicles
Heavy-duty population and
activity information3"9
Update information on heavy-duty vehicle populations based on vehicle registration
data, FHWA vehicle miles traveled data, and updated projections from the Annual
Energy Outlook 2018312
Speciation Updates311313
Update speciation profiles (used to estimate total organic gases, volatile organic
compounds, and individual organic compounds, such as methane and toxics) for
model year 2010 and later heavy-duty diesel vehicles
Light-duty vehicle and
other changes3"9311
Other changes to MOVES, including updated gasoline fuel properties, light-duty
vehicle emissions rates, light-duty activity, and other changes outlined in technical
memo to the docket
5.2.1 Methodology Overview
We used MOVES CTI NPRM to estimate the emissions impacts of the proposed and
alternative control scenarios. First, we estimated emissions for a baseline scenario in which there
are no new heavy-duty engine emission standards. We then estimated emissions for each of the
three control scenarios: the proposed Option 1, proposed Option 2, and the Alternative. For
description of the proposed options and alternative program aspects, see Chapter 1. The
emissions impacts of the proposed Options 1 and 2 and the Alternative control scenarios were
estimated by calculating the difference between the emissions estimated in the baseline and each
of the control scenarios. All of the model inputs, 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.308
205
-------
For the baseline scenario, we used the heavy-duty exhaust emission rates in the default
database for the MOVES CTINPRM, which are described in detail in the MOVES CTINPRM
technical report.310 Note that because the national emissions inventories used a single national
modeling domain, the baseline scenario did not account for different emission standards in
California or other states that have adopted the California emission standards, for either heavy-
duty or light-duty vehicles. At the time of the analysis for this proposed rule, California had not
finalized its proposed HD lowNOx Omnibus rule314; thus, we modeled only the existing federal
heavy-duty standards for all US States. For light-duty vehicles, MOVES can account for States
that have adopted California's Low Emission Vehicle (LEV) program under Section 177 of the
Clean Air Act. However, the single modeling domain used in national-scale modeling does not
allow us to include the LEV program in California and Section 177 States emission estimates.
Expanding the number of MOVES runs to include the state-by-state LEV programs would
greatly increase the number of MOVES runs and was deemed outside the scope of this
analysisMMM because the purpose of this analysis is to estimate the emission reductions from the
heavy-duty sector due to the proposed rule. Note that the LEV program in California and Section
177 States are included as inputs in the county-scale MOVES runs conducted to develop the
emissions inventory used for air quality analyses documented in Chapter 5.4.
The vehicle activity and fuel inputs were kept the same for baseline and control scenarios. For
vehicle activity, we used the default inputs in MOVES CTI NPRM (e.g., fleet age distributions,
vehicle miles traveled by vehicle type and road type, vehicle speeds, off-network idling,
hotelling hours, and start activity). The default activity data are described in detail in a recent
MOVES model technical report.309 For example, as shown in Table 5-1 above, future year
projections of vehicle populations and vehicle miles travelled were updated to reflect the
estimates from the Department of Energy's Annual Energy Outlook 2018.312 Note that the
vehicle activity data for the emissions inventory used for the air quality analyses included local
activity data as documented in Chapter 5.4.
The fuel inputs used in the MOVES CTI NPRM emission inventory runs were updated from
the default fuel database in MOVES2014b using revised assumptions and more recently
available compliance data. These changes include setting the diesel fuel supply nationwide to B5
biodiesel, setting the gasoline fuel supply for calendar years 2012 and later to E10 nationwide
(no E0 or El5 gasoline), updating gasoline fuel properties on latest available data from EPA's
fuel compliance system and updating the characterizations of local fuel programs. We used the
same fuel usage and fuel property data for all the MOVES runs for calendar years from 2027
through 2060.311 Additional details on updates to the fuels in MOVES CTI NPRM runs are
documented in the MOVES CTI NPRM Technical Report.311
The same fuel inputs are used in both the national emission inventory runs and the emission
inventory runs used for air quality modeling. The national emissions inventory runs aggregate
the gasoline fuel properties into a single region representing the United States, whereas the
county-level MOVES runs for the air quality modeling analysis account for county and regional
differences in gasoline fuel properties. Diesel and CNG fuel properties are represented with one
set of fuel properties to represent the entire nation for each calendar year and fuel type. As such,
141414 The state-specific LEV programs are included in the emission inventories used for the air quality modeling
analysis of the proposed rule as discussed in Section 5.4.
206
-------
no precision on the diesel and CNG fuel effects are averaged when running MOVES in national-
scale.
The emission rate inputs developed for modeling the control scenarios (proposed Option 1
and 2 and the Alternative) are discussed in detail in the following section.
5.2.2 MOVES Emission Rates for Control Scenarios
We developed separate MOVES emission rates for the proposed Options 1 and 2 and the
Alternative (collectively referred to as the control scenarios). The methodologies used to develop
the emission rates reflect the effects of this rulemaking's program elements (duty-cycle
standards, off-cycle standards, closed crankcase requirements, refueling standards, regulatory
useful life, and emissions warranty) on vehicles subject to the proposed rule as described below.
We did not estimate the emission impacts of our proposed compliance provisions that target
long-term compliance assurance. As we describe in Section IV of the preamble to this rule, we
expect improved serviceability and updates to our current inducement policy will discourage
owners from tampering their engines or emission control systems; however, we had insufficient
data to estimate the impact of these proposals.
MOVES has separate emission rates for heavy-duty vehicles according to three different fuel
types—diesel, gasoline, and natural gas (NG) —and six different regulatory vehicle classes,
shown in Table 5-2. The proposed rule includes new duty-cycle standards for all heavy-duty
vehicles in the LHD45, MHD, HHD, and Urban Bus regulatory classes for all fuel types. The
Urban Bus regulatory class was modeled using the same zero-mile emission rates as HHD for the
control scenarios.NNN
Light heavy-duty Class 2b and 3 trucks (LHD2b3) are composed of vehicles that are both
chassis- and engine-certified. The proposed standards would apply to all engine-certified
LHD2b3 vehicles, which are estimated to be a small fraction of the diesel LHD2b3 vehicles and
are discussed in Chapter 5.2.2.6. All Class 2b and 3 gasoline-fueled vehicles are chassis-certified
and would not be affected by the proposed rule.
A glider vehicle is a new motor vehicle produced with a used or remanufactured engine, and
typically does not include the aftertreatment systems needed to meet the 2007 or 2010 heavy-
duty emission standards.000 In MOVES, glider vehicles are modeled as heavy heavy-duty diesel
vehicles (Class 8 Trucks) that are assumed to emit at a level equivalent to the model year 2000
HHD vehicles310. Annual glider sales are fixed at 1,500 units per year for 2018 and later based
on the 2018 and later sales volume cap for glider vehicles set forth in the Heavy-duty
Greenhouse Gas Phase 2 Rulemaking315, as well as the number of glider manufacturers and their
historic production levels, as documented in the population and activity report.309 For the
proposed rule analysis, we assumed there is no change to the sales of glider vehicles and no
change to glider vehicle emission rates from the baseline scenario.
NNN Urban Bus vehicles are also considered HHD vehicles. They are modeled as a separate regulatory class in
MOVES because they had stricter PM emission standards for 1994 through 2006 model years. For the control
scenarios, the Urban Bus emission rates were assumed to be equivalent to HHD except for the age effects discussed
in Section 5.2.2.1.2.
uuu See the definition of "glider vehicle" in 40 CFR 1037.801 and our discussion in the preamble of the heavy-duty
GHG Phase 2 rulemaking (81 FR 73941, October 25, 2016).
207
-------
Table 5-2: MOVES Heavy-duty Regulatory Classes and Relevant MOVES Fuel Types
MOVES Regulatory Class Description
regClassName
regClassID
Gross Vehicle
Weight Rating
(GVWR) [lb.]
MOVES
Fuel Types
Light Heavy-Duty Class 2b and 3 trucks
LHD2b3
41
8,501 - 14,000
Gasoline,
Diesel
Light Heavy-Duty Class 4 and 5 Trucks
LHD45
42
14,001 - 19,500
Gasoline,
Diesel
Medium Heavy-Duty (Class 6 and 7 Trucks)
MHD
46
19,501 -33,000
Gasoline,
Diesel
Heavy Heavy-Duty (Class 8 Trucks)
HHD
47
> 33,000
Diesel, NG
Urban Bus
Urban Bus
48
> 33,000
Diesel, NG
Gliders (Class 8 Trucks)
Glider Vehicles
49
> 33,000
Diesel
For heavy-duty diesel vehicles, we developed nitrogen oxide (NOx) emission rates for
running, start, and extended idle processes, as discussed in Chapters 5.2.2.1, 5.2.2.2, and 5.2.2.3,
respectively, in response to the proposed duty-cycle and off-cycle standards and the alternatives.
The lower control case NOx emission rates are anticipated to be achieved using the technologies
discussed in Chapter 3.1.3, including cylinder deactivation (CDA) and dual selective catalytic
reduction exhaust aftertreatment system. We also revised heavy-duty diesel emission running
rates for NOx, total hydrocarbons (THC), carbon monoxide (CO), and particulate matter below
2.5 microns (PM2.5), due to the proposed changes to the regulatory useful life and warranty, by
modifying MOVES' tampering and mal-maintenance (T&M) calculations, as discussed in
Chapter 5.2.2.1.2. We also eliminated the PM2.5 crankcase emissions from heavy-duty diesel
vehicles due to the proposed closed crankcase design requirement for heavy-duty diesel engines
as discussed in Section 5.2.2.4.
For heavy-duty gasoline vehicles, we developed revised NOx, THC, CO, and PM2.5 emission
rates for running exhaust in response to the proposed duty-cycle standards, as discussed in
Chapter 5.2.2.6. We also developed revised THC refueling emission rates in response to the
proposed refueling standard discussed in Chapter 5.2.2.7. MOVES does not model extended idle
emissions for heavy-duty gasoline vehicles. We also did not revise the start emission rates for
heavy-duty gasoline vehicles even though we expect reductions in the start emissions due to the
proposed standards. We plan to revisit these assumptions in the final rulemaking analysis.
For heavy-duty NG vehicles, we did not estimate reductions in the NOx or other emission
rates due to the proposed heavy-duty spark-ignition duty-cycle standards (discussed in Section
III of the Preamble). As shown in the MOVES heavy-duty exhaust reportFPP, the average FTP
emissions for MY 2010-2017 CNG engine families is close to 0.1 g/hp-hr. We expect reductions
in NOx emissions from heavy-duty NG engines with the proposed and alternative scenarios,
however, in part due to the small contribution of NG emissions to total emissions, we did not
estimate the reductions in NG engines in this analysis. We can revisit these assumptions in the
final rulemaking analysis.
ppp See Table 4-2 in Exhaust Emission Rates for Heavy-Duty Onroad Vehicles in MOVESCTINPRM310
208
-------
Similarly, we did not estimate emission reductions from the proposed and alternative useful
life and warranty requirements for gasoline and NG vehicles. This is because, in MOVES, we do
not tie emission estimates to heavy-duty gasoline or NG vehicle warranty and useful life periods,
but rather estimate age effects from emissions data, or adapt light-duty gasoline or heavy-duty
diesel age effects. Gasoline and NG-fueled vehicles' contributions to the heavy-duty vehicle NOx
emissions inventory are small (see Figure 5-16), and thus we expect minimal impacts on our
emission inventory estimates from our conservative approach of estimating reductions solely
from running emissions for these vehicles.QQQ Depending on the availability of emissions data
from gasoline and NG-fueled vehicles that are operated past their regulatory useful life, as well
as data on heavy-duty gasoline start emissions, we will consider accounting for emission
reductions from the lengthened warranty and useful life requirements, as well as reductions from
the start emissions from these vehicles, in the final rulemaking analysis.
While we estimated different emission rates for the control scenarios, we did not estimate
differences in the speciation of the total organic gases and particulate matter emissions between
the baseline and control scenarios. Emissions speciation refers to the calculation of individual
compounds or classes of compounds within a broader pollutant. For example, MOVES conducts
speciation to estimate benzene from volatile organic compound (VOC) emissions and particle-
phase naphthalene from particulate matter (PM) emissions. Speciation is important to estimate
individual toxics and necessary for air quality modeling. We do not have data to support
differences in the emissions speciation between the control case and baseline scenarios. Heavy-
duty diesel vehicles in the control case scenarios apply the same speciation values as the baseline
scenario for heavy-duty diesel vehicles model year 2010 and later. Similarly, emissions from
heavy-duty gasoline vehicles in the control scenarios use the same speciation as heavy-duty
gasoline vehicles in the baseline scenario. Changes in emissions for individual compounds
presented in Section 5.3 (acetaldehyde, benzene, formaldehyde, methane, and naphthalene) are
due to changes in the hydrocarbon and particulate matter emission rates for the control scenarios
presented in this section.
5.2.2.1 Heavy-Duty Diesel Running Emission Rates
This section documents the approach used to develop the heavy-duty diesel emission running
rates for the control scenarios. We first estimated new zero-mile NOx emission rates in response
to the proposed Option 1, Option 2 and the Alternative duty-cycle and off-cycle standards as
discussed in Chapter 5.2.2.1.1. The zero-mile emission rate in MOVES is defined as the
emission rate of a new vehicle without any tampering and mal-maintenance effects. We then
estimated the effects of lengthened regulatory useful life and warranty periods and applied them
to the aged NOx, THC, CO and PM2.5 emission rates for the proposed and alternative standards
in Chapter 5.2.2.1.2.
Emission rates in MOVES are defined by the operating mode, regulatory class, model year,
and fuel type. The running operating modes are defined in Table 5-3.
QQQ Based on the MOVES national run, the percentages of the NG vehicle contributions to HD NOx inventory in
calendar year 2045 are: 0.59% (baseline), 1.43% (proposed Option 1), 1.24% (proposed Option 2), and 1.53%
(Alternative). These contributions are dependent on the current projections of NG in the future heavy-duty fleet and
the emission rates of NG vehicles, many of which are already meeting the 0.02 g/hp-hr NOx standard.
209
-------
Table 5-3: MOVES Running Operating Mode Definitions
OpModelD
Operating Mode
Vehicle Speed
(v, mph)
Scaled Tractive
Power (STP, skW)
0
Deceleration/Braking
All speeds
1
Idle
v< 1.0
11
Coast
1 < v < 25
STP< 0
12
Cruise/Acceleration
0 < STP<3
13
Cruise/Acceleration
3 < STP<6
14
Cruise/Acceleration
6 < STP<9
15
Cruise/Acceleration
9 < STP< 12
16
Cruise/Acceleration
12 < STP
21
Coast
25 < v < 50
STP< 0
22
Cruise/Acceleration
0 < STP<3
23
Cruise/Acceleration
3 < STP<6
24
Cruise/Acceleration
6 < STP<9
25
Cruise/Acceleration
9 < STP< 12
27
Cruise/Acceleration
12 < STP< 18
28
Cruise/Acceleration
18 < STP< 24
29
Cruise/Acceleration
24 < STP< 30
30
Cruise/Acceleration
30 < STP
33
Cruise/Acceleration
50
-------
Test - Ramped Modal Cycle (SET-RMC). The proposed Option 1 standards have two steps, with
initial standards starting in Model Year (MY) 2027 vehicles and more stringent standards
starting with MY 2030 vehicles. The proposed Option 2 and the Alternative do not include any
phase-in and start in MY 2027. For the proposed Option 1, HHD vehicles have two separate FTP
duty-cycle test standards for MY 2031 and later vehicles; one standard that applies through an
intermediate useful life of 435,000 miles and another standard through the extended full useful
life of 800,000 miles.
Table 5-4: Heavy-duty Compression Ignition Duty-Cycle Cycle NOx Standards for the Proposed Options and
Alternative Scenarios
Scenario
Applicable Model Years
Regulatory
Classes
LLC (g/hp-hr)
FTP
(g/hp-hr)
SET-RMC
(g/hp-hr)
Baseline
Model Year 2010+
LHD, MHD,
HHD
-
0.2
0.2
Proposed
Option 1
Model Year 2027-2030
LHD, MHD,
HHD
0.09
0.035
0.035
Model Year 2031+
LHD, MHD
0.05
0.02
0.02
HHDA
(435,000/800,000)
0.05/0.1
0.02/0.04
0.02/0.04
Proposed
Option 2
Model Year 2027+
LHD, MHD,
HHD
0.1
0.05
0.05
Alternative
Model Year 2027+
LHD, MHD,
HHD
0.1
0.02
0.02
A In the proposed Option 1, HHD engines have a standard at the intermediate useful life of 435,000 miles, and at the
extended useful life of 800,000 miles. LHD and MHD only have standards at the extended useful life.
The proposed and alternative scenarios also include revised standards for PM emissions as
discussed in Section III of the draft preamble. We did not model the revised PM standards for
heavy-duty diesel vehicles in MOVES because the revised PM standards are intended to prevent
backsliding of PM reductions already achieved with current heavy-duty diesel emission control
systems. As discussed in Chapter 5.2.2.1.2, we estimate reductions in heavy-duty diesel PM
emissions would occur due to the proposed lengthened warranty and useful life periods and due
to the proposed crankcase emissions control, as discussed in Chapter 5.2.2.4.
To develop running exhaust emission rates for the HHD MY 2031 and later vehicles in the
proposed Option 1, we calculated the average emission rate from duty-cycle test standards during
the useful life of the vehicles by weighting each standard by the miles they apply to during the
useful life period using Equation 5-1.
211
-------
Equation 5-1
HHD Weighted Average Engine Certification Standards
( 435K \ ,
= -—- ——— x (Standard at 435K)
VExtended Useful Life/
/ 435K \
+ 1 — ———r , T r x (Standard at Extended Useful Life)
\ Extended Useful Life/
For the FTP and SET-RMC standard, this equation yields:
HHD MY 2031 Weighted Average Engine Certification Standards
/435K\ ( g \ ( 435K\ ( g \
~ (book) X 0-02 [hp ¦ hr) + l* ~ 800k) X °'°4 [hp ¦ hr)
= 0029fe)
The weighted average LLC, FTP and SET-RMC emission standards for HHD MY 2031 and
later model years in the proposed Option 1 are shown in Table 5-5.
Table 5-5: Proposed Option 1 Weighted Average Heavy Heavy-duty Compression Ignition Duty-Cycle Test
NOx Standards
Applicable Model Years
LLC
(g/hp-hr)
FTP
(g/hp-hr)
SET-RMC
(g/hp-hr)
2031+
0.073
0.029
0.029
We used the proposed and alternative standards for the FTP and (SET-RMC) to estimate the
effect of the duty-cycle standards on MOVES NOx emission rates. Because we do not have
sufficient (LLC) test data on existing heavy-duty diesel vehicles to develop the modeling inputs
specific for the proposed LLC standard in MOVES, we used the FTP standard to model the
impact of the standards on low-power operation.
Equation 5-2 through Equation 5-7 were used to incorporate the effects of more stringent FTP
and SET-RMC engine duty-cycle emission standards on MOVES running exhaust NOx emission
rates for model years subject to the proposed and alternative standards. The term Rduty is the
ratio between the proposed or alternative emission standard and the current FTP and SET-RMC
duty-cycle standards (0.2 g/hp-hr).
Equation 5-2
Proposed or Alternative FTP or SET RMC standard
duty Current standard
Rduty ranges between 10 and 25 percent for the control scenarios considered, as shown in
Table 5-6.
212
-------
Table 5-6: Rduty Ratios Calculated for Each Scenario
Scenario
Applicable Model
Years
Regulatory Classes
FTP and SET
standard
(g/hp-hr)
Rduty
Proposed
Option 1
2027-2030
LHD, MHD, HHD
0.035
18%
2031+
LHD, MHD
0.02
10%
HHD
0.029
15%
Proposed
Option 2
2027+
LHD, MHD, HHD
0.05
25%
Alternative
2027+
LHD, MHD, HHD
0.02
10%
To estimate the effect of proposed and alternative engine dynamometer duty-cycle standards
on in-use emissions, we used the relationship between reductions in the most recent duty-cycle
standard compared to reductions in in-use emissions. The 2010 0.2 g/hp-hr NOx emission
standard316 is the most recent heavy-duty NOx emission standard. To evaluate the in-use
effectiveness of the 2010 standard, we compared the in-use NOx emission rates from vehicles
that were certified to the previous heavy-duty NOx standard and the 2010 standard. Equation 5-3
defines Rin uSe as the ratio between the percent change observed in-use from vehicles compliant
with the 2010 NOx standard relative to vehicles compliant with the previous standard, and the
percent change in the 2010 standard FTP standards relative to the previous standard.RRR In other
words, Rin use is the ratio between the relative effectiveness of reducing in-use emissions
compared to the relative reduction in the FTP duty-cycle emission standard.
Equation 5-3
% Change in the inuse emission rates from 2010 compliant vehicles
^in-use % Change in the 2010 FTP standard
The percent change in in-use emission rates from vehicles certified to the 2010 standard (the
numerator in Equation 5-3) was estimated using Equation 5-4. The MOVES emission rates for
HHD vehicles certified to the 2010 0.2 g/hp-hr standard (the numerator) are calculated from 93
MY 2010-2015 HHD vehicles with engine family emission limit (FEL) certified below the 0.2
g/hp-hr NOx emissions level and tested as part of the Heavy-Duty In-Use Testing program.310
The MOVES emission rates for HHD vehicles certified to the 2004-2006 standard (the
denominator) are based on 91 MY 2003-2006 trucks from two in-use datasets: ROVER data
collected by the US EPA and the Heavy-Duty Diesel Consent Decree data collected by West
Virginia University. The in-use datasets and analysis used to derive the emission rates in
MOVES CTINPRM are documented in the MOVES CTINPRM heavy-duty exhaust technical
report.
111111 In 2004-2006, the NMHC+NOx emission standard was 2.4 g/hp-hr; the 0.2 g/hp-hr NOx standard began to be
phased-in starting in 2007, with a full-phase in 2010.
213
-------
Equation 5-4
% Change in the in_use emission rates from 2010 compliant vehicles
Emission rate from HHD 2010 compliant vehicles
HHD MY 2006 MOVES emission rate
The percent change from Equation 5-4 was applied to each MOVES operating mode to
evaluate the effectiveness of the 2010 standard across different ranges of in-use operating
conditions, as shown in Figure 5-1 and Table 5-7. The emission reduction is larger for the higher
speed and higher load MOVES operating modes, with the largest decrease observed for speeds
above 50 mph (operating modes 33 through 35). The lowest effectiveness of the standards is
observed for low speed and several low power operating modes (operating mode 1, 11, and 21),
with an exception of the deceleration bin (operating mode 0).
Equation 5-5 was used to estimate the percent reduction between the 2010 standard and the
2004-2006 emission standard. This term is also the denominator of Equation 5-3. The percent
reduction in the NOx emission standard was estimated assuming that NOx emissions consist of
70 percent of the combined NMHC + NOx 2004-2006 emission standard (2.4 g/hp-hr),
consistent with the assumption used in MOVES2014b.317 This value is also plotted as a line in
Figure 5-1 to compare to the in-use emission rate reductions.
Equation 5-5
0.2
% Change in the 2010 FTP standard = (-ygo/ x 2 4) ~ ^ = ~
214
-------
0>
4-»
03
tlQ
X
O
Cl)
to
c
CD
_c
u
no
ON
Speed<25
¦ g i ¦ r
> i ¦ ¦
% Reductions in the FTP 0.2 bhp-hr standard
% In-use emission reductions from 0.2 bhp-
hr certified vehicles
2550
215
-------
Table 5-7: Calculation of Rm_uSe by MOVES Operation Mode
MOVES
OpMode
HHD MOVES
MY 2006 NOx
emission rates
(g/hr)
NOx Emission rate
from 2010 compliant
HHD vehicles (g/hr)a
Percent change in in-use
NOx emission rates from
2010 compliant vehicles
(%)
Rin use
0
0.038
0.0031
-91.8
1.04
1
0.015
0.0063
-57.9
0.66
11
0.015
0.0106
-28.8
0.33
12
0.058
0.0210
-63.5
0.72
13
0.093
0.0339
-63.7
0.72
14
0.127
0.0453
-64.4
0.73
15
0.145
0.0565
-60.9
0.69
16
0.188
0.0734
-60.9
0.69
21
0.010
0.0066
-32.1
0.36
22
0.064
0.0163
-74.4
0.84
23
0.093
0.0243
-73.9
0.84
24
0.132
0.0359
-72.7
0.83
25
0.165
0.0485
-70.6
0.80
27
0.225
0.0633
-71.8
0.82
28
0.244
0.0558
-77.2
0.88
29
0.314
0.88b
30
0.384
0.88
33
0.051
0.0062
-87.8
1.00
35
0.148
0.0130
-91.2
1.04
37
0.226
0.0323
-85.7
0.97
38
0.268
0.0306
-88.6
1.01
39
0.345
1.01
40
0.422
1.01
Notes:
a The HHD rates in this table are based on fscale of 17.1 metric tons to be consistent with
the fscale of the HHD MOVES CTINPRM MY 2006 emission rates. Note that the fscale
for model year 2010 and later in MOVES CTI NPRM is 10 for HHD, 7 for MHD, and 5 for
LHD45 and LHD2b3.310
b For operating modes lacking data, we used the same Rinuse for the closest operating
mode.
Equation 5-6 is used to estimate the percentage reduction to NOx running emissions from the
change in the duty-cycle standard for each operating mode.sss
sss \ye assumed that the Rinusevalues calculated by MOVES operating mode can be applied to the MOVES rates
that are derived using a different fscaie. The change in fscaie does not change the definition of operating modes that are
not defined by Scaled Tractive Power, STP (deceleration operating mode 0, and idle operating mode 1), or operating
modes with negative STP values (operating mode 11 and 21), defined in Table 5-3. Changing the fscaie values does
change the definition of vehicle operation in the other operating modes. However, because the Rinusevalues are
relatively constant for the positive power operating modes within each speed range as observed in Table 5-7Figure
5-1, we deemed it was not necessary to attempt to account for the fscaie differences when applying the RinUse values.
216
-------
Equation 5-6
D _
^dutyjn.use
in_use
Where:
Rduty_in_use= the percent emission reductions in the in-use running NOx emissions estimated from changing the FTP
duty-cycle standard.
Equation 5-7 is used to estimate the heavy heavy-duty diesel NOx running emission rates
from the changes in the duty-cycle standard. The same calculations were applied to estimate the
other heavy-duty diesel regulatory classes.TTT
Where:
ERduty_in use= the MOVES running NOx emission rates for the control scenarios based on reduction in the duty-
cycle standard
Rduty_in_use= the percent emission reductions in running NOx emissions estimated from changing to FTP duty-cycle
standard
ERM0VES_baseiine= the MOVES baseline running NOx emission rates for each regulatory class
The estimated HHD MOVES running NOx emission rates for the control scenarios are shown
in Figure 5-2.
TTT We applied the Rin_use developed on HHD data to both the LHD45 and MHD regulatory classes, which have
different fscaie values in MOVES CTINPRM for the 2010 and later model years. We believe this approximation is
defensible for the similar reason provided in footnote SSS.
MOVES_baseline
217
-------
300
Base Rate
Option 2 MY 2027+
Option 1 MY 2027-2030 - Option 1 MY 2031 +
Alternative MY 2027+
LA*
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES Operating Mode
Figure 5-2: Duty-cycle-based running NOx emissions, ERduty_in_use, for HHD diesel for the control scenarios
5.2.2.1.1.2 Emission Rates Based on Off-cycle Standards
In this section, we document the methods used to estimate MOVES NOx running emission
rates based on the proposed and alternative off-cycle standards for heavy-duty diesel vehicles.
Table 5-8 presents the calculated off-cycle standards used to develop MOVES inputs for the
control scenarios. The proposed and alternative off-cycle standards all include requirements for
operating conditions in three bins: idle, low-load, and medium-to-high load. The off-cycle
operation is defined as follows: idle is less than 6 percent of maximum power; low-load is 6 to
20 percent of maximum power; and medium-to-high load is above 20 percent of maximum
power. The off-cycle standards for the proposed Option 1 are set to be twice the engine-cycle
standards for model year 2027-2030, and 1.5 times the engine-cycle standards for 2031+ that
correspond to similar engine loads. For the proposed Option 2 and Alternative scenarios, the off-
cycle standards are 1.5 times the engine-cycle standards for 2027+. In developing inputs to
MOVES, we did not apply a scaling factor to the off-cycle idling operation, and assumed
manufacturers would comply with the voluntary idle standard in all off-cycle idle operation. For
developing inputs to MOVES, we multiplied the duty-cycle standards, including the proposed
Option 1 weighted average HHD 2031+ NOx standards from Table 5-5, by the corresponding
off-cycle scaling factors shown in Table 5-8.
218
-------
Table 5-8. Calculated Off-Cycle NOx Standards used for the Control Scenarios
Scenario
Model
Year
Regulatory
Class
Engine Cycle
Reference
Off-cycle
NOx
Standard
Off-cycle Bin
Off-cycle
scaling
factor
Off-Cycle NOx
Standards
(g/hr for idling,
g/hp-hr for
low-load and
medium to
high-load)
Proposed
Option 1
2027-
2030
LHD,
MHD,
HHD
Idle (g/hr)
5 A
Idle, < 6% power
1
5
LLC (g/hp-hr)
0.09
Low-load, 6-20%
power
2
0.18
FTP & SET
(g/hp-hr)
0.035
Medium to High
Load, >20% power
2
0.07
2031+
LHD,
MHD
Idle (g/hr)
5
Idle, < 6% power
1
5
LLC (g/hp-hr)
0.05
Low-load, 6-20%
power
1.5
0.075
FTP & SET
(g/hp-hr)
0.02
Medium to High
Load, >20% power
1.5
0.03
HHD
Idle (g/hr)
5
Idle, < 6% power
1
5
LLC (g/hp-hr)
0.073 B
Low-load, 6-20%
power
1.5
0.110
FTP & SET
(g/hp-hr)
0.029 B
Medium to High
Load, >20% power
1.5
0.044
Proposed
Option 2
2027+
LHD,
MHD,
HHD
Idle (g/hr)
10
Idle, < 6% power
1
10
LLC (g/hp-hr)
0.1
Low-load, 6-20%
power
1.5
0.15
FTP & SET
(g/hp-hr)
0.05
Medium to High
Load, >20% power
1.5
0.075
Alternative
2027+
LHD,
MHD,
HHD
Idle (g/hr)
10
Idle, < 6% power
1
10
LLC (g/hp-hr)
0.1
Low-load, 6-20%
power
1.5
0.15
FTP & SET
(g/hp-hr)
0.02
Medium to High
Load, >20% power
1.5
0.03
A As discussed in the Preamble Section III.C, the proposed Option 1 idle off-cycle bin standards are 10 g/hr for MY 2027-
2030 and 7.5 g/hr for MY 2031+ However, we assumed vehicles comply with the voluntary idle standard presented in
Preamble Section III. B.2.iv rather than the off-cycle idle bin standard. The voluntary idle standard is 5 g/hr for both steps
of the proposed Option 1 and 10 g/hr for the proposed Option 2 and the Alternative.
B For HHD vehicles we used the calculated average standard from the intermediate and full useful life standard
presented in Table 5-5.
We calculated the voluntary idle NOx g/hr standard in units of NOx g/C02 kg using Equation
5-8, and the resulting values are displayed in Table 5-9.
219
-------
Equation 5-8
Voluntary Idle standard = [voluntary Idle NOx standard
Idle CO
(H
Where Idle C02 (^r)= the MOVES average CO2 (kg/hr) emission rate for HHD diesel vehicles for MOVES idle
(operating mode 1 defined in Table 5-3).
Table 5-9: Calculation of Voluntary Idle NOx/CCh Standard (g/kg)
Scenario
Applicable
Model
Years
Voluntary Idle NOx standard (g/hr)
Average HHD Idle
(Operating Mode=l)
CO2 emission rate
(kg/hr)
Voluntary
Idle
NOx/COi
standard
(g/kg)
Proposed
Option 1
2027-2030
5
8.024
0.62
2031+
5
8.024
0.62
Proposed
Option 2
2027+
10
8.024
1.25
Alternative
2027+
10
8.024
1.25
Next, we converted the reference off-cycle NOx standards into units of gram per hour (g/hr)
for each MOVES operating mode. We refer to g/hr rates as the off-cycle standard compliant
emission rates, which are shown in Table 5-10 in Columns (F) and (G) for HHD vehicles for
proposed Option 1.
220
-------
Table 5-10: Calculation of the Off-cycle NOx Standard Compliant Emission Rate for HHD Diesel Vehicles for
the Proposed Option 1
A
B
C
D
E
F
G
MOVES
operating
mode
MOVES
MY 2027
HHD
C02
emission
rate
(kg/hr)
Mean
power
(hp)
Percent
load
Power classification
MY
2027-
2030 off-
cycle
compliant
emission
rate (g/hr)
MY
2031+
off-cycle
compliant
emission
rate (g/hr)
0
5
4
0.90%
Idle
3.33
3.33
1
8
6
1.30%
Idle
5.00
5.00
11
13
0
-0.10%
Idle
8.04
8.04
12
23
29
6.10%
Idle
14.51
14.51
13
39
75
16.10%
Low-Load
13.58
8.22
14
53
122
25.90%
Medium to High Load
8.51
5.29
15
67
167
35.60%
Medium to High Load
11.70
7.27
16
109
285
60.60%
Medium to High Load
19.92
12.38
21
12
-5
-1.00%
Idle
7.24
7.24
22
32
34
7.20%
Idle
20.24
20.24
23
45
78
16.50%
Low-Load
13.96
8.46
24
60
122
26.00%
Medium to High Load
8.54
5.31
25
76
167
35.60%
Medium to High Load
11.72
7.29
27
99
231
49.10%
Medium to High Load
16.16
10.04
28
137
327
69.60%
Medium to High Load
22.90
14.23
29
174
408
86.90%
Medium to High Load
28.56
17.75
30
200
473
100.70%
Medium to High Load
33.10
20.57
33
29
34
7.30%
Idle
17.55
17.55
35
72
147
31.20%
Medium to High Load
10.28
6.39
37
104
228
48.60%
Medium to High Load
15.99
9.94
38
141
321
68.40%
Medium to High Load
22.50
13.98
39
175
402
85.60%
Medium to High Load
28.16
17.50
40
197
470
100.00%
Medium to High Load
32.89
20.44
In Table 5-10, Column (B) contains the MOVES CO2 emission rate for Model Year 2027
HHD diesel vehicles for each MOVES operating mode. Column (C) includes the mean power for
each operating mode bin calculated from the Heavy-Duty In-Use Testing data, which is the same
data set that was used to derive the MOVES CO2 emission rates.310 The percent load, Column
(D), is calculated for each operating mode bin using Equation 5-9.
221
-------
Percent Load0pMode=i =
Equation 5-9
Mean Power 0pMode=i
Mean PowerOpMode=40
Where i= one of the 23 MOVES running exhaust operating modes from 0 to 40
Mean Power 0pMode= t= the mean power for each of the MOVES operating modes shown in Column C.
Mean Power0pMode=40= assumed maximum power bin = 470 hp for HHD diesel vehicles.
We then assigned each MOVES operating mode into a power classification (Column (E))
based on the percent load (Column (D)), where percent load less than 8 percent of maximum
power is idle, 8 to 25 percent of maximum power is low-load, and above 25 percent of maximum
power is medium-to-high load.uuu
The off-cycle NOx standard compliant emission rate in Column (F) and (G) is then calculated
based on the power classification and the stringency of the off-cycle standard. For operating
modes classified as idle, we multiplied the MOVES CO2 emission rate in Column (B) with the
N0x/C02 off-cycle idle standard calculated in Table 5-9 for the corresponding control scenario
using Equation 5-10.
Equation 5-10
Idle Emission Rate
= MOVES MY 2027 HHD C02 Emission Rate
hr^
NOX /g\
x Idle in use standard -—
C02 - Vkg/
For the operating modes classified as low-load and medium- to high-load, we multiplied the
off-cycle (g/hp-hr) standard in Equation 5-11 for the corresponding control scenario and power
classification by the mean power (Column C), as shown in Equation 5-11.
Equation 5-11
Low Load and Medium to High Load Emission Rate
= Mean Power(hp) x In_use standard ^ ^ )
The estimated off-cycle NOx standard compliant emission rates for heavy heavy-duty diesel
vehicles are shown in Figure 5-3. Similarly, we applied Equation 5-9 through Equation 5-11 to
1,1111 At the time of our emission inventory analysis, we used a definition of less than 8% maximum power to define
idle operation in-use (Bin 1), and 8% to 25% of maximum power for low-load (Bin 2), and greater than 25%
maximum power as medium-to-high power (Bin 3). This definition of idle places more operation into low-load bins
with more stringent emission standards than the idle bin (Bin 2 is more stringent than Bin 1). In the case of HHD,
three operating modes (operating mode 12, 22 and 33) should be classified in the low-load bin rather than the idle
bin. Due to the coarseness of the MOVES operating mode bins, no operating mode would be changed from low-load
to medium-to-high-power and this change in definitions is expected to have a minor impact on the final emission
rates. As shown in Figure 5-4 for two of the three HHD low-load operating modes affected by this change (operating
mode 22 and 33), the final emission rate is determined by the certification standard, not the in-use standard.
Nevertheless, we plan to update our emission inventory analysis for the final rulemaking based on the definitions of
in-use operations selected for the final rule, which will be informed by stakeholder feedback on the proposed rule.
222
-------
estimate the off-cycle standard compliant emission rates for the other MOVES regulatory classes
using corresponding CO2 rates and the mean power for those vehicles.
300
250
200
150
100
50
i Base Rates
Option 1 MY 2027-2030 Off-Cycle Standard Compliant Rate
Option 1 MY 2031+ Off-Cycle Standard Compliant Rate
Option 2 MY 2027 Off-Cycle Standard Compliant Rate
Alternative MY 2027 Off-Cycle Standard Compliant Rate
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES Operating Mode
Figure 5-3: Base NOx rates and off-cycle NOx standard compliant emission rates for HHD diesel
5.2.2.1.1.3 Emission Rates Based on Combination of Duty-Cycle and Off-Cycle
Standards
In this section, we document the methods used to develop MOVES NOx emission rates for
heavy-duty diesel vehicles that reflect the effects of both duty-cycle standards and off-cycle
standards. As an example, Figure 5-4 shows that the proposed Option 1 HHD duty-cycle and off-
cycle standards in MYs 2031 and later affect running emission rates differently across MOVES
operating modes. The duty-cycle standard is estimated to have a larger impact than the off-cycle
standard in five operating modes (operating modes 0, 22, 33, 35, and 37), while the off-cycle
standard is estimated to have a larger impact in the remaining operating modes.
223
-------
MY 2031+ Off-Cycle Compliant Rates
MY 2031+ Duty-Cycle Compliant Rates
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES Operating Mode
Figure 5-4: Comparison of Running NOx emission rates for diesel-fueled HHD compliant with the MY2030
proposed Option 1 duty-cycle and off-cycle standards
Because manufacturers would need to comply with both the duty-cycle and off-cycle
standards, we estimated the final MOVES NOx emission rate for each operating mode as the
lower of the two rates generated from the duty-cycle and the off-cycle standards (e.g., the
emission rate based on the proposed Option 1 off-cycle standards is selected for operating mode
12, but the emission rate based on the proposed Option 1 duty-cycle standards is selected for
operating mode 33). Figure 5-5 presents the estimated emission rates for HHD diesel vehicles
that meet both the proposed Option 1 duty-cycle and off-cycle standards. The same approach
was used to estimate the emission rates for proposed Option 2 and alternative scenarios. The
proposed Option 1 emission rates for MHD, LHD45 regulatory classes are shown in Appendix
5.5.1.
224
-------
300
~250
CD
05
cr
.1 200
CO
(/)
X 150
O
z
«
£ 100
o
50
¦ Base Rate Option 1 MY 2027-2030 "Option 1 MY 2031 +
i Option 2 MY 2027+ "Alternative MY 2027+
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES Operating Mode
Figure 5-5: Estimated zero-mile NOx emission rates for HHD diesel vehicles due to the proposed Option 1,
Option 2 and Alternative duty-cycle and off-cycle standards
5.2.2.1.2 Emission Rates Based on Proposed Changes in Warranty and Useful Life
The MOVES NOx, THC, CO, and PM2.5 emission rates for heavy-duty diesel engines in the
control scenario are adjusted to reflect the useful life and warranty periods for the proposed and
alternative scenarios shown in Table 5-11.
225
-------
Table 5-11: Useful Life and Warranty Periods for Heavy-duty DieselA Engines and Aftertreatment Systems in
the Control Scenarios
Scenario
Applicable Model
Years
Warranty
Useful Life
LHD
MHD
HHD
LHD
MHD
HHD
Baseline
Model Year 2010+
5yr/ 50k
5yr/ 100k
5yr/ 100k
10yr/
110k
10yr/
185k
10yr/
435k
Proposed
Option 1
Model Year 2027-2030
7yr/ 150k
7yr/ 220k
7yr/ 450k
12yr/
190k
llyr/
270k
llyr/
600k
Model Year 2031+
10yr/
210k
10yr/
280k
lOy/600k
12yr/
270k
12yr/
350k
12yr/
800k
Proposed
Option 2
Model Year 2027+
5yr/ 110k
5yr/ 150k
5yr/ 350k
10yr/
250k
10yr/
325k
10yr/
650k
Alternative
Model Year 2027+
10yr/
280k
10yr/
360k
10yr/
800k
10yr/
350k
10yr/
450k
10yr/
850k
A Age effects for heavy-duty gasoline vehicles in MOVES are estimated directly from emissions data or
adapted from light-duty gasoline or heavy-duty diesel age effects, and are not tied to warranty and useful life periods;
thus we have not adjusted heavy-duty gasoline or NG engine emission rates due to the proposed and alternative
warranty and useful life periods.
B The MOVES baseline emission rates for LHD diesel were incorrectly developed using a warranty period
of 100K instead of 50K, and thus we underestimated the emissions impact of the longer warranty periods for LHD
diesel vehicles in the proposal and alternative scenarios. We will update the baseline warranty period for LHD diesel
in the FRM analysis.
We used the existing methodologyvvv in MOVES to estimate the impact of lengthened useful
life and warranty periods on heavy-duty diesel engine emissions for each of the three control
scenarios (proposed Options 1 and 2 and Alternative).www In that methodology, new
vehicles/engines have zero-mile emission rates for each operating mode and maintain that rate
until the age of the vehicle/engine matches the warranty period (Figure 5-6). Once the warranty
period ends, the emission rate increases linearly until the vehicle/engine reaches its useful life
age. At the end of the useful life, the emissions rates remain constant at a level calculated from
the tampering and mal-maintenance (T&M) adjustment factor. In MOVES, we assume that
tampering and mal-maintenance effects are the dominant source of emissions deterioration of
fleet-wide heavy-duty diesel emissions. MOVES does not account for normal deterioration of
heavy-duty diesel emissions, such as due to catalyst aging.
vvv The existing methodology is documented in Appendix B "Tampering and Mal-maintenance" of the reference
"Exhaust Emission Rates for Heavy-Duty Onroad Vehicles in MOVES CTINPRM"310
Emission rate inputs for the proposed Option 1 and 2 and Alternative scenarios without aging effects (zero-
mile rates) are included with the other MOVES input tables in the docket.308 The zero-mile proposal emission rates
are used to quantify the emission reductions due to the different control aspects of the proposed rulemaking as
shown in Figure 5-24.
226
-------
Emission rate
Final emission rate
due to T&M
Zero-mile
emission rate
End of warranty End 0f useful life
period
Age
Figure 5-6: Methodology to model the effects of tampering and mal-maintenance (T&M) on emission rates
according to warranty and useful life
For the baseline and control scenarios, we estimated the vehicle age at which heavy-duty
diesel vehicles would reach the end of their warranty period and the end of their useful life
period (Table 5-13). Table 5-12 shows example calculations for the proposed Option 1 MY 2027
standards. Row (A) shows the age limit of the standards: 5-year warranty period, 10-year useful
life period. Row (B) shows the mileage limit of the standards. Row (C) shows the typical miles
driven per year,xxx which is used to calculate Row (D), the calculated age rounded to the nearest
whole number when the mileage limit is reached. Row (E) is the smaller of the age at which the
vehicle meets the end of its age limit, Row (A), or mileage limit, Row (D).
xxx The typical miles per year used in Table 5-12 are the same values used in deriving the vehicle age at the end of
the warranty period and baseline warranty and useful life assumed in MOVES CTINPRM for the baseline case.31"
227
-------
Table 5-12: Estimated Vehicle Age at the End of the Warranty Period and the Useful Life for Each Heavy-
duty Diesel Regulatory Class for the Proposed Option 1 Model Year 2027 Scenario
Row
Warranty
Useful Life
LHD
MHD
HHD
Bus
LHD
MHD
HHD
Bus
(A)
Age limit
7
7
7
7
12
11
11
11
(B)
Mileage
limit
150,000
220,000
450,000
450,000
190,000
270,000
600,000
600,000
(C)
Typical
miles/year
driven
26,000
41,000
105,000
44,000
26,000
41,000
105,000
44,000
(D)
Calculated
age when the
mileage
limit is
reached
6
5
4
10
7
7
6
14
(E)
Estimated
age
6
5
4
7
7
7
6
11
Similar calculations were performed for other regulatory classes for the baseline and control
scenarios using the same estimated mileage per year; the resulting estimates of vehicle age at the
end of the warranty and useful life periods are shown in Table 5-13.
Table 5-13: Estimated Vehicle Age at the End of the Warranty Period and the Useful Life for Each Heavy-
duty Diesel Regulatory Class in the Baseline and Control Scenarios
Warranty
Useful Life
MOVES age
LHD
MHD
HHD
Bus
LHD
MHD
HHD
Bus
Baseline
4
2
1
2
4
5
4
10
Proposed Option 1 Model Year
2027 - 2030
6
5
4
7
7
7
6
11
Proposed Option 1 Model Year
2031+
10
8
8
12
10
8
8
12
Proposed Option 2 Model Year
2027
4
4
4
4
4
10
8
8
Alternative Model Year 2027
10
9
8
10
10
10
8
10
The T&M adjustment factor is calculated as the sum of the product of the T&M frequency for
each failure z, and the corresponding T&M emission effect, as shown in Equation 5-12.
Equation 5-12
X (T&M frequency; x T&M emission effectp j)
Where:
fx&M= the tampering and mal-maintenance adjustment factor for pollutant p
T&M frequency; = estimated fleet average frequency of a tampering & mal-maintenance failure i.
T&M emission effectj= estimated emission effect for pollutant p associated with tampering & mal-maintenance
failure i.
228
-------
The emission rate at the end of useful life is then calculated using Equation 5-13.
Equation 5-13
E^End of useful life,p,r,o — ERzero mile,p,r,o ^ (l
Where:
ERusefui iife,P,r,o = the heavy-duty diesel emission rate at the end of warranty for each pollutant p, regulatory class, r,
and operating mode, o
ERZero miie~ the zero-mile heavy-duty diesel emission rate for each pollutant p, regulatory class, r, and operating
mode, o
fx&M= the tampering and mal-maintenance adjustment factor for each pollutant p (Equation 5-12)
We used the existing T&M frequency values from MOVES CTINPRM for the baseline and
control scenarios.310 For THC, CO, and PM2.5, we used the existing T&M emission effects from
MOVES CTI NPRM for both the baseline and control scenarios.310
For the NOx T&M emissions effects, the baseline scenario is estimated using the existing
MOVES CTI NPRM emission effects shown in Table 5-14. For the control scenarios, however, a
different approach was taken. As NOx emissions become more tightly controlled due to the
application of advanced technologies, we anticipate the NOx T&M emission effects will
increase. To estimate the NOx T&M emission effects for the control scenarios, we first
calculated the average zero-mile NOx emission rate ERzero miie,NOx, the weighted average of the
different operating modes o, and regulatory class r, using Equation 5-14.
Equation 5-14
¦prpf _ 2r,o(ERzero mile,NOx,r,o ^ ^r,o )
k^zero mile,NOX — y Z
Zjt.o tr,o
Where:
ERzero mile,nox = the average heavy-duty diesel NOx emission rate
ERzero miie,Nox,r,o = the zero-mile heavy-duty diesel NOx emission rate for regulatory class, r, and operating mode, 0
tr,o= operation time by regulatory class and operating mode estimated by MOVES CTI NPRM for calendar year
2045
Next, we estimated the NOx emission rate of vehicles with a tampering and mal-maintenance
failure i, using Equation 5-15, which was derived from Equation 5-13 using the fleet average
emission rate from Equation 5-14 and assuming the T&M frequency is 100 percent.
Equation 5-15
ERt&m i.Nox = ERzero mile,NOX x (l + T&M emission effect; N0X)
We then derived Equation 5-16, assuming that a NOx aftertreatment equipment failure i, in
the control scenario, would cause the average of the failed emission rates, ERt8lM ,-i/VOx, to be the
same as a NOx aftertreatment failure in the baseline case, Baseline ERx&Mi N0X.
229
-------
Equation 5-16
Baseline ERx&Mi NOx — Control ERx&Mi NOx
Baseline ERzero mne,Nox x (l + Baseline T&M emission effect; NOx)
= Control ERzero miie,Nox x (l + Control T&M emission effect; NOx)
By rearranging Equation 5-16, we derived Equation 5-17 to estimate the control scenario NOx
T&M emissions effects.
Equation 5-17
Control T&M emission effect; N0X
Baseline ERzeromiie,nox x (l + Baseline T&M emission effect; NOx)
Control ERzero miie,Nox
Table 5-14 presents the T&M NOx emission effects for the NOx aftertreatment failures for
the control scenarios calculated from Equation 5-17. The T&M NOx emission effects for the
NOx aftertreatment failures are much larger than the baseline scenario, because the zero-mile
NOx emission rate in the control scenarios are lower than the baseline zero-mile NOx emission
rates. The NOx T&M emission effects for the other T&M failures shown in Table 5-14 (e.g.,
Timing Advanced and EGR Disabled/Low-Flow) in the control scenarios use the same NOx
T&M emissions effects in the baseline and proposed and alternative control scenarios.
230
-------
Table 5-14: NOx Tampering & Mal-maintenance (T&M) Emission Effects
Baseline
Proposed
Option 1 MY
2027-2030
Proposed
Option 1
MY 2031+
Proposed
Option 2 MY
2027+
Alternative
MY 2027+
Timing Advanced
6%
6%
6%
6%
6%
Timing Retarded
-20%
-20%
-20%
-20%
-20%
Injector Problem (all)
-1%
-1%
-1%
-1%
-1%
Puff Limiter Mis-set
0%
0%
0%
0%
0%
Puff Limited Disabled
0%
0%
0%
0%
0%
Max Fuel High
0%
0%
0%
0%
0%
Clogged Air Filter
0%
0%
0%
0%
0%
Wrong/Worn Turbo
0%
0%
0%
0%
0%
Intercooler Clogged
3%
3%
3%
3%
3%
Other Air Problem
0%
0%
0%
0%
0%
Engine Mechanical Failure
-10%
-10%
-10%
-10%
-10%
Excessive Oil Consumption
0%
0%
0%
0%
0%
Electronics Failed
0%
0%
0%
0%
0%
Electronics Tampered
8%
8%
8%
8%
8%
EGR Stuck Open
-20%
-20%
-20%
-20%
-20%
EGR Disabled/Low-Flow
5%
5%
5%
5%
5%
NOx Aftertreatment Sensor
200%
1648%
2101%
1097%
1814%
Replacement NOx Aftertreatment Sensor
200%
1648%
2101%
1097%
1814%
NOx Aftertreatment Malfunction
500%
3395%
4303%
2294%
3728%
PM Filter Leak
0%
0%
0%
0%
0%
PM Filter Disabled
0%
0%
0%
0%
0%
Oxidation Catalyst Malfunction/Remove
0%
0%
0%
0%
0%
Using the NOx T&M emission effects in Table 5-14, we then calculated T&M adjustment
factors fx&M.NOx f°r each scenario using Equation 5-12 and the baseline T&M frequency values.
For THC, CO, and PM2.5, we used the existing T&M adjustment factors fx&M,p in MOVES CTI
NPRM. Then, we calculated the heavy-duty diesel emission rate for each pollutant p, age a,
regulatory class r, and operating mode o, using Equation 5-18.
Equation 5-18
ERp,r,a,o — ERzero mile,p,r,o ^ (1 ^ ^T&m)
Where:
ERP,r,o,a = the heavy-duty diesel emission rate for each pollutant p, regulatory class r, age a, operating mode, 0,
ERzero miie~ the zero-mile heavy-duty diesel emission rate for each pollutant p, regulatory class r, operating mode, 0
sa= scaled age effect at age a
fx&M= the tampering and mal-maintenance adjustment factor (Equation 5-12)
The scaled age effect, sa, is calculated using the age of the vehicle in comparison to the
warranty and useful life requirements, as shown in Table 5-15. When the vehicle age is between
the end of the warranty and the useful life, sa is interpolated between 0 and 1.
231
-------
Table 5-15: Calculation of sa
Where:
0
age < end of warranty age
(age — end of warranty age)
(Useful life age — end of warranty age)
end of warranty age < age < useful life
1
age > useful life
As the final step, the age-adjusted emission rates calculated in Equation 5-18 were averaged
according to the age ranges shown in Table 5-16 that are used to define emission rates in
MOVES for LHD45, MHD, HHD, and Urban Bus regulatory classes. The resulting age-adjusted
running emissions have a relationship with vehicle age as shown in Figure 5-7 and Figure 5-8 for
HHD NOx emissions.^ ^
Table 5-16: MOVES ageGroupID Which Are Used to Define Running and Start Emission Rates
ageGroupID
Lower bound
(years)
Upper bound
(years)
3
0
3
405
4
5
607
6
7
809
8
9
1014
10
14
1519
15
19
2099
20
30
5.2.2.1.3 Summary of Diesel NOx Running Emission Rates
Figure 5-7 and Figure 5-8 show average running NOx emission rates (g/mile) in MOVES CTI
NPRM for the model year 2027 and 2031 fleets across vehicle age for the baseline and control
scenarios.zzz The MOVES running emission rates for the control scenarios reflect the
adjustments to the duty-cycle and off-cycle standards, and the extended warranty and useful life
as discussed above in Chapter 5.2.2.1.1 and Chapter 5.2.2.1.2, respectively. The gram per mile
average running emissions are also a function of the default activity assumptions in MOVES CTI
NPRM.309
Figure 5-7 and Figure 5-8 show that the average zero-mileAAAA NOx emission rates for the
control scenarios are significantly lower than the baseline scenario. The figures also demonstrate
YYY The average emission rate accounts for the frequency of different operating modes according to MOVES
estimate of in-use vehicle activity. The trend in individual operating modes will be slightly different than the
average trend shown in Figure 5-7 and Figure 5-8. For example, the zero-mile idle operating mode is not reduced as
much as the average emission rates in the control scenarios. Because the control case T&M emission effects were
calculated using the average emission rates in Equation 5-17, individual emission rates in the control case, such as
for idle, can be higher than the baseline scenario when fully aged. This is a feature of the method used to derive the
aging effects, but the effect is averaged out when conducting emission inventory analysis.
zzz The average emission rates are estimated from the national MOVES runs described in Section 0.
AA AA The age 0-3 emission rates shown in Figure 5-7 for the proposed Option 1 and Option 2 do not represent the
true zero-mile rates because the warranty period for HHD vehicles ends at age 3 (Table 5-13), and the MOVES
232
-------
the larger T&M NOx emission effect for the control scenarios than for the baseline scenario as
explained in Chapter 5.2.2.1.2. Although not shown, the emission rate is constant from age 15
through age 30.. The impact of the longer emission warranty and useful life periods is reflected
by the zero-mile NOx emission rate extending for additional time periods, which corresponds to
the emission warranty periods in the proposed Options 1 and 2 and Alternative.
Figure 5-7: NOx g/mile emission rates estimated from MOVES for HHD diesel long-haul combination trucks
for the model year 2027 fleet across vehicle age for the baseline and control scenarios
emission rates are averaged according to the age groups shown in Table 5-16. In contrast, the emission rate for the
proposed Option 1 scenario for the age 0-3 group in Figure 5-8 is equivalent to the zero-mile rate because the
warranty period extends to age 6.
233
-------
Figure 5-8: NOx g/mile emission rates estimated from MOVES for HHD diesel long-haul combination trucks
for the model year 2031 fleet across vehicle age for the baseline and control scenarios
5.2.2.2 Heavy-Duty Diesel Start Emission Rates
In this section, we describe the methods used to estimate lower start NOx emission rates in
MOVES CTINPRM for the control scenarios due to the proposed and alternative duty-cycle
standards. We did not estimate the impact of the off-cycle standard on start emissions, in part
because the baseline MY 2010 and later start emission rates in MOVES CTI NPRM are not
based on in-use data but are based on emissions data from the FTP duty-cycle.310 Additionally,
because the baseline heavy-duty diesel start emission rates estimated in MOVES CTI NPRM do
not vary with age due to insufficient data310, we did not estimate changes due to the proposed
changes in warranty and useful life.
Start emission rates in MOVES are defined by regulatory class, fuel type, vehicle age and
operating mode. Start operating modes in MOVES are also defined by different lengths of engine
time (the time between the preceding engine off event and the engine start). The length of soak
time for each MOVES operating mode bin is defined in Figure 5-10. Operating mode 108
represents a start with a soak longer than 720 minutes or 12 hours and is referred to as a 12-hour
cold-start.
To estimate the start emissions under the control scenarios, we estimated the NOx cold start
emission rate (g/start) from the CARB Stage 1 HDD engine tested on the FTP duty-cycle cycle.
Table 5-17 contains the NOx Cold and Hot FTP measurements in Columns (B) and (C) for
different aging lengths. Cold - Hot, Column (E), is calculated as the difference between the
Columns (B) and (C). The cold start, Column (F), is then calculated by multiplying the
difference in the Cold and Hot Start, Column (E), by the work performed on the FTP cycle,
Column (D), as shown in Equation 5-19.
234
-------
Equation 5-19
NOx Cold Start (-M = [cold f, g, ) - Hotf, g, )
x Vstart/ L Vhp ¦ hr/ Vhp ¦ hr/
x FTP work (hp ¦ hr)
Table 5-17: Calculation of NOx 12-hour Cold Starts from the CARB Stage 1 HHD Engine from the Cold and
Hot FTP Cycle
(A)
(B)
(C)
(E)
(D)
(F)
Aged hours
FTP composite
(g/hp-hr)
Cold
(g/hp-hr)
Hot
(g/hp-hr)
Cold - Hot
(g/hp-hr)
FTP Work
(hp-hr)
Cold Start
(g/start)
0
0.008
0.025
0.005
0.02
31.4
0.63
333
0.012
0.042
0.006
0.036
31.4
1.13
656
0.018
0.061
0.009
0.052
31.4
1.64
1000
0.024
0.092
0.01
0.082
31.4
2.58
1000 hr Post Ash
Clean
0.026
0.109
0.009
0.1
31.4
3.14
The Stage 1 HHD engine is deemed representative of an engine-certified to a 0.02 g/hp-hr
NOx standard based on the FTP composite measurements in Column (A). Table 5-17
demonstrates that there was larger cold start measured with increase in aged hours, and after the
ash clean out at 1000 hours. We used the 1000 hr, Post Ash Clean cold start emission rate (3.14
g/start shown in Table 5-17) to represent the 12-hour cold-start (operating mode 108) emission
rate for the Alternative scenario.
To estimate the 12-hour cold-start NOx emission rate FtHD diesel vehicles for the control
scenarios, we interpolated the FtHD 12-hour cold-start between the Stage 1 cold start (3.14
g/start) and the MOVES baseline 12-hour cold-start (8.4 g/start), and their respective FTP duty-
cycle standards using Equation 5-20 as shown in Figure 5-9 and Table 5-18 . For example, the
interpolation yielded an estimated 12-hour cold start of 4.02 g/start for the 0.05 g/hp-hr FTP
standard.
Equation 5-20
Start ERFXPx HHD12 hour
^MOVES start hhd,12 hour-Stage lstart^
= I Baseline FTP - Stagel FTP j X (FTP* " BaSelme FTP)
+ MOVES Start HHD,12 hour
Where:
Start ERDuty_cycle standard x,hhd,i2 hour = the estimated NOx start emissions for an FTP duty-cycle standard, x, for
heavy heavy-duty diesel emissions for a 12-hour cold-start (operating mode 108).
Stagel start = 1000 Post Ash Clean start emission rate from the CARB Stage 1 HHD diesel engine = 3.14 g/start
(Table 5-17)
Stagel FTP = Composite FTP level of the CARB Stage 1 engine = 0.02 g/hp-hr
MOVES start Hhd,i2 hour= MOVES CTINPRM baseline start emission rate for MY 2027 heavy heavy-duty diesel
engine for a 12 hour soak (operating mode 108)
Baseline FTP = baseline FTP composite NOx standard = 0.2 g/hp-hr
235
-------
FTPX = composite FTP standard in the control scenarios, either 0.035, 0.029, 0.05, or 0.02 g/hp-hr (Table 5-4 and
Table 5-5)
9
___ 8
4-»
i—
OJ 7
-M /
i/)
^6
4-»
L_
OJ r
in 5
-a
o 4
u
= 3
o
2
r^j
H
o 1
X
xo
C
) 0.05 0.1 0.15 0.2 0.25
FTP Composite Standard (g/hp-hr)
Figure 5-9: Estimated relationship between the HHD NOx 12-hour cold-start and the composite FTP NOx
standards
Table 5-18: HHD Cold Start Emissions for Proposed and Alternative Scenarios
Scenario
Applicable Model
Years
Weighted
Average FTP
standard (g/hp-
hr)
Cold Start
emissions (g/start)
Baseline
Model Year 2010+
0.2
8.40
Proposed Option 1
Model Year 2027-2030
0.035
3.58
Model Year 2031+
0.029
3.41
Proposed Option 2
Model Year 2027+
0.05
4.02
Alternative
Model Year 2027+
0.02
3.14
We then used Equation 5-21 to estimate the MOVES NOx emission rates for each MOVES
heavy-duty regulatory class (LHD45, MHD, and HHD) and for each MOVES start operating
mode classified by different soak times (Figure 5-10). We assumed that the relative difference in
emission rates by regulatory class and by operating mode is the same in the baseline and control
scenarios.
Equation 5-21
Start ERFXP=x reg class=y,soak=z
— Start ER[)uty cycle standard x,HHD,12
/MOVES Start reg class=y,soak=z\
Xl MOVES Start HHD,12-hour )
236
-------
Where:
Start ERfxp= the start N0X emission rates for the control scenarios with FTP x (0.035, 0.029, 0.05, or 0.02) for
regulatory class y (LHD45, MHD, and HHD), and soak length z
Start ERDuty cycle standard x,hhd,i2—hour = the estimated start emissions for an FTP duty-cycle standard, x, for heavy
heavy-duty diesel emissions for a 12-hour soak (operating mode 108)
MOVES start reg ciass=y,soak=z = MOVES CTINPRM baseline start emission rate for MY 2027 for regulatory class
y (LHD45, MHD, and HHD), and soak length z
MOVES start HHD,i2-hour= MOVES CTI NPRM baseline start emission rate for MY 2027 HHD diesel engine for a
12-hour soak (operating mode 108)
Figure 5-10 shows the estimated MOVES NOx start emission rates for HHD diesel vehicles
for the baseline scenario, as well as the proposed and alternative scenarios.
¦ Base Rate ¦ Option 1 MY 2027-2030 ¦ Option 1 MY 2031 +
is Option 2 MY 2027+ "Alternative MY 2027+
9
¦e
~ 8
en u
3
« 7
OJ
cm
<6min 6-30 min 30-60 min 60-90 min 90-120 min 120-360 360-720 720+ min
min min
MOVES Start Operating Mode
Figure 5-10: Duty-cycle-based NOx start emissions for HHD Diesel for the baseline, proposed, and alternative
control scenarios
5.2.2.3 Heavy-Duty Extended Idle Emission Rates
In MOVES, extended idling is defined as idling more than an hour, which occurs during
hotelling activity when long-haul combination trucks idle during rest periods. MOVES has
extended idle emission rates for long-haul combination trucks that include HHD, MHD,BBBB and
BBBB HHD and MHD have the same extended idle emission rates in MOVES.
237
-------
glider vehiclescccc. All idling activity by other regulatory classes is modeled using the running
idle emission rates (Table 5-3), which are different than extended idle emission rates. We
anticipated that reductions in the HHD and MHD NOx extended idle emissions rates would be
driven by the proposed idle standard, rather than the proposed duty-cycle standards. The duty-
cycle standards do not contain high duration extended idling (> 1 hour) that is representative of
truck hotelling activity. In addition, we did not estimate lower extended idle emission rates due
to the lengthened warranty or useful life periods.DDDD
We estimated extended idle emission rates that would comply with the off-cycle NOx/CCh
g/kg standard calculated in Table 5-9. We then used Equation 5-10 to calculate the extended idle
off-cycle NOx g/hr emission rate based on the MOVES extended idle CO2 g/hr emission rate, as
shown in Table 5-19.
Table 5-19: Calculation of HHD and MHD Extended Idle NOx g/hr Emission Rates
Control Scenario
Model Years
2027 MY
Baseline
Rates
NOx
(g/hr)
2027 MY
Baseline
Rates
CO2
(g/hr)
Idle
Standard
(g/hr)
Idle
Standard
N0x/C02
(g/kg)
Idle-
standard
compliant
NOx
emission
rate
(g/hr)
%
Change
in NOx
emission
rate
Proposed
Model Year 2027-2030
42.6
7191
5
0.6232
4.48
-89%
Option 1
Model Year 2031+
42.6
7191
5
0.6232
4.48
-89%
Proposed
Option 2
Model Year 2027+
42.6
7191
10
1.25
8.99
-79%
Alternative
Model Year 2027+
42.6
7191
10
1.25
8.99
-79%
5.2.2.4 Heavy-duty Diesel Crankcase Emissions
MOVES CTINPRM estimates that crankcase emissions contribute 38% of the PM2.5 exhaust
emissions for MY 2007 and later heavy-duty diesel vehicles and assumes that all heavy-duty
diesel vehicles have open crankcase systems in the baseline.310 However, approximately one-
third of heavy-duty diesel vehicles have closed crankcase systems, as described in Section III.B
of the preamble and Chapter 1.1.4 of this draft RIA. Under the control scenarios, the PM2.5
crankcase emissions from the HHD, MHD, and LHD45 diesel vehicles are set to zero. For
LHD2b3 diesel vehicles, we reduced the diesel emissions by 94.9%, because we assumed that
5.1%) of the LHD2b3 diesel vehicles are engine-certified (Chapter 5.2.2.5). Because we are
assuming that all heavy-duty diesel engines have open crankcase systems in the baseline, we
anticipate that we are slightly over-estimating the PM2.5 benefits of the closed crankcase
requirement in the control scenarios.
cccc We assumed there are no changes to glider emission rates due to the proposed rulemaking.
dddd Extended idle emission rates in MOVES are not differentiated by vehicle age.
238
-------
MOVES CTINPRM does not estimate the contribution of crankcase emissions to THC, NOx,
and CO exhaust emissions from heavy-duty diesel vehicles. Because of this, we did not estimate
reductions in these gaseous pollutants from crankcase control.
For the final rulemaking, we anticipate updating MOVES baseline crankcase emission rates to
account for the current fraction of closed crankcase systems; incorporate estimates of THC, NOx,
and CO crankcase emissions into the baseline; and incorporate newer crankcase emission
estimates from a more recent test program. With these updates, we anticipate that compared to
this analysis, estimated benefits from the closed crankcase requirement could potentially increase
for THC, NOx, and CO, and decrease for PM2.5.
5.2.2.5 Light Heavy-Duty Class 2b and 3 Diesel Emission Rates
We assumed that in 2027 and later model years, 5.1 percent of the diesel-fueled LHD2b3
vehicles will be engine-certified and would be impacted by the proposed rule,EEEE>FFFF To
develop the emission rates for LHD2b3 vehicles, for the control scenarios, we assumed that 5.1
percent of the emissions from LHD2b3 are equivalent to the controlled emissions of LHD45
regulatory class vehicles. This is consistent with the analysis for model year 2010 and later diesel
vehicles, where we used the same data from the HDIUT program to estimate emission rates for
both LHD2b3 and LHD45 vehicles.310 In addition, we assumed that the proposed duty-cycle, off-
cycle, and warranty and useful life requirements are the same for all engine-certified LHD
vehicles.
We did not estimate the contribution of engine-certified vehicles on the emission rates for
diesel-fueled LHD2b3 in the baseline scenario.GGGG Because the LHD2b3 diesel emission rates
certified to engine standards are higher than the emission rates certified to chassis standards, the
control scenarios generally increase NOx emissions compared to the baseline scenario for diesel
LHD2b3 vehicles for most calendar years, even though we anticipate the proposed rule would
reduce NOx emissions for engine-certified LHD2b3 vehicles. We included the emission
contribution from engine-certified diesel-fueled LHD2b3 vehicles in the control scenarios to
provide our most accurate estimation of future NOx emissions, while acknowledging that we are
underestimating the benefits of controlling these vehicles due to their absence from the baseline
eeee as noted in Section I, Class 2b and 3 vehicles with GVWR between 8,500 and 14,000 pounds are primarily
commercial pickup trucks and vans and are sometimes referred to as "medium-duty vehicles". The majority of Class
2b and 3 vehicles are certified as complete vehicles and will be included in a future combined light-duty and
medium-duty rulemaking action, per Executive Order 14037, Section 2a.
FFFF In Appendix 5.5.1, we present the analysis suggesting that 4.2% of MY 2027 diesel-fueled LHD2b3 vehicles
would be engine-certified. However, we mistakenly used 5.1% in the development of the MOVES rates for LHD2b3
vehicles and subsequent inventory analysis. Given the small contribution of engine-certified LHD2b3 to the total
emissions inventory, we expect the difference between the intended and used values to have a negligible impact on
the NPRM analysis. In addition, we deem that both values (4.2% and 5.1%) are within the range of feasible values
for the fraction of engine-certified LHD2b3 vehicles in future years. Nonetheless, for the final rulemaking analysis,
we may revisit our estimate of the fraction of engine-certified engines in 2027 and later model years, depending on
available data and resources.
gggg jn (]lc baseline case created for MOVES CTI NPRM, we assumed, for simplicity, that 100% of diesel-fueled
LHD2b3 vehicles are chassis-certified and are subject to the light-duty Tier 3 emission standard. The estimated NOx
emission rates for engine-certified diesel LHD2b3 vehicles (subject to the proposed rule) are higher than chassis-
certified diesel LHD2b3 vehicles (subject to the light-duty Tier 3 standard).
239
-------
scenario. We plan on revisiting our assumptions to exclude the contribution of engine-certified
vehicles from baseline emission rates for diesel-fueled LHD2b3 vehicles in the final rulemaking
analysis.
5.2.2.6 Heavy-Duty Gasoline Running Emission Rates
In this section, we describe the methods used to develop the running exhaust emission rates
for NOx, THC, CO, and PM2.5 from heavy-duty gasoline vehicles in MOVES for the control
scenarios. In this rulemaking, we are not proposing off-cycle standards for vehicles fueled by
gasoline or NG. Furthermore, even though we anticipate emission benefits from the lengthened
warranty and useful life periods from gasoline and NG-fueled vehicles, they were not included in
the current analysis.
The proposed FTP duty-cycle standards shown in Table 5-4 apply to both heavy-duty
compression-ignition engines and heavy-duty spark-ignition engines.™™ For the control
scenarios, we updated the NOx exhaust emission rates for gasoline, assuming that emissions are
reduced for all operating modes based on the reduction in the NOx FTP standards from the
current 0.2 g/hp-hr standard. Table 5-20 shows the estimated reduction in NOx emission rates,
which is consistent with the ratio of the current FTP emission standards and the proposed FTP
standards shown in Table 5-6.
In addition to modeling the proposed standards for NOx, we estimated emission rate
reductions due to the proposed standards for HC, CO and PM2.5. As discussed in the Preamble
section III.D, the proposed emissions standards for HC, CO, and PM2.5 heavy-duty spark ignition
are lower than the current MY 2010 standards. We estimated reduced THC and CO emission
rates assuming that those emissions would be reduced due to improvements in the three-way
catalyst emission controls. We used initial data from our baseline testing and from the heavy-
duty gasoline technology demonstration program presented in Chapter 3.2 to estimate our
modeled emissions levels. We assumed a 65 percent reduction in THC emissions would occur at
a NOx standard of 0.1 g/hp-hr.1111 We assumed additional decreases in THC emissions to reflect
tighter proposed NOx standards in proposed Option 1 in MY 2031 and the Alternative in MY
2027. We derived Equation 5-22 assuming a linear decrease in THC emissions between the
estimated THC emissions emitted at the 0.1 g/hp-hr NOx FTP level, and zero THC emissions at
a hypothetical 0 g/hp-hr NOx FTP level. We then used Equation 5-22 to estimate the reductions
in THC emissions using the NOx levels for the Proposed Option 1 MY 2027-2030, Proposed
Option 1 MY 2031+, Proposed Option 2, and the Alternative scenarios (Table 5-20).
11111111 Our inventory analysis for HD SI engines only evaluated the impact of the proposed FTP duty-cycle standards.
We did not analyze the impact of our proposed SET duty-cycle standards or idle provisions for HD SI engines, but
will consider including them in the final rulemaking analysis.
1111 The proposal analyzed for the air quality modeling assumed an FTP standard at 0.1 g/hp-hr (Table 5-27)
240
-------
Equation 5-22
n ^ INOx FTP Standard \ ^ n ^
Rgasoiine,THC,NOx FTP = 1 — n X _ ^gasoline,THC,0.1 NOx FTP )
\ ®'^bhp-hr J
= 1 - (NOx ^ stgandard\ x (x _ 65%)
\ " bhp hr J
Where:
Rgasoiine, 77/c,NOx ftp = percent emission reductions in heavy-duty gasoline THC emissions for NOx FTP standards
more stringent than the 0.1 NOx FTP standard, calculated values shown in Table 5-20
NOx FTP Standard = NOx FTP standards in the Proposed and Alternative scenarios
We are proposing a single CO standard for MY 2027 and later HD SI engines and we
maintained a 60 percent reduction in CO for all scenarios (see Table 5-20). To meet the proposed
PM standards, manufacturers are expected to improve fuel control and limit the need for catalyst
protection. We assumed a 50 percent reduction in PM2.5, consistent with the proposed 50 percent
lower PM standard, for all scenarios. Table 5-20 contains the emission rate reductions, Rgasoiine-.
applied in MOVES for the emission inventory analysis.
Table 5-20: Running Emission Rate Reductions From Heavy-duty Gasoline Vehicles Due to Proposed and
Alternative Standards, Rg^oiine, Across All Heavy-duty Gasoline Regulatory Classes and Operating Modes
Control
Scenario
Model
Years
Regulatory
ClasssA
FTP/SET NOx standard
(g/hp-hr)
NOx
THC
CO
PM25
Proposed
Option 1
2027-
2030
LHD,
MHD,
HHD
0.035
82.5%
87.8%
60%
50%
2031+
LHD,
MHD,
HHD
0.02
90.0%
93.0%
60%
50%
Proposed
Option 2
2027+
LHD,
MHD,
HHD
0.05
75.0%
82.5%
60%
50%
Alternative
2027+
LHD,
MHD,
HHD
0.02
90.0%
93.0%
60%
50%
A The proposed and alternative spark-ignition engine standards are the same for LHD, MHD, and HHD
engines, unlike the proposed Option 1 standards for compression-ignition engines (Preamble Section III.2)
We used Equation 5-23 to estimate the MOVES NOx emission rates for the control scenarios
using the Rgasoiine values for heavy-duty gasoline vehicles. Since we are not proposing to require
an in-use testing program for spark-ignition engines, we did not estimate operating mode-
specific effectiveness of reductions of the in-use emissions compared to duty-cycle standard
emissions, as was done for diesel running emissions. Instead, we assumed these reductions apply
uniformly across all running exhaust operating modes. As such, we used Equation 5-23 to
estimate the MOVES emission rates proportionally for all operating modes.
241
-------
Equation 5-23
ERcontrol ~ (l — ^gasoline) ^ ^^MOVES_baseline
Where:
ERcontmi = MOVES running exhaust emission rates for the control scenarios based on the reduction in the FTP duty-
cycle standard
Rgasoime = percent emission reductions in heavy-duty gasoline emissions from Table 5-20
ERmoies baseline = MOVES running exhaust emission rates for the baseline
The estimated heavy-duty gasoline MOVES running emission rates for the Baseline, proposed
Options 1 and 2, and the Alternative scenarios are shown for NOx and THC emissions in Figure
5-11 and Figure 5-12, respectively. CO and PM2.5 were similarly estimated from the reductions
shown in Table 5-20, but they have the same emission rates within each regulatory class for all
the control scenarios and, therefore, are not plotted.
100
' 90
)
80
70
60
50
40
30
20
10
0
Base Rate
Option 2 MY 2027+
Option 1 MY 2027-2030 "Option 1 MY 2031 +
Alternative MY 2027+
-
1
¦
1.
L
_
1.
1.
1,
1.
i,
_
_
L.
1.
1.
Lit
1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES Operating Mode
Figure 5-11: Duty-cycle-based running NOx emission rates for LHD gasoline for the control scenarios
242
-------
-C
s
aj
ro
tr
c
o
0
1
F
w
LU
>
O
50
45
40
35
30
25
20
15
10
5
0
i Base Rate
Option 2 MY 2027+
Option 1 MY 2027-2030 "Option 1 MY 2031 +
Alternative MY 2027+
1
i
1
,|
1
1
1
L.
L
a
ii
i
1
i.
1.
i
L
Lx
bL.
1
1
1
1.
ii
1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES Operating Mode
Figure 5-12: LHD gasoline Duty-cycle-based running THC emission rates for LHD gasoline for the control
scenarios
5.2.2.7 Heavy-Duty Gasoline Refueling Emission Rates
In this section, we describe the methods used to estimate lower refueling emission rates in
MOVES for the control scenarios due to the proposed Onboard Refueling Vapor Recovery
(ORVR) requirements. Refueling emissions result when the pumped gasoline displaces the vapor
in the vehicle tank. The THC emissions are a function of temperature and the gasoline Reid
Vapor Pressure (RVP).JJJJ The emissions control technology which collects the vapor from the
refueling events is the ORVR system. ORVR requirements for light-duty vehicles started
phasing in as part of EPA's Refueling Emission Regulations for Light-Duty Vehicles and Light-
Duty Trucks Final Rule318 in 1998. Under the EPA's Tier 2 vehicle program, all complete
vehicles with a gross vehicle weight rating (GVWR) of 8,500 lbs up to 14,000 lbs (MOVES
regulatory class LHD2b3) were required to meet the ORVR requirements between 2004 and
2006 model years.319 With the Tier 3 rulemaking, all heavy-duty trucks up to 14,000 lbs and all
complete vehicles greater than 14,000 lbs are required to meet a refueling standard of 0.2 grams
of HC per gallon of gasoline dispensed by 2022.320
Table 5-21 shows the ORVR adoption rates applied to all heavy-duty gasoline trucks in
MOVES. For the baseline scenario, we estimated that all heavy-duty gasoline trucks up to
14,000 lbs will have ORVR control by 2018 (as shown in Table 5-21). No heavy-duty gasoline
vehicles over 14,000 lb GVWR are being certified todayKKKk as complete vehicles, and our
See additional discussion of refueling updates in the Updates to MOVES for Emissions Analysis of the HD 2027
NPRM3"
kkkk \\;e expect that only one complete vehicle model will exist in 2022 and it is not yet certified.
243
-------
baseline scenario reflects a population of 100 percent incomplete vehicles that have not adopted
ORVR technologies and are not expected to adopt ORVR without a regulatory action, due to the
costs and added complexities.
As part of this rulemaking, we are proposing that all heavy-duty gasoline vehicles, including
those sold as incomplete vehicles, be required to have an ORVR system and be certified to the
same standard as light-duty by model year 2027. For all three control scenarios modeled for the
proposed rule, we assumed manufacturers would fully implement ORVR technologies in 2027 in
the over 14,000 GVWR vehicle categories as shown in Table 5-21. For a 100 percent phase-in of
ORVR, we estimated a 98 percent reduction in refueling emissions, because we assume that
some ORVR systems would fail or may not be fully effective, similar to our assumptions made
for current ORVR systems in light-duty vehicles.311 The emissions inventory impact of the
proposed ORVR control is summarized in Chapter 5.3.3.
Table 5-21: Phase-In of Onboard Refueling Vapor Recovery (ORVR) for Heavy-duty Trucks
Model Year
Light Heavy-Duty Trucks
8,500-10,000 lbs GVWR
Heavy-Duty Trucks
10,000-14,000 lbs GVWR
Heavy-Duty Trucks
> 14,000 lbs GVWR
(LHD45 and MHD)
(Class 2b)
(Class 3)
including incompletes
Baseline
Control
2003 and earlier
0%
0%
0%
0%
2004
40%
0%
0%
0%
2005
80%
0%
0%
0%
2006-2017
100%
0%
0%
0%
2018-2026
100%
100%
0%
0%
2027 and newer
100%
100%
0%
100%
5.3 National Emissions Inventory Results
In the following sections, we present the emissions impacts of the control scenarios (proposed
options and alternative) in three select calendar years.LLLL The national (50 states and
Washington DC, excluding Puerto Rico and the Virgin Islands) highway heavy-duty vehicle
emission inventory was generated using the national-scale option in MOVES CTINPRM with
the methodology and the model inputs as described in Chapter 5.2.
LLLL The proposed criteria pollutant program is expected to have minimal impacts on C02 emissions. We estimated a
small fuel savings for heavy-duty gasoline vehicles in Section 7.2.2 due to ORVR control (Chapter 5.2.2.7).
However, for MOVES emissions inventories, we estimated no differences in the CO2 emission rates for the baseline
and control scenarios. See draft RIA Chapter 1 for more discussion of the technologies evaluated to control NOx
emissions without impacting CO2 emissions. See Section XI for our analysis of CO2 emission impacts of the
proposed revisions to the Heavy Duty GHG Phase 2 program.
244
-------
5.3.1 Proposed Options
Table 5-22 summarizes the emission impacts of the proposed Option 1 for three select
calendar years. Chapter 5.5.5 shows NOx, VOC, PM2.5, and CO inventories for calendar years
between 2027 and 2045.
Table 5-22: National Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,2040, and 2045
— Proposed Option 1 Program Emissions Relative to Heavy-Duty Vehicle Emissions Baseline
2030
2040
2045
Pollutant
US Short
%
US Short
%
US Short Tons
%
Tons
Reduction
Tons
Reduction
Reduction
NOx
153,608
16.4%
491,318
55.9%
558,780
60.5%
VOC
4,681
5.0%
15,199
18.7%
17,975
21.0%
Primary Exhaust PM2.5
- Total
408
3.4%
1,741
23.7%
2,005
26.4%
Carbon Monoxide (CO)
51,154
3.2%
241,974
15.2%
289,835
17.2%
1,3-Butadiene*
0
0.0%
0
0.0%
0
0.0%
Acetaldehyde
8
0.4%
46
2.5%
52
2.7%
Benzene
42
4.1%
181
23.1%
221
26.8%
Formaldehyde
12
0.5%
63
4.1%
75
4.6%
Methane (CH4)
166
0.2%
881
0.7%
1,025
0.7%
Naphthalene
1.3
0.9%
6.5
14.3%
8
16.7%
* No change is observed in 1,3-butadiene emissions in the control scenarios, because, in MOVES CTI NPRM, 1,3-
butadiene emissions do not contribute to VOC emissions from 2027 and later diesel running and start, heavy-duty
gasoline running, and gasoline refueling311321
More details about the impacts of the proposed Option 1 can be found in Chapter 5.5.3 where
the emission reductions are categorized by vehicle fuel type with further splits by emission
process and by heavy-duty regulatory class. Additional details are also provided in Chapter 5.5.4
regarding the NOx emissions in calendar year 2045 from different engine operational processes
with and without the proposed Option 1 or 2 standards. Table 5-23 summarizes the emissions
impacts in three selected calendar years for proposed Option 2.
245
-------
Table 5-23: National Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,2040, and 2045
— Proposed Option 2 Program Emissions Relative to Heavy-Duty Vehicle Emissions Baseline
2030
2040
2045
Pollutant
US Short
%
US Short
%
US Short
%
Tons
Reduction
Tons
Reduction
Tons
Reduction
NOX
140,691
15.0%
383,350
43.6%
437,869
47.4%
VOC
4,645
5.0%
14,623
18.0%
17,283
20.2%
Primary Exhaust PM2 5 -
408
3.4%
1,600
21.8%
1,856
24.4%
Total
Carbon Monoxide (CO)
51,154
3.2%
216,413
13.6%
262,574
15.6%
1,3-Butadiene
0
0.0%
0
0.0%
0
0.0%
Acetaldehyde
8
0.4%
32
1.8%
37
1.9%
Benzene
41
4.0%
167
21.3%
202
24.5%
Formaldehyde
12
0.5%
51
3.3%
61
3.7%
Methane (CH4)
160
0.1%
654
0.5%
770
0.6%
Naphthalene
1.2
0.8%
5.7
12.6%
7
14.6%
5.3.2 Alternative
Table 5-24 summarizes the emissions impacts in three selected calendar years for the
Alternative scenario.
Table 5-24: National Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,2040, and 2045
— Alternative Program Emissions Relative to Heavy-Duty Vehicle Emissions Baseline
CY2030
CY2040
CY2045
Pollutant
US Short
%
US Short
%
US Short
%
Tons
Reduction
Tons
Reduction
Tons
Reduction
NOX
155,954
16.7%
500,367
56.9%
566,100
61.3%
VOC
4,716
5.0%
15,312
18.9%
18,069
21.1%
Primary Exhaust PM2 5 -
408
3.4%
1,822
24.8%
2,090
27.5%
Total
Carbon Monoxide (CO)
51,154
3.2%
247,475
15.5%
295,561
17.5%
1,3-Butadiene
0
0.0%
0
0.0%
0
0.0%
Acetaldehyde
9
0.4%
49
2.7%
56
2.9%
Benzene
44
4.3%
183
23.3%
222
26.9%
Formaldehyde
13
0.6%
66
4.3%
78
4.7%
Methane (CH4)
172
0.2%
934
0.7%
1,076
0.8%
Naphthalene
1.4
0.9%
6.6
14.6%
8
16.9%
5.3.3 Impacts of Heavy-Duty Gasoline Refueling Controls
Table 5-25 shows the estimated impact on refueling emissions from heavy-duty vehicles due
to the proposed refueling emission standard. For heavy-duty vehicles, MOVES CTINPRM only
estimates refueling emissions from gasoline-fueled vehicles. Thus, the reductions reflect the
control of the refueling emissions from only the heavy-duty gasoline vehicles above 14,000 lbs.
Because benzene is calculated as a fraction of VOC emissions, the percent reductions are the
same for both pollutants as shown in Table 5-25.
246
-------
Table 5-25: Emission Reductions Due to Adoption of ORVR for Heavy-duty Vehicles Relative to Heavy-Duty
Vehicle Emissions Baseline
Calendar
Year
Pollutant
Reductions in US
Short Tons
% Reduction
2027
Benzene
3
7.0%
VOC
955
7.0%
2030
Benzene
14
29.2%
VOC
3980
29.2%
2040
Benzene
41
80.2%
VOC
11637
80.2%
2045
Benzene
48
88.5%
VOC
13678
88.5%
5.4 Emissions Inventories for Air Quality Modeling
The air quality modeling analysis completed for this proposal requires emission inventories
with greater geographical and temporal resolution than the national-scale inventories described
in Chapter 5.3. The approach for estimating emission inventories for an air quality modeling
analysis is extremely complex and time and resource intensive since it involves modeling of each
12 km grid cell, for each hour of the day, for the entire year.
The methodology for developing emission inventories for air quality modeling, also referred
to as SMOKE-MOVES emission inventories, is summarized here. Additional details, including
information for sectors other than onroad vehicles, are available in the Air Quality Modeling
Technical Support Document (AQM TSD).322 Figure 5-13 illustrates the process involved in
generating the onroad emissions inventories for use in air quality modeling. First, the
meteorological preprocessor, Met4Moves, is used to generate the temperature ranges, relative
humidity, and temperature profiles that are needed for MOVES CTINPRM is used to generate
county-level onroad emission factors (EF) by temperature and speed bins. As discussed below,
the MOVES emission rates for the air quality modeling inventory differed from the emission
rates discussed in Chapter 5.2.2. Additionally, other MOVES data inputs for the county-level
runs sometimes differ from the inputs used for the national inventory. For example, the county-
level MOVES runs include county-specific inputs including: fuel programs, inspection and
maintenance programs, adoption of LEV standards, and the age distribution of the local vehicle
fleet. The emission factors for each representative county were generated with multiple runs of
the MOVES CTI NPRM version of the model.
The MOVES-generated onroad emission factors were then combined with activity data to
produce emissions within the Sparse Matrix Operator Kernel Emissions (SMOKE) modeling
system. The collection of tools that compute the onroad mobile source emissions (as one of the
sectors included in the air quality modeling) are known as SMOKE-MOVES. SMOKE-MOVES
uses a combination of vehicle activity data, emission factors from MOVES, meteorology data,
and temporal allocation information to estimate hourly onroad emissions. Additional types of
ancillary data are used for the emissions processing, such as spatial surrogates which spatially
allocate emissions to the 12 km grid cells used for air quality modeling.
247
-------
Use representative
EFs and county/
grid-specific
activity data and
meteorology to
create emissions
for all counties
i v.nipv.iULui v i cii iyv.j,
rel. humidity
Run MOVES to get
emission factors
(EF) for
representative
counties for each
temperature and
speed needed
SMOKE
-0-
Activity
Data
AQ model-ready files
Figure 5-13: Modeling process of onroad emissions as part of the input for air quality modeling
5.4.1 Control Scenario Evaluated for the Air Quality Modelin2 Analysis
The control scenario evaluated for the air quality modeling analysis contains differences from
the proposed Option 1 that is outlined in Preamble Sections III and IV and used to develop the
national emissions inventories discussed in Chapter 5.2.2. Because the air quality modeling
analysis is a time and resource intensive process, the work was conducted using a preliminary
control case that differed from the proposal. We used the same methodology as documented in
Chapter 5.2.2 to develop the running, start, and extended idle emission rates for the air quality
modeling analysis, but different standards, emission warranty, useful life periods, and phase-in
schedules. In addition, we did not account for the closed crankcase control in the air quality
analysis. There are no differences in the refueling control in the proposed Option 1 and the
scenario analyzed for air quality modeling. Table 5-26 summarizes the differences between the
MOVES inputs and emission rates between proposed Option 1 and the air quality modeling
scenario, including the directional impact for the air quality modeling scenario emissions as
compared to proposed Option 1. The net impact of the differences in the control scenarios and
the differences in the emission inventory methodology are presented in Chapter 5.4.3.
248
-------
Table 5-26. Summary of Differences between Emissions in Proposed Option 1 and the Control Scenario
Analyzed for Air Quality Modeling
Program Component
Fuel Types
Processes
Directional Impact for the Air
Quality Modeling Scenario
compared to Option 1
Duty cycle and off-cycle
standards
Diesel
Running, Start, Extended
Idle
Increase in NOx emissions
Gasoline
Running, Start
Increase in NOx and THC
emissions
Warranty Periods
Diesel
Running
Slight increase in NOx, THC, CO,
and PM2 5 emissions
Useful Life Periods
Diesel
Running
Slight decrease in NOx, THC, CO,
and PM2 5 emissions
Crankcase Control
Diesel
Crankcase
Increase in PM2 5 emissions
Delayed phase-in (2031 vs.
2030)
Diesel
Running, Start, Extended
Idle
Decrease in NOx, THC, CO, and
PM2.5 emissions
Gasoline
Running, Start
Refueling Standard
Gasoline
Refueling
No change
Table 5-27 compares the differences in the duty-cycle standards used in developing the
running, start, and extended idle emission rates in the proposed Option 1 and the scenario
analyzed for the air quality modeling analysis.
Table 5-27: Duty-Cycle NOx Standards for Proposed Option 1 and the Control Scenario Analyzed for Air
Quality ModelingA
Model Year
Engine
Duty
Cycle
Duty-Cycle NOx Standards (mg/hp-hr)
Scenario Analyzed for
Air Quality Modeling
Option 1
2027
HHD, MHD,
LHD
FTP
100
35
SET
50
35
LLC
200
90
Idle0
18 g/hr
5 g/hr
HD SI
FTP
100
35
SET
50
35
2030 or 2031A
HHD, MHD,
LHD
FTP
50
20
[40 HHD]B
SET
20
20
[40 HHD]B
LLC
100
50
[ 100 HHD]B
Idle0
10 g/hr
5 g/hr
HD SI
FTP
50
20
SET
20
20
A The MY 2030 applies to the scenario for air quality modeling while MY 2031 applies to proposed Option 1.
B Values in brackets [ ] denote standards that only apply to HHD engines.
c In both scenarios, we assumed compliance with the Voluntary Idle standard which is more stringent than the off-
cycle standard as discussed in Chapter 5.2.2.1.1.2 .
In the control scenario analyzed for air quality modeling, the FTP and SET standards are
different from one another. The Rduty values calculated from the FTP are applied to MOVES
249
-------
running operating modes for vehicle speeds below 50 miles per hour, which aligns with the
transient behavior of the FTP cycle. The i?duty from the SET standard are applied to MOVES
operating modes above 50 mph (operating mode 33 and above), which aligns with the high-
speed activity that is targeted with the SET standard. In the proposed options and the alternative,
the FTP and SET standards are equivalent, so we used the same i?duty values to calculate the
emission rates for all running operating modes as discussed in Section 5.2.2.1.1.1.
Table 5-28. Rduty Ratios Calculated for the Control Scenario Analyzed for Air Quality Modeling
Scenario
Applicable Model
Years
Emission standard
Rduty
(g/hp-hr)
FTP
SET-RMC
FTP
SET-RMC
Air Quality Modeling
Model Year 2027-2029
0.1
0.05
50%
25%
Model Year 2030+
0.05
0.02
25%
10%
The differences in the warranty and useful life periods used in the control scenario analyzed
for air quality modeling are shown in Table 5-29 and Table 5-30. Aside from the differences in
the warranty and useful life periods, we used the same methodology to estimate the impact on
diesel running emission rates. In general, the warranty periods are longer in proposed Option 1
compared to the scenario analyzed for air quality modeling. Furthermore, Option 1 has shorter
useful life periods than in the control scenario analyzed for air quality modeling.
Table 5-29: Warranty Mileages and Years in Option 1 and the Control Scenario Analyzed for Air Quality
Modeling
Model
Year
Engine
Warranty Mileage
Warranty Years
Air Quality
Modeling
Control
Scenario
Option 1
Air Quality
Modeling
Control
Scenario
Option 1
2027
HHD
350k
450k
5 years
7 years
MHD
150k
220k
LHD
110k
150k
HD SI
110k
200k
5 years
7 years
2030 or
2031A
HHD
600k
600k
7 years
10 years
MHD
260k
280k
LHD
200k
210k
HD SI
160k
160k
7 years
10 years
A The MY 2030 applies to the air quality modeling control scenario while MY 2031 applies to Option 1.
250
-------
Table 5-30:Useful Life Mileages and Years in Option 1 and the Control Scenario Analyzed for Air Quality
Modeling
Model
Year
Engine
Useful Life Mileage
Useful Life Years
Air Quality
Modeling
Control Scenario
Option 1
Air Quality
Modeling
Control Scenario
Option 1
2027
HHD
650k
600k
10 years
11 years
MHD
325k
270k
LHD
250k
190k
12 years
HD SI
200k
155k
10 years
12 years
2030 or
2031A
HHD
850k
800k
10 years
12 years
MHD
450k
350k
LHD
350k
270k
15 years
HD SI
200k
200k
10 years
15 years
A The MY 2030 applies to the air quality modeling control scenario while MY 2031 applies to Option 1.
5.4.2 Estimated Differences in the Emission Reductions between the SMOKE-MOVES
and National-Scale Emission Inventories
Table 5-31 presents the emission reductions from all onroad vehicles (both light-duty and
heavy-duty) estimated from the SMOKE-MOVES inventories; for comparison, the table also
includes the estimated reductions from the national-scale inventories using the MOVES inputs
developed for the air quality modeling control scenario. Comparison between the SMOKE-
MOVES inventories and national-scale inventories shows that the percent emission reduction
trends are generally similar, although the SMOKE-MOVES inventory reductions are smaller
relative to the national-scale inventories. There are several factors that contribute to the
differences between the SMOKE-MOVES inventories and the national inventories. The most
important factor is the methodological differences — SMOKE-MOVES utilizes more localized
input data (such as local vehicle age distributions), and local meteorology data — but there is
also a domain difference (48 contiguous states for the SMOKE-MOVES inventories and 50
states for the national inventories). As a result, we also present the SMOKE-MOVES inventories
for the 50 states to compare to the national emissions inventory in Table 5-31.
251
-------
Table 5-31: Onroad Vehicle Emission Reductions from the Air Quality Modeling Control Scenario Using
SMOKE-MOVES Inventories and National Inventories
Pollutant
CY2045 Reduction in
SMOKE-MOVES
Inventory
(Contiguous US - 48
states; used in AQ
modelingA)
CY2045 Reduction in
SMOKE-MOVES
Inventory (50 states)
CY2045 Reduction in
MOVES National
Inventory (50 states)
US Short
Tons
%
Reduction
US Short
Tons
%
Reduction
US Short
Tons
%
Reduction
NOx
447,436
48.1%
449,408
48.0%
539,468
52.7%
VOC
7,756
1.7%
7,854
1.7%
17,750
3.4%
PM2.5 - Primary
544
1.4%
548
1.4%
746
1.8%
Carbon Monoxide (CO)
165,973
3.6%
167,241
3.6%
289,756
5.9%
1,3-Butadiene*
0
0.0%
0
0.0%
0
0.0%
Acetaldehyde
35
0.9%
35
0.9%
51
1.1%
Benzene
112
1.6%
113
1.6%
210
2.8%
Formaldehyde
45
1.6%
46
1.6%
72
2.4%
Naphthalene
4
1.5%
4
1.5%
7
2.4%
Notes: The 48-state SMOKE-MOVES Emissions Inventory was used as emissions inputs to CMAQ, which
currently has a domain that only covers the contiguous 48 states.
* No change is observed in 1,3-butadiene emissions in the control scenarios, because in MOVES CTI NPRM,
1.3 -butadiene emissions do not contribute to VOC emissions from 2027 and later diesel running and start, heavy-
duty gasoline running, and gasoline refueling.311
5.4.3 Estimated Differences in the Emission Reductions between Option 1 and the
Control Scenario Analyzed for Air Quality Modeling
Table 5-32 contains a comparison of the national emissions inventory reduction for proposed
Option 1 compared to the control scenario analyzed for air quality modeling, for all onroad
vehicles (both light-duty and heavy-duty). As discussed in Chapter 5.4.1, we anticipate larger
reduction in NOx emissions from Option 1 compared to what was analyzed for air quality
modeling, due to the lower NOx duty-cycle emission standards. In addition, we anticipate larger
reductions in PM2.5 emissions because Option 1 accounts for crankcase control. Slightly larger
reductions are observed in VOC emissions, which is attributed to the lower gasoline tailpipe
THC emissions with the more stringent NOx emission standards. Small changes are observed for
CO emissions due to the differences in the emissions warranty and useful life standards.
252
-------
Table 5-32 Comparison of the Onroad Vehicle Emission Reductions from the Air Quality Modeling Control
Scenario vs. Option 1
CY2045 Reduction in MOVES
National Inventory (50 states) using
CY2045 Reduction in MOVES National
Pollutant
the Air Quality Modeling Control
Scenario
Inventory (50 states) from Option 1
US Short Tons
% Reduction
US Short Tons
% Reduction
NOx
539,468
52.7%
558,780
54.6%
VOC
17,750
3.4%
17,975
3.5%
PM2.5 - PrimaryA
746
1.8%
2,005
4.8%
Carbon Monoxide
289,756
5.9%
5.9%
(CO)
289,835
1,3-Butadiene*
0
0.0%
0
0.0%
Acetaldehyde
51
1.1%
52
1.2%
Benzene
210
2.8%
221
3.0%
Formaldehyde
72
2.4%
75
2.5%
Naphthalene
7
2.4%
8.0
2.5%
Notes: A The PM2 5 reductions in the Option 1 include additional reductions from accounting for the closed
crankcase requirement for HD diesel vehicles.
* No change is observed in 1,3 -butadiene emissions in the control scenarios, because in MOVES CTINPRM, 1,3-
butadiene emissions do not contribute to VOC emissions from 2027 and later diesel running and start, heavy-duty
gasoline running, and gasoline refueling.311
5.5 Chapter 5 Appendix
5.5.1 Estimation of Engine-Certified Fraction among Model Year 2027 Diesel-Fueled
LHD2b3 Vehicles
We estimated the fraction of engine-certified diesel vehicles in MY 2027 using a combination
of the vehicle activity data in MOVES CTI NPRM and national vehicle registration data. The
vehicle population in MOVES CTI NPRM for current and future years, detailed by regulatory
class, fuel type, vehicle age, and MOVES vehicle activity type, are derived from vehicle
registration data and sales and fleet population projections as documented in a MOVES technical
report.309 Table 5-33 contains the national sales volumes of LHD2b3 diesel-fueled vehicles
estimated for MY 2027 among four MOVES vehicle activity types referred to as source types.
All Class 3 vehicles are classified in the vocational source types (single-unit short-haul and
single-unit long-haul trucks).
Table 5-33: Sales Volumes of Model Year 2027 LHD2b3 Diesel-Fueled Vehicles Estimated by MOVES CTI
NPRM
Source Type
Sales Volume
Passenger Truck
95,308
Light-Commercial Truck
15,820
Single Unit Short-haul Truck
113,959
Single Unit Long-haul Truck
5,026
Total
230,114
Using the sales volume in Table 5-33, we estimated the fraction of Class 3 vocational diesel
vehicles of the total diesel LHD2b3 sales with Equation 5-24. We first estimated the fraction of
253
-------
Class 3 vocational vehicles, because we assume that most of the engine-certified vehicles exist
within Class 3 vehicles.
Equation 5-24
MY 2027 Diesel Class 3 vocational
MY 2027 Diesel Class 2b3
Single Unit Shorthaul + Single Unit Longhaul (Table 5-33)
Total
113,959 + 5,026
= 230.114 = 517%
To estimate the fraction of Class 3 diesel vehicles in model year 2027 that are engine-
certified, we estimated the current fraction of engine-certified Class 3 diesel vehicles in the
national fleet using national vehicle registration data. Fiat Chrysler Automobiles (FCA) is
currently the only vehicle manufacturer to assemble engine-certified 2b3 vehicles. FCA engine-
certified vehicles are assembled with Cummins diesel engines. We did not estimate the fraction
of engine-certified vehicles from the manufacturer-submitted production volumes to EPA,
because the use of this data could compromise the sales data that was claimed to be confidential
business information.MMMM Using the 2014 national registration data that was used to develop
the vehicle activity in MOVES CTINPRM, we estimated that 16.1 percent of Class 3 diesel-
fueled vocational vehicles were produced by FCA (across all model years). We then used
Equation 5-25 to estimate the fraction of engine-certified vehicles among all diesel-fueled 2b3
vehicles in MY 2027. Because FCA manufacturers both chassis and engine-certified diesel 2b3
vehicles, we assumed 50 percent of them are engine-certified.
Equation 5-25
MY 2027 Engine certified Diesel Class 3 vehicles
MY 2026 Diesel 2b3 vehicles
/MY 2027 Diesel Class 3 diesel vocational^
~ V MY 2026 Diesel 2b3 )
x
'FCA Class 3 diesel vocational vehicles in CY 2014 \
1
Class 3 diesel vocational vehicles in CY 2014
/Engine certified FCA Class 3 diesel vocational vehicles\
V FCA Class 3 diesel vocational vehicles /
= 51.7% x 16.1% x 50%
= 4.2%
Equation 5-25 yields an estimate fraction of engine-certified vehicles of diesel 2b3 vehicles in
MY 2027 as 4.2 percent. We also assume the fraction of engine-certified diesel 2b3 vehicles will
remain constant across the model years analyzed for the proposed rule (2027-2045 model years).
14141414 See Section XII. A.I of the preamble to this rulemaking for our proposed determinations for confidential
business information.
254
-------
5.5.2 Zero-Mile Emission Rates for the Control Scenarios
The zero-mile NOx emission rates for HHD diesel vehicles in the proposed and alternative scenarios due
to the duty-cycle and off-cycle standards are displayed in Figure 5-5.
Figure 5-14 and Figure 5-15 display the zero-mile NOx emission rates for LHD45 and MHD
diesel vehicles in the proposed and the alternative scenarios.
120
Base Rate
Option 2 MY 2027+
Option 1 MY 2027-2030 "Option 1 MY 2031 +
Alternative MY 2027+
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES Operating Mode
Figure 5-14: Estimated zero-mile emission rates for LHD45 diesel vehicles due to the proposed and
alternative duty-cycle and off-cycle standards
255
-------
120
_C
^100
CD
ro
cr
.1 80
u>
U)
E
LU
X
O
60
w
£ 40
o
20
i Base Rate Option 1 MY 2027-2030 ¦ Option 1 MY 2031 +
Option 2 MY 2027+ ¦ Alternative MY 2027+
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES Operating Mode
Figure 5-15: Estimated zero-mile emission rates for MHD diesel vehicles due to the proposed and alternative
duty-cycle and off-cycle standards
5.5.3 Details of the Emission Impacts of the Proposed Option 1 Program
In this section, we provide details of the national emission reductions from the heavy-duty
vehicles due to the proposed Option I program (previously summarized in Chapter 5.3.1).
256
-------
¦ Running ¦ Start ¦ Extended Idle & APU |
600,000
CY2030 CY2040 CY2045
Figure 5-16: National NOx Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,2040,
and 2045 — Proposed Option 1 Program Reductions from the Baseline for Each Fuel Type Category by
Emission Process
LHD2b3 ¦ LHD45 ¦ MHD ¦ HHD ¦ Urban Bus
£ 400,000
to
z>
~ 300,000
o
o
200,000
en
100,000
I
& J? j? . 0?
CY2030 CY2040 CY2045
Figure 5-17: National NOx Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,2040,
and 2045 — Proposed Option 1 Program Reductions from the Baseline for Each Fuel Type Category by HD
Regulatory Class
257
-------
Running ¦ Start ¦ Extended Idle & APU ¦ Evap ¦ Refueling |
20,000
18,000
16,000
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
I
CY2030
//
CY2040
• <& c$>
J** *
CY2045
Figure 5-18: National VOC Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,2040,
and 2045 — Proposed Option 1 Program Reductions from the Baseline for Each Fuel Type Category by
Emission Process
LHD2b3
LHD45
MHD
HHD
Urban Bus
20,000
18,000
16,000
£ 14,000
£
12,000
™ 10,000
o
"t> 8,000
3
"O
£ 6,000
4,000
2,000
0
I
/ / C#
? <3~
CY20B0
to"
¦
_
& #
CY2040
/
-------
¦ Running ¦ Start ¦ Extended Idle & APU |
2,000
CY2030 CY2040 CY2045
Figure 5-20: National Exhaust PM2.5 Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,
2040, and 2045 — Proposed Option 1 Program Reductions from the Baseline for Each Fuel Type Category by
Emission Process
¦ LHD2b3 ¦ LHD45 ¦ MHD ¦ HHP ¦ Urban Bus |
2,000
CY2030 CY2040 CY2045
Figure 5-21: National Exhaust PM2.5 Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,
2040, and 2045 — Proposed Option 1 Program Reductions from the Baseline for Each Fuel Type Category by
HD Regulatory Class
259
-------
¦ Running ¦ Start ¦ Extended Idle & APU |
250,000
CY2030 CY2040 CY2045
Figure 5-22: National CO Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,2040, and
2045 — Proposed Option 1 Program Reductions from the Baseline for Each Fuel Type Category by Emission
Process
¦ LHD2b3 ¦ LHD45 ¦ MHD ¦ HHP ¦ Urban Bus |
250,000
CY20B0 CY2040 CY204S
Figure 5-23: National CO Emission Reductions from Heavy-duty Vehicles in Calendar Years 2030,2040, and
2045 — Proposed Option 1 Program Reductions from the Baseline for Each Fuel Type Category by HD
Regulatory Class
5.5.4 Onroad Heavv-Dutv NOx Emissions by Engine Operational Process for the
Baseline, Proposed Option 1, and Proposed Option 2 Standards
Figure 5-24 displays the estimated national onroad heavy-duty NOx emissions in 2045 in the
baseline and in the proposed Option 1 and 2 by engine operation process for the MY 2027 and
later fleet impacted by the proposed rule. See Section VI of the preamble for more discussion on
these comparisons.
260
-------
1,000,000
900,000
800,000
700,000
600,000
CO
=>
.E 500,000
x
O
=3
£=
C
<
400,000
300,000
200,000
100,000
0
Extended Idle + APU
Starts
Running, Medium/High-Load
Running, Low-Load
Running, Age Effects
MY 2010-2026 Fleet
Baseline
Proposed Option 1 Proposed Option 2
Figure 5-24: Comparison of Calendar Year 2045 Onroad Heavy-Duty NOx Emissions from Different Engine
Operational Process NNNN for the Baseline, Proposed Option 1, and Proposed Option 2 Standards
5.5.5 Year-Over-Year Criteria Pollutant Emissions for Calendar Years Between 2027
and 2045
In this section, we present MOVES national inventory emissions (for selected criteria
pollutants) across multiple calendar years (2027-2045) for baseline and control scenarios.
The national heavy-duty vehicle emissions inventories are summarized in Table 5-34 through
Table 5-37 below for NOx, VOC, PM2.5 (exhaust), and CO, respectively, for the baseline,
proposed Options 1 and 2, and alternative scenarios. The same results are also displayed
graphically in Figure 5-25 through Figure 5-28.
NNNN In this graph, the "low-load running" emissions refer to the running exhaust emissions (for model year 2027
and later without age effects) from MOVES running operating mode 0, 1, 11-14, 21-24, 33, plus 50% of operating
mode 35 (see Table 5-3 for MOVES operating mode definitions). The remainder are considered "medium/high-load
running" emissions. The contribution of the "Running, Age Effects" was estimated by conducting a series of
MOVES runs with and without running aging effects for the baseline and proposed Options 1 and 2. The MOVES
inputs without the aging effects are available in the rulemaking docket.3"8
261
-------
Table 5-34: National Heavy-duty Vehicle NOx Emissions (Annual US Tons) For Calendar Years Between
2027 and 2045
Calendar Year
Baseline
Proposed Option 1
Proposed Option 2
Alternative
2027
1,043,853
1,008,012
1,011,033
1,007,470
2028
998,254
924,616
930,819
923,500
2029
964,637
851,723
861,225
850,004
2030
935,698
782,090
795,007
779,744
2031
913,916
709,061
732,174
702,657
2032
896,771
642,716
675,748
632,455
2033
879,841
594,289
631,471
564,540
2034
873,529
558,114
599,299
509,977
2035
869,923
522,722
573,172
475,752
2036
870,446
493,595
552,832
448,090
2037
867,129
450,951
531,055
422,827
2038
871,297
418,629
518,199
406,754
2039
875,406
403,131
506,957
392,702
2040
879,258
387,940
495,908
378,891
2041
887,603
379,711
490,737
371,129
2042
895,708
373,391
487,126
365,198
2043
904,248
369,393
485,498
361,553
2044
913,557
366,489
484,953
358,940
2045
923,026
364,245
485,157
356,926
Baseline Proposed Option 1 Proposed Option 2 Alternative
1,200,000
200,000
o
V f f T T V f V T T T y T T "V T y T T
Calendar Year
Figure 5-25: National Heavy-duty Vehicle NOx Emissions (Annual US Tons) For Calendar Years Between
2027 and 2045
262
-------
Table 5-35: National Heavy-Duty Vehicle VOC Emissions (Annual US Tons) For Calendar Years Between
2027 and 2045
Calendar Year
Baseline
Proposed Option 1
Proposed Option 2
Alternative
2027
109,222
108,099
108,107
108,090
2028
101,990
99,707
99,725
99,690
2029
98,801
95,335
95,362
95,309
2030
93,833
89,153
89,189
89,117
2031
90,551
84,635
84,714
84,577
2032
87,467
80,355
80,477
80,277
2033
85,156
76,812
76,961
76,630
2034
83,542
74,014
74,189
73,734
2035
81,916
71,275
71,498
70,982
2036
81,897
70,213
70,483
69,911
2037
79,966
67,269
67,656
67,050
2038
80,680
67,036
67,532
66,896
2039
81,242
66,803
67,340
66,677
2040
81,205
66,006
66,582
65,893
2041
82,137
66,267
66,871
66,159
2042
82,991
66,530
67,158
66,426
2043
83,770
66,780
67,431
66,680
2044
84,572
67,076
67,748
66,980
2045
85,520
67,545
68,237
67,451
^—Baseline ^—Proposed Option 1 ^—Proposed Option 2 ^—Alternative
120,000
110,000
C 100,000
o
on
3
90,000
80,000
70,000
60,000
-------
Table 5-36: National Heavy-duty Vehicle PM2.5 (Exhaust Only) Emissions (Annual US Tons) For Calendar
Years Between 2027 and 2045
Calendar Year
Baseline
Proposed Option 1
Proposed Option 2
Alternative
2027
17,174
17,078
17,078
17,078
2028
14,747
14,550
14,550
14,550
2029
13,412
13,111
13,111
13,111
2030
11,997
11,590
11,590
11,590
2031
10,979
10,374
10,387
10,368
2032
10,078
9,284
9,308
9,272
2033
9,349
8,412
8,432
8,346
2034
8,833
7,763
7,778
7,646
2035
8,400
7,210
7,231
7,058
2036
8,119
6,818
6,844
6,635
2037
7,303
5,857
5,936
5,723
2038
7,332
5,750
5,879
5,663
2039
7,350
5,686
5,821
5,602
2040
7,335
5,593
5,734
5,512
2041
7,381
5,575
5,718
5,494
2042
7,427
5,565
5,709
5,483
2043
7,479
5,568
5,714
5,485
2044
7,539
5,580
5,728
5,496
2045
7,599
5,594
5,744
5,509
Baseline Proposed Option 1 Proposed Option 2 Alternative
18,000
^ ^
Calendar Year
Figure 5-27: National Heavy-duty Vehicle PM2.5 (Exhaust Only) Emissions (Annual US Tons) For Calendar
Years Between 2027 and 2045
264
-------
Table 5-37: National Heavy-Duty Vehicle CO Emissions (Annual US Tons) For Calendar Years Between 2027
and 2045
Calendar Year
Baseline
Proposed Option 1
Proposed Option 2
Alternative
2027
1,731,776
1,719,562
1,719,562
1,719,562
2028
1,665,353
1,640,479
1,640,479
1,640,479
2029
1,649,108
1,611,269
1,611,269
1,611,269
2030
1,622,980
1,571,826
1,571,826
1,571,826
2031
1,607,158
1,537,808
1,540,247
1,535,479
2032
1,585,670
1,498,705
1,503,512
1,494,112
2033
1,579,461
1,471,175
1,476,072
1,457,798
2034
1,573,019
1,444,301
1,449,261
1,422,576
2035
1,569,490
1,419,437
1,427,042
1,398,932
2036
1,580,182
1,410,067
1,420,203
1,390,853
2037
1,575,206
1,381,753
1,399,327
1,369,502
2038
1,581,911
1,366,746
1,391,269
1,360,993
2039
1,592,078
1,363,205
1,388,241
1,357,584
2040
1,593,814
1,351,840
1,377,400
1,346,338
2041
1,616,360
1,362,968
1,388,881
1,357,419
2042
1,633,915
1,370,352
1,396,583
1,364,747
2043
1,650,715
1,377,945
1,404,453
1,372,301
2044
1,668,084
1,386,567
1,413,401
1,380,889
2045
1,686,295
1,396,460
1,423,721
1,390,734
Baseline Proposed Option 1 Proposed Option 2 Alternative
1,800,000
1,700,000
£ 1,600,000
£
LO
3 1,500,000
~ro
c 1,400,000
1,300,000
1,200,000
<3^ <3? <3^ <3^ <3? S'
f f V T y T y y T f f f f ¦(? f f f ¦v
Calendar Year
Figure 5-28: National Heavy-Duty Vehicle CO Emissions (Annual US Tons) For Calendar Years Between
2027 and 2045
265
-------
5.5.6 Sensitivity Analysis for Battery-Electric Vehicles and Fuel Cell Electric Vehicles
As described in Section 1.4.2, we conducted a sensitivity analysis to estimate the potential
impact of battery-electric vehicles (BEV) and fuel cell electric vehicles (FCEV) on the national
emission inventory for the Baseline, proposed Options 1 and 2, and the Alternative.
Table 5-38 presents the fractions of zero tailpipe emission vehicle populations from the EIA's
2018 Annual Energy Outlook and the NREL study, respectively (see Section 1.4.2 for more
discussion on these data sources). Note that the AEO2018 data included both BEV and FCEVs,
while the NREL study only included BEV data.
To obtain the fractions of BEVs (plus FCEVs if applicable) in HD vehicle categories for each
model year, we first summed the number of battery-electric and fuel-cell electric vehicles in each
of the heavy-duty vehicle categories (Buses/Single Unit Trucks/Combination Trucks) by model
year. Then, we divided that number by the total number of new vehicles (using all fuel types) for
the corresponding category by model year.
266
-------
Table 5-38: Fractions of the BEVs and FCEVs Based on AEO2018 Compared with the Fractions of the BEVs
Based on NREL Study in Each of the Heavy-Duty Vehicle Categories by Model Year
MY
Buses
Single Unit Trucks
Combination Trucks
AEO2018
NREL
AEO2018
NREL
AEO2018
NREL
2016
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
2017
0.0000
0.0040
0.0000
0.0034
0.0000
0.0012
2018
0.0058
0.0060
0.0074
0.0045
0.0029
0.0016
2019
0.0060
0.0080
0.0077
0.0060
0.0030
0.0021
2020
0.0063
0.0100
0.0080
0.0079
0.0031
0.0027
2021
0.0065
0.0100
0.0082
0.0105
0.0032
0.0036
2022
0.0068
0.0100
0.0086
0.0138
0.0033
0.0047
2023
0.0071
0.0100
0.0088
0.0180
0.0034
0.0062
2024
0.0072
0.0100
0.0090
0.0235
0.0035
0.0081
2025
0.0074
0.0100
0.0093
0.0305
0.0036
0.0105
2026
0.0077
0.0480
0.0095
0.0392
0.0037
0.0135
2027
0.0079
0.0860
0.0098
0.0499
0.0038
0.0172
2028
0.0081
0.1240
0.0101
0.0628
0.0039
0.0217
2029
0.0084
0.1620
0.0104
0.0780
0.0041
0.0269
2030
0.0086
0.2000
0.0108
0.0953
0.0042
0.0329
2031
0.0089
0.2250
0.0111
0.1144
0.0043
0.0394
2032
0.0091
0.2500
0.0114
0.1347
0.0044
0.0464
2033
0.0094
0.2750
0.0118
0.1553
0.0046
0.0536
2034
0.0097
0.3000
0.0121
0.1756
0.0047
0.0606
2035
0.0100
0.3250
0.0125
0.1947
0.0048
0.0671
2036
0.0103
0.3400
0.0129
0.2120
0.0050
0.0731
2037
0.0106
0.3550
0.0132
0.2272
0.0051
0.0783
2038
0.0109
0.3700
0.0136
0.2401
0.0053
0.0828
2039
0.0112
0.3850
0.0140
0.2508
0.0054
0.0865
2040
0.0116
0.4000
0.0144
0.2595
0.0056
0.0895
2041
0.0119
0.4150
0.0149
0.2665
0.0058
0.0919
2042
0.0123
0.4300
0.0153
0.2720
0.0059
0.0938
2043
0.0127
0.4450
0.0158
0.2762
0.0061
0.0953
2044
0.0130
0.4600
0.0162
0.2795
0.0063
0.0964
2045
0.0134
0.4750
0.0167
0.2821
0.0065
0.0973
2046
0.0138
0.4800
0.0172
0.2840
0.0067
0.0979
2047
0.0142
0.4850
0.0177
0.2855
0.0069
0.0984
2048
0.0147
0.4900
0.0182
0.2866
0.0071
0.0988
2049
0.0151
0.4950
0.0187
0.2874
0.0073
0.0991
2050
0.0155
0.5000
0.0192
0.2881
0.0075
0.0993
The estimated annual percent emission reductions for proposed Option 1 adjusted for
BEV/FCEVs based on AEO2018 and NREL study are shown in Table 5-39.0000 For
comparison, the emission reductions from proposed Option 1 (previously presented in draft RIA
Chapter 5.3) are also included. Similarly, Table 5-40 and Table 5-41 show the estimated annual
percent reductions adjusted for BEV/FCEVs for proposed Option 2 and the Alternative
0000 Although BEVs and FCEVs have brakewear and tirewear particulate emissions, we are not accounting for
them in this sensitivity analysis due to insufficient data.
267
-------
scenarios, respectively. These reductions are compared to the reductions from the main analysis
for proposed Option 2 and the Alternative (previously presented in draft RIA Chapter 5.3).
Table 5-39: Percent Reduction from the Baseline for Proposed Option 1 and Sensitivity Cases Based on
AEO2018 and NREL in Calendar Years 2030,2040, and 2045
Pollutant
% Reduction in 2030
% Reduction in 2040
% Reduction in 2045
Option 1
AEO2018
NREL
Option 1
AEO2018
NREL
Option 1
AEO2018
NREL
NOx
16.4%
16.4%
15.9%
55.9%
55.8%
54.4%
60.5%
60.5%
59.5%
VOC
5.0%
5.0%
4.7%
18.7%
18.6%
17.1%
21.0%
20.9%
19.0%
Primary Exhaust
3.4%
3.4%
3.2%
23.7%
23.7%
22.7%
26.4%
26.3%
25.2%
PM2 5 - Total
Carbon Monoxide
3.2%
3.1%
3.0%
15.2%
15.1%
14.2%
17.2%
17.1%
15.9%
(CO)
1,3-Butadiene
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Acetaldehyde
0.4%
0.4%
0.4%
2.5%
2.5%
2.4%
2.7%
2.7%
2.6%
0.5%
0.5%
0.5%
4.2%
4.2%
4.1%
4.6%
4.6%
4.4%
Benzene
4.1%
4.1%
3.8%
23.1%
23.0%
21.3%
26.8%
26.7%
24.7%
Formaldehyde
0.5%
0.5%
0.5%
4.1%
4.1%
3.9%
4.6%
4.6%
4.4%
Methane (CH4)
0.2%
0.2%
0.1%
0.7%
0.7%
0.7%
0.7%
0.7%
0.7%
Naphthalene
0.9%
0.9%
0.8%
14.3%
14.3%
13.4%
16.7%
16.7%
15.5%
Table 5-40: Percent Reduction from the Baseline for Proposed Option 2 and Sensitivity Cases Based on
AEO2018 and NREL in Calendar Years 2030,2040, and 2045
Pollutant
% Reduction in 2030
% Reduction in 2040
% Reduction in 2045
Option 2
AEO2018
NREL
Option 2
AEO2018
NREL
Option 2
AEO2018
NREL
NOx
15.0%
15.0%
14.6%
43.6%
43.6%
42.5%
47.4%
47.4%
46.7%
VOC
5.0%
4.9%
4.6%
18.0%
17.9%
16.4%
20.2%
20.1%
18.2%
Primary Exhaust
3.4%
3.4%
3.2%
21.8%
21.8%
20.8%
24.4%
24.4%
23.3%
PM2 5 - Total
Carbon Monoxide
3.2%
3.1%
3.0%
13.6%
13.5%
12.6%
15.6%
15.5%
14.2%
(CO)
1,3-Butadiene
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Acetaldehyde
0.4%
0.4%
0.4%
1.8%
1.7%
1.7%
1.9%
1.9%
1.8%
0.5%
0.5%
0.5%
3.1%
3.1%
2.9%
3.4%
3.4%
3.2%
Benzene
4.0%
3.9%
3.7%
21.3%
21.2%
19.7%
24.5%
24.5%
22.6%
Formaldehyde
0.5%
0.5%
0.5%
3.3%
3.3%
3.1%
3.7%
3.7%
3.5%
Methane (CH4)
0.1%
0.1%
0.1%
0.5%
0.5%
0.5%
0.6%
0.6%
0.5%
Naphthalene
0.8%
0.8%
0.8%
12.6%
12.6%
11.8%
14.6%
14.6%
13.5%
268
-------
Table 5-41: Percent Reduction from the Baseline for the Alternative and Sensitivity Cases Based on AEO2018
and NREL in Calendar Years 2030,2040, and 2045
Pollutant
% Reduction in 2030
% Reduction in 2040
% Reduction in 2045
Alternativ
e
AEO201
8
NRE
L
Alternativ
e
AEO201
8
NRE
L
Alternativ
e
AEO201
8
NRE
L
NOX
16.7%
16.6%
16.2
%
56.9%
56.9%
55.6
%
61.3%
61.3%
60.4
%
VOC
5.0%
5.0%
4.7%
18.9%
18.8%
17.2
%
21.1%
21.1%
19.1
%
Primary Exhaust
PM2 5 - Total
3.4%
3.4%
3.2%
24.8%
24.8%
23.8
%
27.5%
27.5%
26.3
%
Carbon
Monoxide
(CO)
3.2%
3.1%
3.0%
15.5%
15.5%
14.5
%
17.5%
17.5%
16.2
%
1,3-Butadiene
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Acetaldehyde
0.4%
0.4%
0.4%
2.7%
2.7%
2.6%
2.9%
2.9%
2.8%
0.5%
0.5%
0.5%
4.5%
4.5%
4.4%
4.9%
4.9%
4.7%
Benzene
4.3%
4.3%
4.0%
23.3%
23.2%
21.6
%
26.9%
26.8%
24.8
%
Formaldehyde
0.6%
0.6%
0.5%
4.3%
4.2%
4.1%
4.7%
4.7%
4.5%
Methane (CH4)
0.2%
0.2%
0.2%
0.7%
0.7%
0.7%
0.8%
0.8%
0.8%
Naphthalene
0.9%
0.9%
0.9%
14.6%
14.6%
13.7
%
16.9%
16.8%
15.7
%
269
-------
5.5.7 National Heavy-duty Vehicle Emissions Inventory Comparison of the California
Heavy-duty Omnibus Regulation and the Proposed Option 1
This section presents a sensitivity analysis of the estimated emission inventory impacts from
two scenarios based on the California Heavy-Duty Omnibus Program (Omnibus). The first
scenario assumes nationwide adoption of the Omnibus starting in MY 2024. The second scenario
assumes nationwide adoption of the MY 2027 and later Omnibus program.pppp We compare
these two scenarios to the EPA proposed Option 1 scenario.QQQQ In summary, the three scenarios
presented include: 1) EPA proposed Option 1, 2) a national program that has the same stringency
level as the Omnibus starting in MY 2024 (Omnibus Nationwide 2024), and 3) Omnibus
Nationwide starting in MY 2027 (Omnibus Nationwide 2027).
The EPA proposed Option 1 largely aligns with the MY 2027 standards, regulatory useful life
periods, and emissions warranty periods in the Omnibus (See Preamble Sections I.D, III, and IV
for details). The results of the sensitivity analysis presented in this Appendix show that the
differences between the Omnibus Nationwide 2027 scenario and EPA proposed Option 1 result
in only slight differences (approximately 0.2%) in NOx emission inventory impacts in 2045,
although there are larger differences in earlier years of the program. Additional details on the
differences between proposed Option 1 and the Omnibus requirements are included in Preamble
Sections III and iv.RRRR
Below, we describe the methods used to analyze the three scenarios included in the sensitivity
analysis, and then present detailed results of the comparison.
5.5.7.1 MOVES input development for Omnibus Nationwide Scenarios
We estimated the emission impacts of the EPA proposed Option 1 and Omnibus Nationwide
scenarios using the same version of MOVES (MOVES CTINPRM); the MOVES inputs for the
EPA proposed Option 1 are described in Chapter 5.2.2. The MOVES inputs for Omnibus
Nationwide scenarios are based on the Omnibus duty-cycle test standards, warranty and useful
life requirements, as detailed in the subsections immediately below.
5.5.7.2 Heavy-duty Diesel Running Emission Rates Based on Duty-Cycle and Off-Cycle
Standards
For estimating NOx running emissions based on duty-cycle test standards for LHD and MHD
vehicles under the Omnibus, we used the standards on the FTP and SET-RMC cycles from the
Omnibus shown in Table 5-42. For HHD vehicles, the duty-cycle test standards in the Omnibus
pppp jjlc Omnibus does not include updated spark-ignition HC and CO emission standards; however, for this
analysis we assumed nationwide adoption of the HC and CO standards for SI engines. See additional discussion of
HD gasoline running exhaust emission rates in this appendix. HD CI engines are already meeting those emissions
levels and we did not model HC and CO reductions for CI engines.
QQQQ As discussed in Section 5.4, there are differences between the proposed Option 1 and the control case analyzed
for air quality modeling. This section (5.5.6) presents emission reductions from the EPA proposed Option 1, which
is discussed in further detail in Section 5.3.
11111111 In addition to the difference in ORVR requirements between our EPA proposed Option 1 and the Omnibus, the
Omnibus Nationwide 2027 scenario includes a slightly lower NOx standard in MY 2027-2030 than proposed Option
11, while proposed Option 1 includes NMHC and CO standards not included in the Omnibus Nationwide 2027
scenario. There are also slight differences between the Omnibus Nationwide 2027 scenario and the proposed Option
1 for duty-cycle and off-cycle standards for SI engines (See Preamble Section III for details).
270
-------
have two separate standards for MY 2027 and later vehicles. The HHD vehicles have a duty-
cycle standard at both an intermediate useful life of 435,000 miles and a full useful life of
600,000 miles for MY 2027-2030. For MY 2031 and later, both the CARB Omnibus program
and the EPA proposed Option 1 have an intermediate useful life of 435,000 miles and a full
useful life of 800,000 miles. For comparison, we also present the duty-cycle NOx standards for
EPA proposed Option 1 in Table 5-42.
Table 5-42: Duty-Cycle NOx Standards for the CARB Omnibus and EPA Proposed Option 1A
Model
Year
Engine
Duty
Cycle
CARB Omnibus
EPA Proposed
Option 1
FTP
50
(200)
HHD, MHD, LHD
SET
50
(200)
2024
LLC
200
-
Idle
10 g/hr
-
HD SI
FTP
50
(200)
SET
-
-
FTP
20 [HHD 20/35]c
35
HHD, MHD, LHD
SET
20 [HHD 20/35]c
35
2027
LLC
50 [HHD 50/90]c
90
IdleB
5 g/hr
5 g/hr
HD SI
FTP
20
35
SET
-
35
FTP
20 [HHD 20/40]c
20 [HHD 20/40]c
HHD, MHD, LHD
SET
20 [HHD 20/40]c
20 [HHD 20/40]c
2031+
LLC
50 [HHD
50/100]°
50 [HHD 50/100]°
IdleB
5 g/hr
5 g/hr
HD SI
FTP
20
20
SET
-
-
A (#) = current standard with no change in the noted model year
B We assumed compliance with the voluntary idle standard
c [HHD intermediate useful life standard/full useful life standard]. The HHD intermediate useful life in the CARB
Omnibus program for MY 2027-2030 is 435,000 miles, and the full useful life is 600,000 miles. For the MY 2031+
model years the HHD intermediate useful life is 435,000 miles and the full useful life is 800,000 miles in both the
CARB Omnibus program and the EPA proposed Option 1.
As was done for the EPA proposed Option 1, we calculated a weighted average HHD duty-
cycle standard from the CARB Omnibus program using the intermediate and full useful life
using Equation 5-1. The resulting values are shown in Table 5-43.
271
-------
Table 5-43: Weighted Average Heavy Heavy-duty Compression Ignition Duty-Cycle Test NOx Standards
used to estimate NOx emission rates for the Omnibus Nationwide Scenarios
Applicable Model Years
LLC
(g/hp-hr)
FTP
(g/hp-hr)
SET-RMC
(g/hp-hr)
Model Year 2024-2026
0.2
0.05
0.05
Model Year 2027-2030
0.061
0.024
0.024
Model Year 2031+
0.073
0.029
0.029
Using the FTP Omnibus duty-cycle test standards for LHD and MHD in Table 5-42 and the
weighted average FTP & SET-RMC duty-cycle standards for HHD in Table 5-43, we applied the
methods outlined in Chapter 5.2.2.1.1.1 to estimate the zero-mile emission running emission
rates based on duty-cycle test standards.
The Omnibus has off-cycle standards with three bins, to represent idle, low-load and medium
to high-load operations, consistent with the EPA proposed program. The off-cycle standards for
the Omnibus program are set to be twice the engine-cycle standards for model year 2024-2029,
and 1.5 times the engine-cycle standards for 2030+ that correspond to similar engine-loads.314
For developing inputs for MOVES, we did not apply a scaling factor to the off-cycle idling
operation, to be consistent with the assumptions used to estimate off-cycle idling emissions from
the EPA proposed Option 1, which assumed that the voluntary EPA idle standard would be
complied with during off-cycle idling operation with no scaling factor (See 5.2.2.1.1.2). For
developing inputs to MOVES, we multiplied the Omnibus duty-cycle standards, including the
weighted average NOx standards from Table 5-43, by the corresponding off-cycle scaling
factors, as shown in Table 5-44.
272
-------
Table 5-44: Calculated Average Off-Cycle Standards for the Omnibus from the Average Idling and Engine
Cycles Standards and Off-Cycle Scaling Factors
Regulatory
Class
Model
Year
Engine
Cycle
Engine
cycle NOx
Standard
In-use Bin
In-use
scaling
factor
Calculated
In-use NOx
Standards
(g/hr for
idling, g/hp-
hr for low-
load and
medium to
high-load)
LHD/MHD/HHD
2024-
2026
Idle (g/hr)
10
Idle, < 6% power
1*
10
LLC (g/hp-
hr)
0.2
Low-load,6-20%
power
2
0.4
FTP & SET
(g/hp-hr)
0.05
Medium to High
Load, >20% power
2
0.1
LHD/MHD
2027-
2030
Idle (g/hr)
5
Idle, < 6% power
1*
5
LLC (g/hp-
hr)
0.05
Low-load,6-20%
power
2
0.1
FTP & SET
(g/hp-hr)
0.02
Medium to High
Load, >20% power
2
0.04
2031+
Idle (g/hr)
5
Idle, < 6% power
1*
5
LLC (g/hp-
hr)
0.05
Low-load,6-20%
power
1.5
0.075
FTP & SET
(g/hp-hr)
0.02
Medium to High
Load, >20% power
1.5
0.03
HHD
2027-
2030
Idle (g/hr)
5
Idle, < 6% power
1*
5
LLC (g/hp-
hr)
0.061
Low-load,6-20%
power
2
0.122
FTP & SET
(g/hp-hr)
0.024
Medium to High
Load, >20% power
2
0.048
2031+
Idle (g/hr)
5
Idle, < 6% power
1*
5
LLC (g/hp-
hr)
0.073
Low-load,6-20%
power
1.5
0.109
FTP & SET
(g/hp-hr)
0.029
Medium to High
Load, >20% power
1.5
0.0435
* The Omnibus includes off-cycle scaling factors of 2 and 1.5 for all off-cycle modes, including idling. For
developing the MOVES inputs, we did not apply the scaling factors for idling, to be consistent with the methods
used to estimate emissions for off-cycle idling operation from the EPA proposed Option 1
We developed MOVES emission rates for the Omnibus scenarios using the methods
discussed in Chapter 5.2.2.1.1 along with the duty-cycle standards (Table 5-42 and Table 5-43)
and off-cycle standards (Table 5-44) discussed in this section.
273
-------
5.5.7.3 Heavy-Duty Diesel Running Emission Rates Based on Changes to Warranty
and Useful Life
The warranty and useful life standards for the Omnibus program are presented in Table 5-45
and Table 5-46. For comparison, we also present the warranty and useful life periods for the EPA
proposed Option 1 which matches the Omnibus program for MY 2027 and later (see Preamble
Section IV for more details on our proposed warranty and useful life standards).ssss
We used the methods outlined in Chapter 5.2.2.1.2 to estimate the impact of the lengthened
warranty and useful life program using the warranty and useful life standards specific to the
Omnibus program. As for the EPA proposal options, we did not account for the impact of the
longer warranty and useful life for heavy-duty gasoline and heavy-duty NG vehicles.
Table 5-45: Warranty Mileages and Years in Omnibus and EPA Proposed Option 1A
Model
Year
Engine
Warranty Mileage
Warranty Years
Omnibus0
Proposed
Option 1
Omnibus0
Proposed
Option 1
2024
HHD
(350k)
(100k)
(5y)
(5 y)
MHD
(150k)
(100k)
LHD
(110k)
(50k)
HD SI
(50k)
(50k)
(5y)
(5 y)
2027
HHD
450k
450k
7 y
7 y
MHD
220k
220k
LHD
150k
150k
HD SI
110k
160k
7 y
7 y
2031
HHD
600k
600k
10 y
10 y
MHD
280k
280k
LHD
210k
210k
HD SI
160k
160k
10 y
10 y
A (#) = current standard with no change in the noted model year
B CARES' Step 1 Warranty program begins in MY 2022
ssss as noted in Preamble Section VI, the warranty and useful requirements included in the Omnibus and the EPA
proposed Option 1 would lower emissions of NOx, exhaust PM, THC, and CO, whereas the standards presented in
Table 5-42 would lower emissions of NOx.
274
-------
Table 5-46: Useful Life Mileages and Years in Omnibus Program and EPA Proposed Option 1A
Model
Year
Engine
Useful Life Mileage
Useful Life Years
Omnibus
EPA
Proposed
Option 1
Omnibus
EPA
Proposed
Option 1
2024
HHD
(435k)
(435k)
(10 y)
(10 y)
MHD
(185k)
(185k)
LHD
(110k)
(110k)
HD SI
(110k)
(110k)
(10 y)
(10 y)
2027
HHD
600k
600k
11 y
11 y
MHD
270k
270k
11 y
11 y
LHD
190k
190k
12 y
12 y
HD SI
155k
155k
12 y
12 y
2031
HHD
800k
800k
12 y
12 y
MHD
350k
350k
12 y
12 y
LHD
270k
270k
15 y
15 y
HD SI
200k
200k
15 y
15 y
A (#) = current standard with no change in the noted model year
5.5.7.4 Heavy-Duty Diesel Start Emission Rates
The heavy-duty diesel start emission rates were developed using the CARB duty-cycle
standards for LHD and MHD in Table 5-42 and the weighted average duty-cycle standards for
HHD in Table 5-43, and the methods outlined in Section 5.2.2.2.
5.5.7.5 Heavy-Duty Diesel Extended Idle Emission Rates
As discussed in Chapter 5.2.2.3, the off-cycle rates for the EPA proposed Option 1 for
running exhaust idle operating mode bin (MOVES operating mode 1) were derived from the
voluntary idle emission standard and converted to g NOx/kg CO2. Similarly, we used the same
method to estimate the extended idle emission rates for the Omnibus program, which has an
idling standard of 10 g/hr for MY 2024-2026 and 5 g/hr for MY 2027 and later. Applying the
extended idle CO2 emission rate in MOVES CTINPRM yielded the extended idle emission rates
forNOx shown in Table 5-47.
Table 5-47: Extended Idle NOx emission rates for the Omnibus Program
Applicable Model Years
NOx Idling
Standard
(g/hr)
MOVES Extended
Idle NOx Emission
Rate (g/hr)
Model Year 2024-2026
10
8.96
Model Year 2027+
5
4.48
5.5.7.6 Light Heavy-Duty Class 2b and 3 Diesel Emission Rates
We assumed that engine-certified light-heavy-duty Class 2b and 3 diesel emission rates in the
Omnibus nationwide scenarios have the same emission rates as the corresponding LHD45
275
-------
emission rates. We also assumed they represent 5.1% of the Class 2b and 3 diesel vehicles in
MY 2024 and later, similar to the assumption made for the EPA proposed Option 1 for MY 2027
and later model years as documented in Chapter 5.2.2.5.
5.5.7.7 Heavy-Duty Gasoline Running Emission Rates
Table 5-48 provides the FTP duty-cycle standards from the Omnibus program for Otto-cycle
engines, which are referred to as spark-ignition engines in the EPA proposed rule. Both heavy-
duty gasoline and heavy-duty NG engines are subject to the Otto-cycle engine standards;
however, we did not assume a reduction in NG emissions from the more stringent standards in
this sensitivity analysis because we did not estimate reductions in NG emissions for the control
scenario analyzed (see Chapter 5.2.2).
We used the methods described in Chapter 5.2.2.6 to estimate the impact of the Omnibus FTP
standard on heavy-duty gasoline emissions as shown in Table 5-48. The Omnibus scenarios
include NOx and PM emission rate reductions for heavy-duty gasoline engines that start in MY
2024 or 2027; we also included CO and HC reductions in these scenarios that are consistent with
the the methodology described in Chapter 5.2.2.6 which assumes lower THC emissions as a
function of lower NOx emission standards. We also assumed 60% lower CO emissions as
consistent with the EPA proposed Options 1 and 2. We note that spark-ignition engine standards
in the Omnibus do not change in MY 2031 and emission rates modeled in both Omnibus
scenarios apply for MY 2027 and later. As discussed in Section III of the Preamble, the Omnibus
is also setting an off-cycle standard for Otto-cycle engines; however, we did not estimate the
impact of the off-cycle standard for Otto-cycle engines in the Omnibus Nationwide scenarios in
this sensitivity analysis. The EPA proposed Options and the Alternative 1 did not include an off-
cycle standard for heavy-duty gasoline engines.
Table 5-48: Running Emission Rate Reductions From Heavy-duty Gasoline Vehicles For the EPA Proposed
Option 1 and Omnibus Nationwide Scenarios
Control
Scenario
Model
Years
Regulatory
Class
FTP/SET
standard
(g/hp-hr)
NOx
THC
CO
PM2.S
Proposed
Option 1
2027-2030
LHD, MHD, HHD
0.035
82.5%
87.8%
60%
50%
2031+
LHD, MHD, HHD
0.02
90.0%
93.0%
60%
50%
Omnibus
Nationwide
2024
2024-2026
LHD, MHD, HHD
0.05
75%
82.50%
60%
50%
2027+
LHD, MHD, HHD
0.02
90%
93%
60%
50%
Omnibus
Nationwide
2027
2027+
LHD, MHD,
HHDA
0.02
90%
93%
60%
50%
5.5.7.8 Heavy-Duty Gasoline Refueling Emission Rates
The Omnibus does not include ORVR requirements. In contrast, the EPA proposed Options
include an ORVR requirement for heavy-duty gasoline vehicles >14,000 lbs. No changes were
made to the baseline refueling vapor emissions in MOVES for modeling the Omnibus
276
-------
Nationwide Scenarios. We expect inclusion of onboard refueling vapor requirements in the EPA
proposed Option 1 results in VOC emission reductions not expected from the Omnibus
Nationwide Scenarios.
5.5.7.9 Comparisons of NOx Emission Reductions
Below, we present national heavy-duty vehicle NOx emission reductions across multiple
calendar years (2024-2045) for the EPA proposed Option 1 and the two Omnibus Nationwide
scenarios.
Table 5-49 National Heavy-duty Vehicle NOx Emission Reductions Relative to the Baseline Case For
Omnibus Nationwide Scenarios (Reductions Relative to EPA Proposed Option 1 Shown for Comparison)
Reductions in US Tons
Omnibus
% Difference:
Calendar
Omnibus
Omnibus
EPA Proposed
Option 1
Nationwide
Omnibus Nationwide
Year
Nationwide
Nationwide
2027-
2027 - EPA Proposed
2024
2027
EPA Proposed
Option 1
Option 1
2024
26,189
0
0
0
0.0%
2025
52,379
0
0
0
0.0%
2026
79,372
0
0
0
0.0%
2027
117,691
37,353
35,841
1,512
4.2%
2028
153,371
76,742
73,638
3,104
4.2%
2029
188,785
117,671
112,914
4,757
4.2%
2030
224,101
160,395
153,608
6,786
4.4%
2031
270,725
211,533
204,855
6,677
3.3%
2032
314,700
260,513
254,055
6,459
2.5%
2033
340,751
291,555
285,553
6,002
2.1%
2034
365,058
320,834
315,415
5,419
1.7%
2035
391,527
352,103
347,201
4,903
1.4%
2036
415,921
381,194
376,851
4,343
1.2%
2037
450,242
420,140
416,178
3,962
1.0%
2038
481,867
456,227
452,667
3,560
0.8%
2039
496,843
475,429
472,275
3,155
0.7%
2040
511,423
494,062
491,318
2,744
0.6%
2041
524,580
510,254
507,893
2,361
0.5%
2042
536,369
524,318
522,318
2,001
0.4%
2043
547,107
536,521
534,855
1,666
0.3%
2044
557,685
548,460
547,067
1,393
0.3%
2045
567,886
559,947
558,780
1,166
0.2%
Total
7,614,572
6,735,251
6,663,280
71,970
1.1%
277
-------
EPA Proposed Option 1 ¦ Omnibus Nationwide 2027 a Omnibus Nationwide 2024
Calendar Year
Figure 5-29: National Heavy-duty Vehicle NOx Emission Reductions (Annual US Tons) Relative to the
Baseline Case For Omnibus Nationwide Scenarios (Symbols) As Compared with EPA Proposed Option 1
(Line)
278
-------
Chapter 6 Air Quality Impacts
This chapter presents information on air quality, including a discussion of current air quality
in Section 6.1, details related to the methodology used for the air quality modeling analysis in
Section 6.2, and results from the air quality modeling analysis which are summarized in Section
6.3. We expect the proposal's reductions in emissions of NOx, VOC, PM2.5. and CO would lead
to decreases in ambient concentrations of ozone, PM2.5, NO2, and CO. We performed air quality
modeling of proposed Option 1 to quantify these impacts.TTTT Specifically, proposed Option 1
would significantly decrease ozone concentrations across the country, with a population-
weighted average decrease of 2 ppb in 2045. Ambient PM2.5, NO2 and CO concentrations are
also predicted to improve in 2045 as a result of proposed Option 1. The emission reductions
provided by the proposed rule would be important in helping areas attain the National Ambient
Air Quality Standards (NAAQS) and prevent future nonattainment. In addition, the air quality
modeling predicts improvements in nitrogen deposition and visibility but relatively little impact
on ambient concentrations of air toxics.
6.1 Current Air Quality
In this section we present information related to current levels of air pollutants, visibility
levels, and deposition amounts. This provides context for the need for this proposed rule and a
comparison for the modeled projections presented in Section 6.3.
6.1.1 Ozone
As described in Chapter 4 of this draft RIA, ozone causes adverse health effects, and EPA has
set national ambient air quality standards (NAAQS) to protect against those health effects. The
primary NAAQS for ozone, established in 2015 and retained in 2020, is an 8-hour standard with
a level of 0.07 ppm.uuuu EPA recently announced that it will reconsider the previous
administration's decision to retain the ozone NAAQS.vvvv EPA is also implementing the
previous 8-hour ozone primary standard, set in 2008 at a level of 0.075 ppm. As of May 31,
2021, there were 34 ozone nonattainment areas for the 2008 primary ozone NAAQS, composed
of 151 full or partial counties, with a population of more than 99 million (see Figure 6-1); there
were 50 ozone nonattainment areas for the 2015 primary ozone NAAQS, composed of 205 full
or partial counties, with a population of more than 122 million (see Figure 6-2). In total, there
were, as of May 31, 2021, 57 ozone nonattainment areas with a population of more than 122
million people.wwww
TTTT As noted in Chapter 5.4 of the draft RIA, while we refer to this modeling as for the proposed Option 1, there are
differences between the proposed Option 1 standards, emission warranty, and useful life provisions presented in
Sections III and IV of the preamble and those included in the control scenario modeled for the air quality analysis.
1,111,11 https://www.epa.gov/ground-level-ozone-pollution/ozone-national-ambient-air-quality-standards-naaqs
ww https://www.epa.gov/ground-level-ozone-pollution/epa-reconsider-previous-administrations-decision-retain-
2015-ozone
wwww j]le total population is calculated by summing, without double counting, the 2008 and 2015 ozone
nonattaimnent populations contained in the Criteria Pollutant Nonattaimnent Summary report
(https://www.epa.gov/green-book/green-book-data-download).
279
-------
8-Hour Ozone Nonattainment Areas (2008 Standard)
05131/2021
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.
8-hour Ozone Classification
Extreme
~ Severe 15
[ | Serious
I | Moderate
I | Marginal
Figure 6-1: 8-Hour Ozone Nonattainment Areas (2008 Standard)
8-Hour Ozone Nonattainment Areas (2015 Standard)
Nonattainment areas are indicated by color.
When only a portion of a county is shown in color,
it indicates that only that part of the county is within
a nonattainment area boundary.
05/31/2021
8-hour Ozone Classification
| Extrerr
Severe-17
I Severe-15
^ Serious
~ Moderate
| Marginal
| Marginal (Rural Transport)
Figure 6-2: 8-Hour Ozone Nonattainment Areas (2015 Standard)
280
-------
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. 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. Attainment dates for areas designated nonattainment for the 2015 ozone NAAQS will
be in the 2021 to 2038 timeframe, again depending on the severity of the problem in each
area.xxxx The proposed rule would begin to take effect in 2027 and would assist areas with
attaining the NAAQS and may relieve areas with already stringent local regulations from some
of the burden associated with adopting additional local controls. The proposed rule would
also provide assistance to counties with ambient concentrations near the level of the NAAQS
who are working to ensure long-term attainment or maintenance of the NAAQS.
6.1.2 PM^s
As described in Chapter 4 of this draft RIA, PM causes adverse health effects, and EPA has
set NAAQS to protect against those health effects. There are two primary NAAQS for PM2.5: an
annual standard (12.0 micrograms per cubic meter ([j,g/m3)) and a 24-hour standard (35 (j,g/m3),
and there are two secondary NAAQS for PM2.5: an annual standard (15.0 jag/ 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, and then retained in 2020. On June 10, 2021, EPA
announced that it will reconsider the previous administration's decision to retain the PM
NAAQS.zzzz
There are many areas of the country that are currently in nonattainment for the annual and 24-
hour primary PM2.5 NAAQS. As of May 31, 2021, more than 19 million people lived in the 4
areas that are designated as nonattainment for the 1997 annual PM2.5 NAAQS. Also, as of May
31, 2021, more than 31 million people lived in the 14 areas that are designated as nonattainment
for the 2006 24-hour PM2.5 NAAQS, and more than 20 million people lived in the 6 areas
designated as nonattainment for the 2012 annual PM2.5 NAAQS. In total, there are currently 17
PM2.5 nonattainment areas with a population of more than 32 million people.AAAAA
Nonattainment areas for the PM2.5 NAAQS are pictured in Figure 6-3.
;;;; lUtps:/Av\v\Y. cpa.gov/ground-lcvcl-o/onc-pol lution/o/onc-naaqs-ti inclines.
yyyy while not quantified in the air quality modeling analysis for this proposed rule, the Early Adoption Incentives
under the proposed program could encourage manufacturers to introduce new emission control technologies prior to
the 2027 model year, which may help to accelerate some benefits of the proposed program (See Preamble Section
IV.H for more details on the proposed Early Adoption Incentives). In addition, the proposed option for
manufacturers to generate NOx emission credits from BEVs and FCEVs as early as MY 2024 may also help to
accelerate some benefits of the proposed program (See Preamble Sections III. A and IV.I for more details on the
proposal to allow manufacturers to generate NOx emission credits from BEVs and FCEVs).
zzzz https://www.epa.gov/pm-pollution/national-ambient-air-quality-standards-naaqs-pm
aaaaa The population total is calculated by summing, without double counting, the 1997, 2006 and 2012 PM2.5
nonattaimnent populations contained in the Criteria Pollutant Nonattaimnent Summary report
(https://www.epa.gov/green-book/green-book-data-download).
281
-------
Counties Designated Nonattainment
for PM-2.5 (1997, 2006, and/or 2012 Standards)
Figure 6-3: Counties Designated Nonattainment for PM2.5 (1997,2006, and/or 2012 standards)
The proposed rule begins to take effect in 2027 and would assist areas with attaining the
NAAQS and may relieve areas with already stringent local regulations from some of the burden
associated with adopting additional local controls.BBBBB The proposed rule would also provide
assistance to counties with ambient concentrations near the level of the NAAQS who are
working to ensure long-term attainment or maintenance of the PM2.5 NAAQS.
6.1.3 NO2
There are two primary NAAQS for NO2: an annual standard (53 ppb) and a 1-hour standard
(100 ppb).ccccc In 2010, EPA established requirements for monitoring NO2 near roadways
bbb§b \yiujg no| quantified in the air quality modeling analysis for this proposed rule, the Early Adoption Incentives
under the proposed program could encourage manufacturers to introduce new emission control technologies prior to
the 2027 model year, which may help to accelerate some benefits of the proposed program (See Preamble Section
IV.H for more details on the proposed Early Adoption Incentives). In addition, the proposed option for
manufacturers to generate NOx emission credits from BEVs and FCEVs as early as MY 2024 may also help to
accelerate some benefits of the proposed program (See Preamble Sections III. A and IV.I for more details on the
proposal to allow manufacturers to generate NOx emission credits from BEVs and FCEVs).
seem j|le statistical form of the 1-hour N AAQS for NO2 is the 3-year average of the yearly distribution of 1-hour
daily maximum concentrations.
282
-------
expected to have the highest concentrations of NO2 within large cities. Monitoring within this
near-roadway network began in 2014, with additional sites deployed in the following years. At
present, there are no nonattainment areas for NO2.
6.1.4 CO
There are two primary NAAQS for CO: an 8-hour standard (9 ppm) and a 1-hour standard (35
ppm). There are currently no CO nonattainment areas; as of September 27, 2010, all CO
nonattainment areas had been redesignated to attainment.
6.1.5 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.323 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 2007 Mobile Source Air Toxics
(MSAT) Rule.324 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 2014, and it was
released in 20 1 8.325 The 2014 NATA 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 National Air Toxic Assessment (NATA) for 2014, mobile sources were
responsible for over 40 percent of outdoor anthropogenic toxic emissions and were the largest
contributor to national average cancer and noncancer risk from directly emitted
pollutants 325 DDDDD Mobile sources are also significant contributors to precursor emissions
which react to form air toxics.326 Formaldehyde is the largest contributor to cancer risk of all 71
pollutants quantitatively assessed in the 2014 NATA. Mobile sources were responsible for more
than 25 percent of primary anthropogenic emissions of this pollutant in 2014 and are significant
contributors to formaldehyde precursor emissions. Benzene is also a large contributor to cancer
risk, and mobile sources account for almost 70 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.
ddddd 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.
283
-------
6.1.6 Visibility
As of May 31, 2021, over 32 million people live in areas that are designated nonattainment
for the PM2.5 NAAQS. Overall, the evidence is sufficient to conclude that a causal relationship
exists between PM and visibility impairment.327 Thus, the populations who live in
nonattainment areas and travel to these areas would likely be experiencing visibility impairment.
Additionally, while visibility trends have improved in Mandatory Class I Federal areas, these
areas continue to suffer from visibility impairment.328 329 EEEEE In summary, visibility
impairment is experienced throughout the U.S., in multi-state regions, urban areas, and remote
Mandatory Class I Federal areas.
6.1.7 Deposition
Over the past two decades, the EPA has undertaken numerous efforts to reduce nitrogen
deposition across the U.S. Analyses of monitoring data for the U.S. show that deposition of
nitrogen compounds has decreased over the last 25 years. At 34 long-term monitoring sites in
the eastern U.S., where data are most abundant, average total nitrogen deposition decreased by
43 percent between 1989-1991 and 2014-2016.330 FFFFF Although total nitrogen deposition has
decreased over time, many areas continue to be negatively impacted by deposition.
6.2 Air Quality Modeling Methodology
This section describes the air quality modeling done to support the proposed rule.
6.2.1 Air Quality Model
CMAQ is a non-proprietary computer model that simulates the formation and fate of
photochemical oxidants, primary and secondary PM concentrations, acid deposition, and air
toxics, over regional and urban spatial scales for given inputs of meteorological conditions and
emissions. CMAQ includes numerous science modules that simulate the emission, production,
decay, deposition and transport of organic and inorganic gas-phase and particle pollutants in the
atmosphere. The CMAQ model is a well-known and well-respected tool and has been used in
numerous national and international applications.GGGGG
The air quality modeling analysis used the 2016vl platform with the most recent multi-
pollutant CMAQ code available at the time of air quality modeling (CMAQ version 5.3.1). The
2016 CMAQ runs utilized the CB6r3 chemical mechanism (Carbon Bond with linearized
halogen chemistry) for gas-phase chemistry, and AER07 (aerosol model with non-volatile
primary organic aerosol) for aerosols. The CMAQ model is regularly peer reviewed, with the
most recent review completed in 2019 on version 5.2 and 5.3beta.331
6.2.2 Model Domain and Configuration
The CMAQ modeling analyses used a domain covering the continental United States, as
shown in Figure 6-4. This single domain covers the entire continental U.S. (CONUS) and large
EEEEE Mandatory Class I Federal areas are the 156 national parks and wilderness areas where state and federal
agencies work to improve visibility, https://www.epa.gov/visibility/regional-haze-program.
fffff Trends data comes from the EPA Report on the Enviromnent. Accessed in 2020,
https://cfpub.epa.gov/roe/indicator.cfm?i=l#4 Based on data from the NADP/National Trends Network, 2018.
ggggg More information available at: https://www.epa.gov/cmaq
284
-------
portions of Canada and Mexico using 12 km x 12 km horizontal grid spacing. The 2016
simulation used a Lambert Conformal map projection centered at (-97, 40) with true latitudes at
33 and 45 degrees north. The model extends vertically from the surface to 50 millibars
(approximately 17,600 meters) using a sigma-pressure coordinate system with 35 vertical layers.
6.2.3 Model Inputs
The key inputs to the CMAQ model include emissions from anthropogenic and biogenic
sectors, meteorological data, and initial and boundary conditions.
The onroad emissions inputs used for the 2045 reference and control scenarios are
summarized in Chapter 5 of this draft RIA, and emissions inputs for other sectors are described
in the documentation for the 2016vl modeling platform.332 The reference scenario represents
projected 2045 emissions without the proposed aile, and the control scenario represents
projected 2045 emissions with proposed Option l.HHHHH The emissions inventories used for the
h0h8h notecj jn Chapter 5.4 of the draft RIA, while we refer to this modeling as for the proposed Optionl, there
are differences between the proposed Option 1 standards, emission warranty, and useful life provisions presented in
Sections III and IV of the preamble and those included in the control scenario modeled for the air quality analysis.
285
-------
air quality modeling control scenario and the national-scale emissions inventories presented in
Chapter 5.3 of the draft RIA and Section VI of the preamble are consistent in many ways, but
there are some differences. Chapter 5.4 of the draft RIA has more detail on the differences
between the air quality control scenario and national-scale inventories. The AQM TSD also
contains a detailed discussion of the emissions inventory inputs used in our air quality
modeling.340
The CMAQ meteorological input files were derived from simulations of the Weather
Research and Forecasting Model (WRF) version 3.8 for the entire 2016 year,333-334 The WRF
Model is a state-of-the-science mesoscale numerical weather prediction system developed for
both operational forecasting and atmospheric research applications.335 The meteorological
outputs from WRF were processed to create 12 km model-ready inputs for CMAQ using the
Meteorology-Chemistry Interface Processor (MCIP) version 4.3. 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.336
The boundary and initial species concentrations were provided by a northern hemispheric
CMAQ modeling platform for the year 2016.337-338 The hemispheric-scale platform uses a polar
stereographic projection at 108 km resolution to completely and continuously cover the northern
hemisphere for 2016. Meteorology is provided by WRF v3.8. Details on the emissions used for
hemispheric CMAQ can be found in the 2016 hemispheric emissions modeling platform TSD.339
The atmospheric processing (transformation and fate) was simulated by CMAQ (v5.2.1) using
the CB6r3 and the aerosol model with non-volatile primary organic carbon (AE6nvPOA). The
CMAQ model also included the on-line windblown dust emission sources (excluding agricultural
land), which are not always included in the regional platform but are important for large-scale
transport of dust.
6.2.4 CMAQ Evaluation
The CMAQ predictions for ozone, fine particulate matter, sulfate, nitrate, ammonium, organic
carbon, elemental carbon, nitrogen deposition, and specific air toxics (formaldehyde,
acetaldehyde, benzene and naphthalene) from the 2016 base scenario were compared to
measured concentrations in order to evaluate the ability of the modeling platform to replicate
observed concentrations. This evaluation was comprised of statistical and graphical comparisons
of paired modeled and observed data. Details on the model performance evaluation, including a
description of the methodology, the model performance statistics, and results, are provided in the
Air Quality Modeling TSD for this proposed rulemaking (AQM TSD).340
6.2.5 Model Simulation Scenarios
As part of our analysis for this rulemaking, the hourly CMAQ outputs were used to calculate
8-hour ozone design values concentrations, daily and annual PM2.5 design value concentrations,
annual NO2 concentrations, annual CO concentrations, annual and seasonal (summer and winter)
air toxics concentrations, visibility levels and annual total nitrogen deposition for each of the
following scenarios:
- 2016 base year
- 2045 reference
- 2045 control
286
-------
Air quality modeling was done for the future year 2045 when the program would be fully
implemented and when most of the regulated fleet would have turned over. We use the
predictions from the air quality model in a relative sense by combining the 2016 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 for the May 1 - Sept 30
ozone season, daily and annual PM2.5 concentrations, and visibility impairment for each of the
2045 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., 2014-2018).341 Additional
predictions from the CMAQ model are used in the distributional analysis (Chapter 6.3.9) and in
the benefits analysis described in Chapter 8.3.1 of the DRIA. The CO, NO2, annual and seasonal
formaldehyde, acetaldehyde, benzene, naphthalene, and annual nitrate deposition projections
were not predicted in a relative sense due to the limited observational data available.
The projected daily and annual PM2.5 design values were calculated using the Speciated
Modeled Attainment Test (SMAT) approach. 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)."342 Several updated
datasets and techniques were used for this analysis. These changes are fully described within the
technical support document for the Final Transport Rule Air Quality Modeling Technical
Support Document.343 The projected 8-hour ozone design values were calculated using the
approach identified in EPA's guidance on air quality modeling attainment demonstrations.344
6.3 Air Quality Modeling Results
This section describes the results of the air quality modeling analysis. The "reference"
scenario represents projected 2045 air quality without the proposed rule and the "control"
scenario represents projected 2045 air quality with the proposed Option 1.11111 This chapter of the
draft RIA presents modeled changes in ambient concentrations of air pollutants when comparing
the "reference" and "control" scenarios. Decreases in concentration mean that the "control"
scenario decreases the pollutant concentration compared to the "reference" scenario.
Everything in the reference and control scenarios was held constant except the onroad
inventories, which reflected the application of the proposed Option 1 at the time we conducted
the modeling.JJJJJ This includes the meteorological data (reflecting calendar year 2016
conditions) and the emissions for all other sources, including boundary conditions and initial
conditions used in the air quality modeling methodology.
The reference and control scenarios include projections of existing mobile source emission
control programs that EPA has already adopted, as well as other federal, state and local programs
which are expected to reduce concentrations of pollutants in the ambient air in the future. These
control programs include (but are not limited to) the Tier 3 Motor Vehicle Emission and Fuel
Standards (79 FR 23414, April 28, 2014), the New Marine Compression-Ignition Engines at or
Above 30 Liters per Cylinder Rule (75 FR 22895, April 30, 2010), the Locomotive and Marine
Compression-Ignition Engine Rule (73 FR 25098, May 6, 2008), the Clean Air Nonroad Diesel
11111 Due to resource constraints, we only conducted air quality modeling for the proposed Option 1.
inn as noted in Chapter 5.4 of the draft RIA, while we refer to this modeling as for the proposed Option 1, there are
differences between the proposed Option 1 standards, emission warranty, and useful life provisions presented in
Sections III and IV of the preamble and those included in the control scenario modeled for the air quality analysis.
287
-------
(69 FR 38957, June 29, 2004), and the Heavy-Duty Engine and Vehicle Standards and Highway
Diesel Fuel Sulfur Control Requirements (66 FR 5002, January 18, 2001).
Not included in the reference or control scenarios are additional federal or state programs that
were not finalized at the time that the air quality modeling analysis for the proposal was initiated.
For example, the CARB Heavy-Duty Low NOx Omnibus rule and the CA Advanced Clean
Trucks (ACT) rule were not final, so the emission reductions associated with these rulemakings
are not included in the air quality modeling analysis for the proposed rule KKKKK-LLLLL
6.3.1 Ozone Design Value Impacts of Proposed Rulemaking
This section summarizes the ozone air quality impacts of the proposed rule in 2045, based on
our CMAQ modeling. Our modeling indicates that ozone design value concentrations will
decrease dramatically in many areas of the country as a result of the proposed rule.
Figure 6-5 presents the changes in 8-hour ozone design value concentrations in 2045.MMMMM
kkkkk Additional information on the CARB Omnibus program is available in Section I.D of the preamble for this
proposed rule. Additional discussion on the CARB ACT program is available in Sections I.D, VI.D, and XI.
lllll Draf( ria Chapter 5 Appendix 6 presents a sensitivity analysis of the estimated emission inventory impacts
from nationwide adoption of the Omnibus rule.
1414141414 An 8-hour ozone design value is the concentration that determines whether a monitoring site meets the
NAAQS for ozone. The full details involved in calculating an 8-hour ozone design value are given in appendix I of
40 CFR part 50.
288
-------
Figure 6-5: Projected Change in 8-hour Ozone Design Values in 2045 due to Proposed Rule
As shown in Figure 6-5, the majority of the design value decreases in 2045 are greater than
1.5 ppb. There are also 82 counties with projected 8-hour ozone design value decreases of more
than 2.5 ppb; many of the counties with the largest design value decreases are in California, and
in the Atlanta and St. Louis urban areas. The maximum projected decrease in an 8-hour ozone
design value in 2045 is 5.1 ppb in Riverside County, California. Not all counties have monitor
data that meets the requirements to calculate a design value concentration; counties without a
calculated design value are left white.
Table 6-1 shows the average projected change, due to the proposed rule, in 2045 8-hour ozone
design values for: (1) all modeled counties (with 2016 base case design values), (2) counties with
2016 base case design values that are above the level of the 2015 ozone NAAQS, (3) counties
with 2016 base case design values that are equal to or within 10% below the level of the 2015
NAAQS, (4) counties with 2045 reference scenario design values that are above the level of the
2015 ozone NAAQS, (5) counties with 2045 reference scenario design values that are equal to or
within 10% below the level of the 2015 ozone NAAQS, (6) counties with 2045 control scenario
design values that are above the level of the 2015 ozone NAAQS, and (7) counties with 2045
control scenario design values that are equal to or within 10% below the level of the 2015 ozone
NAAQS. Counti es within 10 percent of the level of the NAAQS are intended to reflect counties
that although not violating the standards, would also be impacted by changes in ambient levels of
ozone as they work to ensure long-term attainment or maintenance of the ozone NAAQS. On a
289
-------
population-weighted basis, the average modeled future-year 8-hour ozone design value is
projected to decrease by over 2 ppb in 2045 due to the proposed rule.
Table 6-1: Average Change in Projected 8-hour Ozone Design Values in 2045 due to Proposed Rule
Projected Design Value Category
Number of
Counties
2045
Population3
Average
Change
in 2045
Design
Value
(ppb)
Population-
Weighted
Average
Change in
Design
Value (ppb)
all modeled counties
457
246,949,949
-1.87
-2.23
counties with 2016 base year design values above the level
of the 2015 8-hour ozone standard
118
125,319,158
-2.12
-2.43
counties with 2016 base year design values within 10% of
the 2015 8-hour ozone standard
245
93,417,097
-1.83
-2.10
counties with 2045 reference design values above the level
of the 2015 8-hour ozone standard
15
37,758,488
-2.26
-3.03
counties with 2045 reference design values within 10% of
the 2015 8-hour ozone standard
56
39,302,665
-1.78
-2.02
counties with 2045 control design values above the level of
the 2015 8-hour ozone standard
10
27,930,138
-2.36
-3.34
counties with 2045 control design values within 10% of the
2015 8-hour ozone standard
42
31,395,617
-1.69
-1.77
a Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. (2015). Complete
Demographic Database. Washington, DC. http://www.woodsandpoole.com/index.php.
These modeling results project that there would be 15 counties with 8-hour ozone design
values above the level of the 2015 ozone NAAQS in 2045 without the proposed rule or any other
additional standards in place. Table 6-2 below presents the changes in design values for these
counties.
290
-------
Table 6-2: Change in 8-hour Ozone Design Values for Counties Projected to be Above the Level of the 2015 8-
hour Ozone NAAQS in 2045
County Name, State
Population in
2045a
Change in 2045
projected 8-hour
Ozone Design
Value (DV) (ppb)
2045 Reference
Ozone Design
Value (ppb)
2045 Control
Ozone Design
Value (ppb)
San Bernardino, California
3,191,663
-4.6
98.0
93.4
Los Angeles, California
11,755,545
-3.3
92.3
89.0
Riverside, California
3,926,478
-5.1
83.3
78.2
Fairfield, Connecticut
1,050,293
-1.4
79.4
78.0
Imperial, California
296,070
-0.3
76.6
76.3
Kern, California
1,251,350
-2.2
76.5
74.3
San Diego, California
4,452,722
-2.5
75.2
72.7
Fresno, California
1,371,355
-2.8
74.9
72.1
Richmond, New York
614,033
-0.8
73.7
72.9
Mariposa, California
20,630
-0.6
72.0
71.4
Salt Lake, Utah
1,387,960
-1.9
71.9
70.0
Tulare, California
601,851
-2.4
71.5
69.1
Shebovgan, Wisconsin
124,284
-1.9
71.4
69.5
Davis, Utah
556,296
-1.9
71.3
69.4
Harris, Texas
7,157,959
-2.2
71.2
69.0
a Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. (2015). Complete
Demographic Database. Washington. DC. http://www.woodsandpoole.com/index.php.
Our modeling predicts that the proposed rule would reduce ozone design values in some
counties from above the level of the standard to below it. While the number of counties with
projected design values above the level of the NAAQS is less certain than the average projected
changes in design values, our modeling projects that in 2045 ozone design values in five counties
(Salt Lake and Davis Counties in Utah, Tulare County in California, Sheboygan County in
Wisconsin, and Harris County in Texas) will change from being above the level of the standard
in the reference scenario to being below the level of the standard in the control scenario. The
projected population in these five counties in 2045 is almost 10 million people.
As described in Chapter 4 of this draft RIA, the science of ozone formation, transport, and
accumulation is complex. The air quality modeling projects ozone design value decreases as a
result of emissions changes from the proposed standards in the vast majority of counties. This
change in ozone results from interactions between photochemistry, background concentrations of
ozone, VOC and NOx, local emissions and meteorology. However, there is one county in 2045
that is projected to have no change in modeled ozone design value concentration (Skagit County,
Washington).
6.3.2 Annual PM2.5 Design Value Impacts of Proposed Rulemaking
This section summarizes the annual average PM2.5 air quality impacts of the proposed rule in
2045, based on our CMAQ modeling. Our modeling indicates that annual PM2.5 design values
will decrease due to the proposed rule. The decreases in annual PM2.5 design values are due to
the projected reductions in NOx, primary PM2.5, and VOC emissions.
291
-------
Figure 6-6 presents the changes in annual PM2.5 design values in 2045.NNNNN
Figure 6-6: Projected Change in Annual PM2.5 Design Values in 2045 due to Proposed Rule
As shown in Figure 6-6, we project that in 2045 most counties will have design value
decreases of between 0.01 jig/m3 and 0.05 ug/m\ There are also 15 counties with projected
annual PM2.5 design value decreases of more than 0.1 ug/m '; these counties are in California and
Utah. The maximum projected decrease in a 2045 annual PM2.5 design value is 0.21 ug/m3 in
Tulare County, California. Not all counties have monitor data that meets the requirements to
calculate a design value concentration; counties without a calculated design value are left white.
Table 6-3 presents the average projected change, due to the proposed rule, in 2045 annual
PM2.5 design values for: (1) all modeled counties (with 2016 base case design values), (2)
counties with 2016 base case design values that are above the level of the 2012 annual PM2.5
standard, (3) counties with 2016 base case design values that are equal to or within 10% below
the level of the 2012 standard, (4) counties with 2045 reference scenario design values that are
above the level of the 2012 annual PM2.5 standard, (5) counties with 2045 reference scenario
design values that are equal to or within 10% below the level of the 2012 annual PM2.5 standard,
(6) counties with 2045 control scenario design values that are above the level of the 2012 annual
«¦* An annual PM2.5 design value is the concentration that determines whether a monitoring site meets the annual
NAAQS for PM3.5. The full details involved in calculating an amiual PM2.5 design value are given in appendix N of
40 CFR part 50.
292
-------
PM2.5 standard and (7) counties with 2045 control scenario design values that are equal to or
within 10% below the level of the 2012 annual PM2.5 standard. Counties within 10 percent of the
level of the standard are intended to reflect counties that although not violating the standards,
would also be impacted by changes in ambient levels of PM2.5 as they work to ensure long-term
attainment or maintenance of the 2012 annual PM2.5 NAAQS. On a population-weighted basis,
the average modeled future year annual PM2.5 design value is projected to decrease by 0.04
|ig/m3 due to the proposed rule.
Table 6-3: Average Change in Projected Annual PM2.5 Design Values in 2045 due to Proposed Rule
Projected Design Value Category
Number
of
Counties
2045
Population3
Average
Change in
2045
Design
Value
(ug/m3)
Population-
Weighted
Average
Change in
Design Value
(ug/m3)
all modeled counties
568
273,604,437
-0.04
-0.04
counties with 2016 base year design values above
the level of the 2012 annual PM2.5 standard
17
26,726,354
-0.09
-0.05
counties with 2016 base year design values
within 10% of the 2012 annual PM2.5 standard
5
4,009,527
-0.06
-0.06
counties with 2045 reference design values above
the level of the 2012 annual PM2.5 standard
12
25,015,974
-0.10
-0.05
counties with 2045 reference design values
within 10% of the 2012 annual PM2.5 standard
6
1,721,445
-0.06
-0.06
counties with 2045 control design values above
the level of the 2012 annual PM2.5 standard
10
23,320,070
-0.10
-0.05
counties with 2045 control design values within
10% of the 2012 annual PM2.5 standard
8
3,417,349
-0.08
-0.09
a Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. (2015). Complete
Demographic Database. Washington, DC. http://www.woodsandpoole.com/index.php.
There are 12 counties, mostly in California, that are projected to have annual PM2.5 design
values above the level of the NAAQS in 2045 without the proposed rule or any other additional
standards in place. Table 6-4 below presents the changes in design values for these counties.
293
-------
Table 6-4: Change in Annual PM2.5 Design Values for Counties Projected to be Above the Level of the Annual
PMis NAAQS in 2045
County Name, State
Population
in 2045a
Change in 2045
Projected Annual
PM2.5 Design Value
(DV) (jig/m3)
2045 Reference
Design Value
(jig/m3)
2045 Control
Design Value
(jig/m3)
Kern, California
1,251,350
-0.15
15.78
15.63
Kings, California
185,866
-0.19
14.82
14.63
San Bernardino, California
3,191,663
-0.02
14.21
14.19
Plumas, California
21,297
-0.05
14.12
14.06
Tulare, California
601,851
-0.21
14.05
13.84
Riverside, California
3,926,478
-0.04
13.43
13.39
Imperial, California
296,070
-0.02
12.93
12.91
Fresno, California
1,371,355
-0.16
12.87
12.71
Los Angeles, California
11,755,545
-0.02
12.26
12.24
Pinal, Arizona
718,595
-0.11
12.24
12.13
Stanislaus, California
716,019
-0.16
12.17
12.02
San Joaquin, California
979,885
-0.11
12.07
11.96
"¦Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. (2015). Complete
Demographic Database. Washington, DC. http://www.woodsandpoole.com/index.php.
Our modeling predicts that the proposed rule would reduce annual PM2.5 design values in
some counties from above the level of the standard to below it. While the number of counties
with projected design values above the level of the NAAQS is less certain than the average
changes in design values, annual PM2.5 design values in two counties (Stanislaus County,
California and San Joaquin County, California) are projected to change from being above the
level of the standard in the reference scenario to being below the level of the standard in the
control scenario. The projected population in these two counties in 2045 is over 1.5 million
people.
6.3.3 24-hour PMis Design Value Impacts of Proposed Rulemaking
This section summarizes the 24-hour PM2.5 air quality impacts of the proposed rule in 2045,
based on our CMAQ modeling. Our modeling indicates that most 24-hour PM2.5 design values
would decrease due to the proposed rule. The decreases in 24-hour PM2.5 design values are due
to the projected reductions in NOx, primary PM2.5, and VOC emissions.
Figure 6-7 presents the changes in 24-hour PM2.5 design values in 2045.00000
uuuuu a 24-hour PM2 5 design value is the concentration that determines whether a monitoring site meets the 24-
hour NAAQS for PM2 5. The full details involved in calculating a 24-hour PM2 5 design value are given in appendix
N of 40 CFR part 50.
294
-------
JT«~ -rS-^ . 1
j«* % >>¦ 4-1 ^ Afi#
«4 ,i 't - xlf y M,
yr - J fcfc" a* _•> j, \}Jr
& rS^ZH .;. d"*i' -
A vjTr i_ ; iv -t '
9^\ ,¦ b®- t8 | % \ vv^-5r*s /¦»
WkmJ*~~Z ¦ , *\ « &fj
« i rr=rVi-iit£fe£?
~ * j ' j ^•YT»y;xy
\jl,jL_I 1 Vf). *f-' «
^ ) t? ^ j&~:xl,a
Legend n-i-ito- *--f>f.n.« , ¦-> .. («_•. • v "wjij "we \- -:\
¦:- -O.S US»'iri3 ifi \ ** w v L:"'- \
• £> a
-{1 5 to ¦=- -CJ y 5 ; ~ j
> -2.261n ¦== -a 1& "35 \ 7 V\tJ
>-3191c *=-0 05 372 _ j?
>•¦5 0510 ••0,M 'Z1
»- £ 061c- •= '3.15 1
» C-.1S 10*0.25 5
v= 0 ,T> 1(5 v a S 0
; County High Site Difference in DsuVy PM2.S OV — 204Sfh_ca_CTI minus IMSthjvf CTI
Figure 6-7: Projected Change in 24-hour PM2.5 Design Values in 2045 due to Proposed Rule
As shown in Figure 6-7, in 2045 there are 170 counties with projected 24-hour PM2.5 design
value decreases greater than 0.15 jig/m3. These counties are in mainly in the midwest, southeast
and western United States. The maximum projected decrease in a 2045 24-hour PM2.5 design
value is 1.79 ug/m! in Tulare County, California. Not all counties have monitor data that meets
the requirements to calculate a design value concentration; counties without a calculated design
value are left white.
Table 6-5 shows the average projected change, due to the proposed rule, in 2045 24-hour
PM2.5 design values for: (1) all modeled counties (with 2016 base case design values), (2)
counties with 2016 base case design values that are above the level of the 2006 24-hour PM2.5
standard, (3) counties with 2016 base case design values that are equal to or within 10% below
the level of the 2006 24-hour PM2.5 standard, (4) counties with 2045 reference scenario design
values that are above the level of the 2006 24-hour PM2.5 standard, (5) counties with 2045
reference scenario design values that are equal to or within 10% below the level of the 2006 24-
hour PM2.5 standard, (6) counties with 2045 control scenario design values that are above the
level of the 2006 24-hour PM2.5 standard, and (7) counties with 2045 control scenario design
values that are equal to or within 10% below the level of the 2006 24-hour PM2.5 NAAQS.
Counties within 10 percent of the level of the standard are intended to reflect counties that
although not violating the standards, would also be impacted by changes in ambient levels of
PM2.5 as they work to ensure long-term attainment or maintenance of the 2006 24-hour PM2.5
295
-------
NAAQS. On a population-weighted basis, the average modeled future-year 24-hour PM2.5
design value is projected to decrease by 0.17 |ig/m3 in 2045 due to the proposed rule.
Table 6-5: Average Change in Projected 24-hour PM2.5 Design Values in 2045 due to Proposed Rule
Projected Design Value Category
Number
of
Counties
2045
Population3
Average
Change
in 2045
Design
Value
(ug/m3)
Population-
Weighted
Average
Change in
Design Value
(ug/m3)
all modeled counties
568
272,852,777
-0.12
-0.17
counties with 2016 base year design values above the
level of the 2006 daily PM2.5 standard
33
28,394,253
-0.40
-0.67
counties with 2016 base year design values within 10%
of the 2006 daily PM2.5 standard
15
13,937,416
-0.18
-0.27
counties with 2045 reference design values above the
level of the 2006 daily PM2.5 standard
29
14,447,443
-0.38
-0.55
counties with 2045 reference design values within 10%
of the 2006 daily PM2.5 standard
12
22,900,297
-0.30
-0.59
counties with 2045 control design values above the
level of the 2006 daily PM2.5 standard
29
14,447,443
-0.38
-0.55
counties with 2045 control design values within 10%
of the 2006 daily PM2.5 standard
10
19,766,216
-0.26
-0.60
a Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. (2015). Complete
Demographic Database. Washington, DC. http://www.woodsandpoole.com/index.php.
There are 29 counties that are projected to have 24-hour PM2.5 design values above the level
of the NAAQS in 2045 without the proposed rule or any other additional controls in place. Table
6-6 below presents the changes in design values for these counties.
296
-------
Table 6-6: Change in 24-hour PM2.5 Design Values for Counties Projected to be Above the 24-hour PM2.5
NAAQS in 2045
County Name, State
Population in
2045a
Change in 24-
hour PM2.5
Design Value
frig/m3)
2045 Reference
Design Value
(jig/m3)
2045 Control
Design Value
(jig/m3)
Okanogan, Washington
47,922
-0.24
57.02
56.79
Ravalli, Montana
52,336
-0.03
56.93
56.90
Kern, California
1,251,350
-0.49
55.78
55.29
Fresno, California
1,371,355
-0.68
49.86
49.17
Jackson, Oregon
281,974
-0.28
49.22
48.94
Kings, California
185,866
-1.42
47.23
45.81
Plumas, California
21,297
-0.16
46.30
46.14
Klamath, Oregon
71,950
-0.16
44.39
44.24
Siskiyou, California
46,491
-0.03
44.04
44.02
Lincoln, Montana
19,924
-0.14
43.05
42.91
Tulare, California
601,851
-1.79
42.63
40.84
Missoula, Montana
139,759
-0.13
42.49
42.36
Lemhi, Idaho
8,830
-0.08
42.46
42.39
Lewis and Clark, Montana
95,256
-0.06
41.17
41.11
Flathead, Montana
150,424
-0.12
40.75
40.64
Yakima, Washington
289,388
-0.22
40.64
40.42
Lake, Oregon
8,605
-0.10
40.43
40.33
Stanislaus, California
716,019
-1.26
39.54
38.28
Lane, Oregon
440,599
-0.13
39.53
39.39
Josephine, Oregon
106,207
-0.27
39.46
39.19
Alameda, California
1,936,700
-0.30
38.81
38.51
Madera, California
208,957
-0.62
38.49
37.87
San Joaquin, California
979,885
-1.22
38.15
36.93
Kittitas, Washington
53,927
-0.17
37.82
37.65
Riverside, California
3,926,478
-0.51
37.35
36.84
Shoshone, Idaho
11,064
-0.16
37.07
36.91
Crook, Oregon
24,645
-0.19
37.00
36.81
Benewah, Idaho
10,426
-0.13
36.72
36.59
Salt Lake, Utah
1,387,960
0.02
36.04
36.06
aPopulation numbers based on Woods & Poole data. Woods & Poole Economics, Inc. (2015). Complete
Demographic Database. Washington, DC. http://www.woodsandpoole.com/index.php.
While the count of modeled nonattainment counties is much less certain than the average
changes in air quality, in 2045, there are no 24-hour PM2.5 design values that are projected to
change from being above the level of the standard in the reference case to being below the level
of the standard in the proposed control case.
As described in Chapter 4 of this draft RIA, PM2.5 in the atmosphere can be primary or
secondary and its composition, transport, and accumulation is complex. The air quality
modeling projects 24-hour PM2.5 design value decreases as a result of emissions changes from
the proposed rule in the vast majority of counties. However, there are a handful of counties
where 24-hour PM2.5 design values are projected to increase. These increases are likely due to
elevated secondary PM2.5 formation rates as a result of increased oxidant levels, which occur
during stagnant cold weather due to reductions in NOx.
297
-------
6.3.4 Nitrogen Dioxide Concentration Impacts of Proposed Rulemaking
This section summarizes the annual average NO2 air quality impacts of the proposed rule in
2045, based on our CMAQ modeling. Our modeling indicates that annual average NO2
concentrations would decrease as a result of the proposed rule, if finalized as proposed. Figure
6-8 presents the changes in annual NO2 concentrations in 2045.
As shown in Figure 6-8, our modeling indicates that by 2045 annual NO2 concentrations in
the majority of the country would decrease between 0.01 and 0.1 ppb due to the proposed rule.
However, decreases in annual NO2 concentrations would be greater than 0.2 ppb along many
highway corridors and greater than 0.3 ppb in most urban areas. The absolute reductions
correspond to reductions of greater than 5 percent in annual NO2 concentrations across much of
the country, see Figure 6-9. Although we didn't model changes in 1-hour concentrations, the
proposed rule would also likely decrease 1-hour NO2 concentrations and help any potential
nonattainment areas attain and maintenance areas maintain the NO2 standard.ppppp
ppppp as noted in Chapter 6.1.3, there are currently no nonattainment areas for the NO2 NAAQS.
298
-------
ppc-
M •- -3.30
¦¦ -0.30 to -O.ZC
a -0.20 Lo -0.1.C
0.20 tf 0.01
D.01 Id O.tll
tl.cil ra CI. 1 tl
0,10-DO 0.20
" 0.30 "0 0 30
Figure 6-9: Percent Change in Annual Ambient NO: Concentrations in 2045
6.3.5 Carbon Monoxide Concentration Impacts of Proposed Ruleniakiii2
This section summarizes the annual average CO air quality impacts of the proposed rule in
2045, based on our CMAQ modeling. Our modeling indicates that annual average CO
concentrations would decrease as a result of the proposed rule. Figure 6-10 and Figure 6-11
present the absolute and percent changes in annual CO concentrations in 2045.
Figure 6-8: Projected Absolute Change in Annual Ambient NQ2 Concentrations in 2045
Hi -30.0
™ -50.0 to-25 D
¦¦ -25.0 LQ -10.D
Hi -1D.0 to -5.0
-5.0 ju -2.5
2.5 zo 1.0
-I.LI Tl 1.0
1.C1 tn 7 rj
2.5to b.D
¦¦ 5 0 to 10 0
¦¦I 10.0 IO 25.0
25.0 ta 50.0
M > 50.0
299
-------
As shown in Figure 6-10, our modeling indicates that by 2045 annual CO concentrations in
the majority of the country would decrease between 0.02 and 0.5 ppb due to the proposed
rulemaking. However, decreases in annual CO concentrations would be greater than 1.5 ppb in
some urban areas. The absolute reductions correspond to percent changes of less than 1 percent
across the country, except for the Phoenix area, where there are some larger decreases between 1
and 2 percent. Although we didn't model changes in 8-hour or 1-hour concentrations, the
proposed standards would also likely decrease 1-hour and 8-hour CO concentrations and help
any potential nonattainment areas attain and maintenance areas maintain the CO standard.QQQQQ
m } : • i ¦
1 3 fT*-'- I.
\J i - - -, L
M
% - -%•
i#'' PPfr-
m¦ K
.. / M -j? ,00 to -2.5C-
~ • ' ; y V M -2.50LO-2.0C
) /y" '\ f '%? " -2.00tD-l.5C
• • ,'X i.bOtD 1.0C
- ; .¦¦ -i.ootD-o.5fi
-/ * 1 £-ObO to -O.Oi
¦ • ' -DO? to 0 0?
. ' * ./ > o.oz
i ' K
Mtft- 3'CP'' Mn: 3 065 '
Figure 6-10: Absolute Change in Annual Ambient CO Concentrations in 2045
QQQQQ aS noted in Chapter 6.1.4, there are currently no nonattainment areas for the CO NAAQS.
300
-------
%
< -DC? -U
-500 ro -75-0
¦¦
-25.0 to-10.0
-20.0 to-5.0
-5.0 jj -2.5
2.5 to 1.0
l.U ta 1.3
1 .CI tn J
2. J t3 'j.O
¦¦
5 0 tD 10 3
10.0 10 25.0
M
25.0 to 50.0
¦¦
> 50.0
Figure 6-11: Percent Change in Annual Ambient CO Concentrations in 2045
6.3.6 Air Toxics Impacts of Proposed Rulemaking
This section summarizes the changes in annual average air toxic (acetaldehyde, benzene,
formaldehyde and naphthalene) concentrations in 2045 due to the proposed rule. Our modeling
indicates that the proposed rule would have relatively little impact on national average ambient
concentrations of the modeled air toxics in 2045. Annual percent changes are less than 1% for
air toxics across most of the country. Annual absolute changes in ambient concentrations are
generally less than 0.001 |ig/m3 for benzene and naphthalene (Figure 6-12 and Figure 6-13
below). There are small increases in acetaldehyde across the country, see Figure 6-14. The
increases in acetaldehyde likely occur because species that lead to production or recycling of
acetaldehyde increase as their reactions with nitrogen oxides decrease. For formaldehyde there
are decreases across most of the country and a few areas with increases, see Figure 6-15. The
increases in formaldehyde concentration due to the proposed rule are likely related to higher
concentrations of OH radicals in areas where ozone increases due to NOx emissions reductions
(see Chapter 4.1.1.1).
301
-------
¦¦ 0.200 to 0.300 M >50.0
¦¦ >0.300
Figure 6-12: Changes in Ambient Benzene Concentrations in 2045 due to Proposed Rule: Absolute Changes
in fig/m' (left) and Percent Changes (right)
B 10.0 to 25.0
M 25.0 to 50.0
B >50.0
Figure 6-13: Changes in Ambient Naphthalene Concentrations in 2045 due to Proposed Rule: Absolute
Changes in jig/m3 (left) and Percent Changes (right)
302
-------
Figure 6-14: Changes in Ambient Acetaldehyde Concentrations in 2045 due to Proposed Rule: Absolute
Changes in jig/rn ' (left) and Percent Changes (right)
2.5 to 5.0
^ 5.0 to 10.0
M 10.0 to 25.0
^ 25.0 to 50.0
Figure 6-15: Changes in Ambient Formaldehyde Concentrations in 2045 due to Proposed Rule: Absolute
Changes in jig/rn1 (left) and Percent Changes (right)
6.3.7 Visibility Impacts of Proposed Rulemaking
Air quality modeling was used to project visibility conditions in 145 Mandatory Class I
Federal areas across the U.S. with and without the proposed rule in 2045. The results show that
in 2045, the proposed rule would improve projected visibility on the 20% most impaired days in
all modeled areas.RRRRR The average visibility on the 20 percent most impaired days at all
ERfil® The level of visibility impairment in an area is based on the light-extinction coefficient and a unitless visibility
index, called a "deciview", which is used in the valuation of visibility. The deciview metric provides a scale for
perceived visual changes over the entire range of conditions, from clear to hazy. Under many scenic conditions, the
ug/m3
im <-0.300
¦¦ -0.300 to -0.200
^ -0.200 to-0.100
^ -0.100 to-0.010
-0.010 to -0.001
-0.001 to 0.001
0.001 to 0.010
0.010 to 0.100
^ 0.100 to 0.200
^ 0.200 to 0.300
mm >0.300
303
-------
modeled Mandatory Class I Federal areas is projected to improve by 0.04 deciviews, or 0.37
percent, in 2045. The greatest improvement in visibility would occur in San Gorgonio and San
Jacinto Wilderness Areas in California, where visibility is projected to improve by 1.56 percent
(0.21 deciviews) in 2045 due to the proposed rule. The AQM TSD contains the full visibility
results from 2045 for the 145 analyzed areas.340
6.3.8 Deposition Impacts of Proposed Rulemaking
Our air quality modeling projects decreases in nitrogen deposition due to the proposed rule.
Figure 6-16 shows that by 2045 the proposed rule would result in decreases in nitrogen
deposition over much of the eastern US and in urban areas of the western US, with the largest
decreases in Atlanta and Los Angeles. Figure 6-17 indicates those decreases correspond to
annual percent decreases of more than one percent over much of the country, with some
localized decreases of over 4 percent.
kgW hn-i
¦¦ -0.40 to -D.3C
¦¦ -0.30 to -D.2C
a -0.20 Lt>-0.1C
0.10 tD 0.05
D.U^ to O.Ui
Q.Clfj -xi (I. Ill
0.10 tD (1.20
™ 0 20 TOO 30
H 0.30 jO 0.40
M > 0.40
Figure 6-16: Absolute Change in Annual Deposition of Nitrogen in 2045
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.
304
-------
Figure 6-17: Percent Change in Annual Deposition of Nitrogen in 2045
6.3.9 Demographic Analysis of Air Quality
When feasible, EPA's Office of Transportation and Air Quality conducts full-scale
photochemical air quality modeling to demonstrate how its national mobile source regulatory
actions affect ambient concentrations of regional pollutants throughout the United States. As
described in Chapter 6.2, the air quality modeling we conducted supports our analysis of future
projections of PM2.5 and ozone concentrations in a "baseline" scenario absent the proposed
standards and in a "control" scenario that assumes the proposed Option 1 is in place.SSSsS The
incremental reductions in estimated air quality concentrations between the two scenarios are
therefore attributed to the proposed rule. These baseline and control scenarios are also used as
inputs to the health benefits analysis. As demonstrated in Chapter 6.3 and Chapter 8.6, the ozone
and PM2.5 improvements that are projected to result from the proposed rule, and the health
benefits associated with those pollutant reductions, would be substantial.
This air quality modeling data can also be used to conduct an analysis of how human
exposure to future air quality varies with sociodemographic characteristics relevant to potential
environmental justice concerns in scenarios with and without the proposed rule in place. To
compare trends, we sorted 2045 baseline air quality concentrations from highest to lowest
concentration and compared two groups - areas within the contiguous U.S. with the worst air
quality (grid cells with the highest 5 percent of concentrations) and the rest of the country
(remaining 95 percent of grid cells). This approach can then answer two principal questions to
determine disparity of air quality on the basis of race and ethnicity: !'! 11
sssss as noted in Chapter 5.4 of the draft RIA, while we refer to this modeling as for the proposed Option 1, there
are differences between the proposed Option 1 standards, emission warranty, and useful life provisions presented in
Sections III and IV of the preamble and those included in the control scenario modeled for the air quality analysis.
tttxt N0te tjiat we c|0 not jiave sjmiiar projections of income in 2045 and therefore cannot conduct a similar analysis
using measures of low income such as poverty status (or other SES measure).
305
-------
1. What is the racial and ethnic composition of areas with the worst baseline air quality in
2045?
2. Are those with the worst air quality likely to benefit more from the proposed rule?
We found that in the 2045 baseline, nearly double the number of people of color live within
areas with the worst ozone and PM2.5 air pollution compared to non-Hispanic Whites (NH-
Whites).uuuuu We also found that (in absolute terms) the largest predicted improvements in both
ozone and PM2.5 are estimated to occur in areas with the worst baseline air quality, where a
substantially larger number of people of color are expected to reside. This section describes the
data and methods used to conduct the demographic analysis and presents our results.
6.3.9.1 Data and Methods
We began with projected 2045 baseline and control scenarios of modeled PM2.5 and ozone
concentration data (described in draft RIA Chapter 8.3.1). Ambient air quality concentration data
(annual average |ig/m3 for PM2.5 and May-September daily maximum 8-hour average ppb for
ozone) was estimated at a standard grid resolution of 12km x 12km across the contiguous United
States (CONUS). Using 2045 baseline air quality data as our reference scenario, we sorted
baseline air quality concentrations from highest to lowest concentration and compared two air
quality concentration groups - grid cells in the highest 5 percent of the distribution of baseline
concentrations and grid cells in the remaining 95 percent.vvvvv The maps in Figure 6-18 display
the spatial distribution of grid cells with baseline concentrations in the highest 5 percent for both
PM2.5 and ozone concentrations. We retain this distinction throughout the analysis in order to
track how air quality is distributed by air quality concentration group in the baseline and how the
proposal impacts air quality in these same grid cells with the rule in place.
The analysis also used population projections based on economic forecasting models
developed by Woods and Poole, Inc. The Woods and Poole database contains county-level
projections of population by age, sex, and race/ethnicity out to 2050, relative to a baseline using
the 2010 Decennial Census. The projected population for 2045 was extracted from the
Environmental Benefits Mapping and Analysis Program - Community Edition (BenMAP-
CE)Wwwww at same 12km x 12km grid resolution as the air quality data. Race and ethnicity
of individuals projected to live in a given area were compiled into two broad categories, "people
of color" and "Non-Hispanic White (NH-White).".xxxxx In 2045, there are 409 million people
projected to be living in the contiguous United States; 208 million are projected to be NH-White
and 201 million are projected to be people of color. To put these projections into perspective,
1,111,1111 The demographic analysis uses air quality modeling that has a contiguous U.S. domain. The analysis does not
characterize distributional trends in areas of the U.S. that fall outside of this domain.
vvvvv using higher and lower percentiles to compare risks, exposures and outcomes has been applied by EPA's
Office of Air and Radiation in previous distributional analyses of regulatory air quality modeling (see MATS,
CSAPR, PM NAAQS) and is consistent with EPA's EJ Technical Guidance.
wwwww More information about BenMAP-CE can be found here: https://www.epa.gov/benmap. Additional
information about the population projections used in this analysis can be found in Appendix J of the BenMAP-CE
User's Manual: https://www.epa.gov/benmap/benmap-ce-manual-and-appendices.
xxxxx "pe0pie 0f color" includes Black, Asian, Native American, Hawaiian/Pacific Islander, and Hispanic
populations.
306
-------
2010 populations for the contiguous United States were 201 million for NH-White and 106
million for people of color.
Additionally, this analysis looked at the distribution of poverty status within the same air
quality concentration groups - 12km x 12km grid cells in the highest 5 percent of the distribution
of baseline concentrations and grid cells in the remaining 95 percent. We applied county-level
poverty status derived from the Census' American Community Survey (ACS) 5-year estimates
from 2015 to 2019, which represents the fraction of county-level population below and above
200% of the poverty line.''''' We note that measures of "current" poverty are not necessarily
predictors of future poverty status; poverty status in the 2045 population could be different in
terms of both scale and geographic location. However, for the purposes of this analysis, we
believe applying a "current" measure of those who live above and below 200% of the poverty
line is illustrative.
I
& * ^
- 'K./V ¦ \b.t
\
» V ... I • -
\w V\ \ f r\
!-a yj 4 \)
(a) (b)
Figure 6-18: Distributional maps of populated 12km grid cells across the contiguous United States in 2045.
Darker areas represent the location of grid cells within the highest 5 percent of baseline concentrations for (a)
PM2.5 and (b) ozone
For each pollutant and air quality concentration group (i.e., highest 5% of concentration or
remaining 95% of grid cells), we calculated the average baseline, control, and reduction in
concentrations. We then summed the population by group (people of color or NH-White;
populations above or below 200% of the poverty line) for each air quality concentration group.
6.3.9.2 Results
Of the approximately 48,000 grid cells that have a population, 2,400 are in the highest 5
percent of the baseline distribution. For PM2.5, the concentration at the 95th percentile is 7.76
ug/m3 (median: 5.18 |ig/m3), and for ozone it is 49.91 ppb (median: 38.34 ppb). In 2045, 144
million people are projected to live within the highest 5 percent of grid cells for PM2.5 and 39
million are projected to live in areas with the highest concentrations of ozone (Figure 6-18).
As shown in Table 6-7, in 2045, the number of people of color projected to live within the
grid cells with the highest baseline concentrations of ozone (26 million) is nearly double that of
NH-Whites (14 million). Thirteen percent of people of color are projected to live in areas with
'Y1 " County-level poverty status was mapped to the 12km x 12km grid cell domain using spatial weighting in
BenMAP-CE.
307
-------
the worst baseline ozone, compared to seven percent of NH-Whites. The proposed rule would
reduce human exposures to ambient ozone for all population groups, but those in areas with the
worst air quality would experience a greater reduction in ozone than those in the remaining 95
percent of grid cells.
Table 6-7: Demographic Analysis of Projected 2045 Ozone Reductions (ppb) from the Proposed Rule, by
Race/Ethnicity
Seasonal Average Ozone Concentrations in ppb
(5% to 95% Range)
2045 Population - millions
Baseline
Control
Reduction
All 12km x
12km Grid Cells
in CONUS
(n-47,795)
Total Population in
CONUS - All Grid Cells
409
39.2
(29.9 - 49.9)
38.7
(29.4 -49.5)
0.468
(0.157 -0.898)
Non-Hispanic White
Population in CONUS
208
People of Color
Population in CONUS
201
Highest 5% of
Baseline Ozone
Concentrations
(n=2,391)
Population in Highest 5%
(% of Total in CONUS)
39
(10%)
52.6
(50.0 - 58.2)
51.9
(49.6 - 57.2)
0.640
(0.273 - 1.441)
Non-Hispanic White
(% of NH-W in CONUS)
14
(7%)
People of Color
(% of POC in CONUS)
26
(13%)
Remaining 95%
of Baseline
Ozone
Concentrations
(n-45,404)
Population in Remaining 95%
(% of Total in CONUS)
370
(90%)
38.5
(29.8 - 48.7)
38.0
(29.3 -48.3)
0.457
(0.156 -0.866)
Non-Hispanic White
(% of NH-W in CONUS)
194
(93%)
People of Color
(% of POC in CONUS)
176
(86%)
PM2.5 results have a similar pattern to what we observe for ozone. As shown in Table 6-8, in
2045, the number of people of color projected to live within the grid cells with the highest
baseline concentrations of PM2.5 (93 million) is nearly double that of NH-Whites (51 million).
Forty-six percent of people of color are projected to live in areas with the worst baseline PM2.5,
compared to 25 percent of NH-Whites. Those in areas with the worst air quality would
experience a greater reduction in PM2.5 than those in the remaining 95 percent of grid cells.
308
-------
Table 6-8: Demographic Analysis of Projected 2045 PM2.5 Reductions (jig/m3) from the Proposed Rule, by
Race/Ethnicity
Annual Average PM2.5 Concentrations in jig/m3
(5% to 95% Range)
2045 Population - millions
Baseline
Control
Reduction
All 12km x
12km Grid
Cells in
CONUS
(n=47,795)
Total Population in
CONUS - All Grid Cells
409
5.23
(2.65 - 7.76)
5.21
(2.65 - 7.72)
0.022
(0.003 - 0.052)
Non-Hispanic White
Population in CONUS
208
People of Color
Population in CONUS
201
Highest 5% of
Baseline
PM2.5
Concentration
s (n-2,391)
Population in Highest 5%
(% of Total in CONUS)
144
(35%)
9.03
(7.80 - 12.07)
8.99
(7.76 - 12.01)
0.044
(0.008 -0.097)
Non-Hispanic White
(% of NH-W in CONUS)
51
(25%)
People of Color
(% of POC in CONUS)
93
(46%)
Remaining
95% of
Baseline
PM2.5
Concentration
s (n=45,404)
Population in Remaining 95%
(% of Total in CONUS)
265
(65%)
5.03
(2.62 - 7.24)
5.01
(2.62 - 7.20)
0.020
(0.003 - 0.049)
Non-Hispanic White
(% of NH-W in CONUS)
156
(75%)
People of Color
(% of POC CONUS)
108
(54%)
In Table 6-9 and Table 6-10, we looked at populations above and below 200% of the federal
poverty line. Using 2045 population estimates, 126 million people are projected to live below
200% of the poverty line, 13 million (10 percent) of whom would also be living in an area with
the worst baseline concentrations of ozone. Similarly, 10 percent of people projected to live
above 200% of the poverty line would also be living in an area with the worst baseline
concentrations of ozone. For PM2.5, 37 percent of those living below 200% of the poverty line
would also be living in areas with the worst baseline concentrations of PM2.5, compared to 35
percent of the population above 200% of the poverty line projected to live in those same areas.
While some disparity exists for PM2.5, overall, the results for those above and below 200% of the
poverty line are not as pronounced in the areas with the worst air quality as they are for race and
ethnicity.
309
-------
Table 6-9 Demographic Analysis of Projected 2045 Ozone Reductions (ppb) from the Proposed Rule, by
Poverty Status
Seasonal Average Ozone Concentrations in ppb
(5% to 95% Range)
2045 Population - millions
Baseline
Control
Reduction
All 12km x
12km Grid
Cells in
CONUS
(n=47,795)
Total Population in
CONUS - All Grid Cells
409
39.2
(29.9 -49.9)
38.7
(29.4 - 49.5)
0.468
(0.157 -0.898)
Population Below 200% of
the Poverty Line in CONUSa
126
Population Above 200% of
the Poverty Line in CONUS
283
Highest 5% of
Baseline
Ozone
Concentrations
(n=2,391)
Population in Highest 5%
(% of Total in CONUS)
39
(10%)
52.6
(50.0 - 58.2)
51.9
(49.6 - 57.2)
0.640
(0.273 - 1.441)
Population Below 200%
of the Poverty Line
(% Below 200% in CONUS)
13
(10%)
Population Above 200%
of the Poverty Line
(% Below 200% in CONUS)
26
(10%)
Remaining
95% of
Baseline
Ozone
Concentrations
(n=45,404)
Population in Lowest 95% (%
of Total in CONUS)
370
(90%)
38.5
(29.8 - 48.7)
38.0
(29.3 -48.3)
0.457
(0.156-0.866)
Population Below 200%
of the Poverty Line
(% Below 200% in CONUS)
112
(90%)
Population Above 200%
of the Poverty Line
(% Below 200% in CONUS)
257
(90%)
a Note that the poverty measure used here is based on ACS 5-year estimates from 2015 to 2019 at the county
level representing the fraction of county-level population below and above 200% of the poverty line. Counts of 2045
population reflect projections based on 2010 Census Data and population growth factors estimated by Woods &
Poole (2015). Measures of "current" poverty are not necessarily predictors of future poverty status.
310
-------
Table 6-10 Demographic Analysis of Projected 2045 PM2.5 Reductions (jig/m3) from the HD 2027 Proposed
Rule, by Poverty Status
Annual Average PM2.5 Concentrations in jig/m3
(5% to 95% Range)
2045 Population - millions
Baseline
Control
Reduction
All 12km x 12km
Grid Cells in
CONUS
(n=47,795)
Total Population in
CONUS - All Grid Cells
409
5.23
(2.65 - 7.76)
5.21
(2.65 - 7.72)
0.022
(0.003 - 0.052)
Population Below 200% of
the Poverty Line in CONUSa
126
Population Above 200% of
the Poverty Line in CONUS
283
Highest 5% of
Baseline PM2 5
Concentrations
(n=2,391)
Population in Highest 5%
(% of Total in CONUS)
144
(35%)
9.03
(7.80 - 12.07)
8.99
(7.76 - 12.01)
0.044
(0.008 -0.097)
Population Below 200%
of the Poverty Line
(% Below 200% in CONUS)
46
(37%)
Population Above 200%
of the Poverty Line
(% Below 200% in CONUS)
98
(35%)
Remaining 95%
of Baseline PM2 5
Concentrations
(n=45,404)
Population in Lowest 95% (%
of Total in CONUS)
265
(65%)
5.03
(2.62 - 7.24)
5.01
(2.62 - 7.20)
0.020
(0.003 - 0.049)
Population Below 200%
of the Poverty Line
(% Below 200% in CONUS)
79
(63%)
Population Above 200%
of the Poverty Line
(% Below 200% in CONUS)
185
(65%)
a Note that the poverty measure used here is based on ACS 5-year estimates from 2015 to 2019 at the county
level representing the fraction of county-level population below and above 200% of the poverty line. Counts of 2045
population reflect projections based on 2010 Census Data and population growth factors estimated by Woods &
Poole (2015). Measures of "current" poverty are not necessarily predictors of future poverty status.
The results of this demographic analysis are dependent on the available input data and its
associated uncertainty. As we note in both the air quality modeling and health benefits chapters,
uncertainties exist along the entire pathway from emissions to air quality to population
projections and exposure. The demographic analysis (including poverty status) is subject to these
same sources of uncertainty.
311
-------
A key source of uncertainty is the accuracy of the projected baseline concentrations of PM2.5
and ozone because we use modeled 2045 baseline air quality as the basis for our comparisons.
Assumptions that influence projections of future air quality include emissions in the future
baseline (stationary source emissions are only projected out to year 2028 and held constant out to
2045) and the meteorology used to model air quality (2016 conditions). With this uncertainty in
mind, we prefer to examine the air quality impacts of the proposed standards by comparing the
baseline scenario to the control scenario in order to highlight incremental changes in air quality
due to the proposed standards. By looking at the incremental change, any underlying uncertainty
present in both the modeled baseline and control air quality data is largely offset. However, when
we rank grid cells from dirtiest to cleanest using 2045 baseline concentrations, the uncertainties
associated with the baseline take on greater importance when interpreting the results of the
analysis.
There is also inherent uncertainty in the populations projected out to 2045. The projections
take into account patterns of economic growth and migration, and to the extent these patterns and
assumptions vary over time, so too will the projections of population. We attempted to address
some of this uncertainty by compiling race and ethnicity into two broad categories, "people of
color" and "NH-White," to avoid overly precise interpretations of inherently uncertain
projections of population and demographics. The Agency continues to investigate how best to
incorporate population projections into our analyses to disaggregate populations of concern by
relevant socioeconomic variables, and to identify the interactions between demographic changes
and air quality changes. The measure of poverty status used in this analysis is based on data from
the American Community Survey representing the rate of poverty between 2015-2019 and is not
projected to reflect poverty status in the future. This assumption is inherently uncertain, since
measures of "current" poverty are not necessarily predictors of future poverty status.
Finally, we note that the control scenario we modeled to support the air quality, benefits, and
demographic analyses is slightly different than the proposed Option 1 standards, useful life, and
warranty provisions presented in Sections III and IV. Chapter 5.4 of the draft RIA has more
detail on the differences between the air quality control scenario and the proposed Option 1
inventories.
312
-------
Chapter 7 Program Costs
In this chapter, EPA presents estimates of the costs associated with the emissions-reduction
technologies that manufacturers could add in response to the proposed rule. We present these not
only in terms of the upfront technology costs per engine as presented in Chapter 3 of this draft
RIA, but also how those costs would change in the years following implementation. We also
present the costs associated with the proposed program elements of extended regulatory useful
life and warranty. These technology costs are presented in terms of direct manufacturing costs
and associated indirect costs such as warranty and research and development (R&D). The
analysis also includes estimates of the possible operating costs associated with the proposed
changes—the addition of new technology and extension of warranty and useful life periods. All
costs are presented in 2017 dollars consistent with AEO 2018 unless noted otherwise.
The costs presented here are grouped into three main categories, as described below:
1) Technology Package Costs: these are the direct costs of new or modified technology—that
EPA projects manufacturers would add—and the associated indirect costs that would be
involved with bringing those technologies to market (research, development, warranty,
etc.). In our analysis, these costs, while first incurred by manufacturers of new engines,
are presumed to be passed on to the consumers of those engines (i.e., heavy-duty truck
makers and, ultimately, their purchasers/owners).
2) Operating Costs: these are the costs associated with the truck and bus operation that are
projected to be impacted by the proposal. For example, costs associated with tire
replacement are not included since the proposal is not expected to impact tire replacement,
but costs associated with repair of emission-related components are included. These costs
are incurred by truck and bus purchasers/owners.
3) Program Costs: these are the new technology package costs and operating costs combined
(the sum of numbers 1 and 2, above). These costs represent our best estimate of the costs
to society. As such, any taxes (e.g., fuel taxes) are excluded since taxes represent a
transfer payment from one member of society to another with no net cost to society. Total
program costs under the two proposed options are presented in terms of calendar year
2045 costs, present value costs, and annualized costs (see Table 7-98 and Table
7.99) zzzzz
The cost analysis is done using a tool written in Python and contained in the docket. The
Python tool along with some documentation is contained in the docket to this rule and on our
website.345
7.1 Technology Package Costs
As noted, individual technology piece costs were presented in Chapter 3. Those costs are, in
general, the direct manufacturing costs (DMC) estimated for the first year of proposed
implementation. Those costs are used here as a starting point in estimating program costs. As
shown in Chapter 3 and in Section 7.1.3 below, the costs associated with the proposed Option 1
zzzzz jjlc costs presented in Table 7-98 and Table 7-99 are presented again in Table ES ,which summarizes the net
benefits of the two proposed options.
313
-------
are implemented in two phases—MY2027 and MY2031. Costs associated with proposed Option
2 are implemented in a single phase—MY2027. Following the year in which costs are first
incurred for each phase, we have applied a learning effect to represent the cost reductions
expected to occur via the "learning by doing" phenomenon.346 This provides a year-over-year
cost for each technology as applied to new engine sales. We have then applied industry standard
"retail price equivalent (RPE)" markup factors industry-wide, with adjustments discussed below,
to estimate indirect costs associated with each technology. Both the learning effects applied to
direct costs and the application of markup factors to estimate indirect costs are consistent with
the cost estimation approaches used in EPA's past transportation-related regulatory programs.347
The sum of the direct and indirect costs represents our estimate of technology costs per vehicle
on a year-over-year basis where MY2031 and later costs include costs associated with the
MY2027 and later. These technology costs multiplied by estimated sales then represent the total
technology costs associated with the proposed and proposed alternative standards.
This cost calculation approach presumes that the expected technologies would be purchased
by original equipment manufacturers (OEMs) from their suppliers. So, while the DMC estimates
include the indirect costs and profits incurred by the supplier, the indirect cost markups we apply
cover the indirect costs incurred by OEMs to incorporate the new technologies into their vehicles
and to cover profit margins typical of the heavy-duty truck industry. We discuss the indirect
costs markups in more detail in Section 7.1.2.
As noted in the introductory text to this chapter, these technology package costs (both direct
and indirect), while first incurred by manufacturers of new engines, are presumed to be passed
on to the consumers of those engines (i.e., heavy-duty truck makers and, ultimately, their
purchasers/owners).
7.1.1 Direct Manufacturing Costs
To produce a unit of output, manufacturers incur direct and indirect costs. Direct costs include
cost of materials and labor costs. Indirect costs are discussed in the following section. The direct
manufacturing costs presented here include individual technology costs for emission-related
engine components and for exhaust aftertreatment systems (EAS).
Notably, for this analysis we include not only the marginal increased costs associated with the
proposed Options 1 or 2, but also the emission control system costs for the "no action" baseline
case (Table 7-5 and Table 7-6),AAAAAA Throughout this discussion we refer to baseline
technology costs, or baseline costs, which are meant to reflect our cost estimate of engine
systems—that portion that is emission-related—and the exhaust aftertreatment costs absent the
impacts of this proposed rule. This inclusion of baseline system costs contrasts with EPA's
approach in recent Greenhouse Gas rules or the light-duty Tier 3 criteria pollutant rule where we
estimated costs relative to a "no action" baseline case, which obviated the need to estimate
baseline costs. We have included baseline costs in this analysis because under both of the
proposed options emissions warranty and regulatory useful life provisions would be expected to
have some impact on not only the new technology added to comply with the proposed rule, but
aaaaaa §ee ciiapter 5 for more information about the baseline and how that baseline is characterized. For this cost
analysis, the baseline, or no action, case consists of engines and emission control systems meeting 2019 era criteria
emission standards but in 2027 and later model years. Why we include costs for the no action case is described in
this section.
314
-------
also on any existing emission control systems (See Chapter 3 for more details on proposed
Emissions Warranty and Regulatory Useful Life).BBBBBB The baseline direct manufacturing costs
detailed below are thus meant to reflect that portion of baseline case engine hardware and
aftertreatment systems for which new indirect costs would be incurred due to the proposed
warranty and useful life provisions, even absent any changes in the level of emission standards.
We have estimated the baseline engine costs based on recently completed studies by the
International Council on Clean Technology (ICCT) as discussed in more detail below. The
baseline EAS costs were presented in Chapter 3 of this draft RIA. The estimated marginal
technology costs associated with the proposed Options 1 and 2 were also presented in Chapter 3
of this draft RIA.
As noted, the costs shown in Table 7-5 and Table 7-6 include costs for the baseline
case.cccccc For the baseline diesel engine-related costs associated with emission control (i.e., a
portion of the fuel system, the EGR system, etc.), we have relied on a white paper done by the
International Council on Clean Transportation (ICCT) entitled, "Costs of Emission Reduction
Technologies for Heavy-Duty Diesel Vehicles."348 In Table 14 of that paper, ICCT presented
technology costs to meet U.S. standards at different stages for a 12L engine. The different stages
of U.S. standards were the 1998, 2004, 2007 and 2010 standards. Relevant portions of ICCT's
Table 14—those portions associated with engine-related technologies—are shown in Table 7-1.
For the fuel system and the turbo charger, ICCT shows only 50 percent of the total cost and
states in the text that, for components that serve other purposes in addition to emission control,
only 50% of the cost is considered in their analysis.349 ICCT notes that their costs are likely
conservative since they do not consider learning effects applied to the cost estimates associated
with each regulatory stage. Lastly, ICCT notes that their cost estimates are stated in 2015 dollars.
Table 7-1: ICCT Cost Estimates of 12L Diesel Engine-Related Emission Control Costs Associated with Past
US Emission Standards (2015 dollars)
Air/fuel control and engine out emissions
US 1998
US2004
US2007
US2010
Fuel system—50% of total cost
-
376
38
41
Variable Geometry Turbo (extra cost)—50% of total cost
-
-
185
-
Exhaust Gas Recirculation (EGR) system
-
439
-
-
EGR cooling
-
108
-
-
Total for air/fuel control and engine out emissions
-
923
223
41
In this analysis, EPA has made use of these ICCT cost estimates by first doubling the fuel
system and turbo charger costs to get the full cost of those systems (i.e., to undo the halving of
those costs done by ICCT). We then added to that result the EGR costs to get a total cost of
$1,827. We have then scaled them based on engine displacement in a manner consistent with our
approach to estimating exhaust aftertreatment costs (EAS, see Chapter 3 of this draft RIA). The
engine displacements used in our EAS cost estimates were 7, 8 and 13 liters for light, medium
bbbbbb j]le proposed warranty and useful life provisions would increase costs not only for the marginal technology
added in response to the proposal, but also for the technology to which the new technology is added because the
proposed warranty and useful life provisions would apply to the emission-control system, not just the marginal
technology added in response to the proposed standards.
cccccc gee R[A Chapters 1.1 and 1.2 for more information on emission control technologies available on
current, or baseline, engines.
315
-------
and heavy heavy-duty engines, respectively. We have estimated the class 2b and 3 engines as
equivalent to the light heavy-duty (7L) and the urban bus engines as equivalent to the medium-
duty (8L) engines. The resultant diesel engine-related costs used in this analysis are shown in
Table 7-2.
Table 7-2: Diesel Engine-Related Emission Control System Costs in the "No Action" Baseline*
LHD2b3
LHD45
MHD67
HHD8
Urban bus
Engine displacement
7
7
8
13
8
Displacement based scalar
7/12=0.58
7/12=0.58
8/12=0.67
13/12=1.08
8/12=0.67
Baseline cost, 2015 dollars
(1,827 times Displacement based
scalar)
$1,066
$1,066
$1,218
$1,979
$1,218
Baseline cost, 2017 dollars
(1.03 GPD deflator Baseline cost
in 2015 dollars)*
$1,097
$1,097
$1,254
$2,038
$1,254
* See Table 7-8 and associated text for information on the GDP deflators used in this analysis; costs shown are by
MOVES regulatory class; there are a small number of diesel engines used in LHD2b3 that are engine (rather than
chassis) certified and are, therefore, expected to incur costs associated with the proposed rule.
For the baseline gasoline engine-related costs associated with emission control (i.e., a portion
of the fuel system, etc.), we have relied on a white paper done by ICCT entitled, "Estimated Cost
of Emission Reduction Technologies for Light-Duty Vehicles."350 In Table 4-10 of that paper,
ICCT presented technology costs to meet U.S. light-duty Tier 2, Bin 5 for a 4.5L engine. The
ICCT estimate shown was $306. ICCT notes that their cost estimates are stated in 2011 dollars.
In this analysis, EPA has made use of this ICCT cost estimate by scaling them based on
engine displacement from the 4.5L light-duty displacement assumed by ICCT to a more typical
gasoline HD engine displacement of 7L. The resultant gasoline engine-related costs used in this
analysis are shown in Table 7-3.
Table 7-3: Gasoline Engine-Related Emission Control System Costs in the "No Action" Baseline*
LHD45
MHD67
HHD8
Engine displacement
7
7
7
Displacement based scalar
7/4.5=1.56
7/4.5=1.56
7/4.5=1.56
Baseline cost, 2011 dollars
(306 times Displacement based scalar)
$476
$476
$476
Baseline cost, 2017 dollars
(1.099 GPD deflator Baseline cost in 2011 dollars)*
$523
$523
$523
* See Table 7-8 and associated text for information on the GDP deflators used in this analysis; note that there are no
engine certified LHD2b3 gasoline engines and, therefore, none are expected to incur costs associated with this
proposed rule.
For the baseline CNG engine-related costs associated with emission control (a portion of the
fuel system, etc.), we have relied on the ICCT baseline gasoline costs presented in Table 7-3, but
have scaled those costs based on more typical diesel engine displacements because CNG engines
tend to be converted diesel engines but with fuel systems more typical of gasoline engines. The
diesel engine displacements used for scaling gasoline costs were presented in Table 7-2. The
resultant baseline engine-related CNG costs are shown in Table 7-4.
316
-------
Table 7-4: CNG Engine-Related Emission Control System Costs in the "No Action" Baseline*
HHD8
Urban bus
Engine displacement
12
9
Displacement based scalar
12/4.5=2.67
8/4.5=2.0
Baseline cost, 2011 dollars
(306 times Displacement based scalar)
$816
$612
Baseline cost, 2017 dollars
(1.099 GPD deflator * Baseline cost in 2011 dollars)*
$896
$672
* See Table 7-8 and associated text for information on the GDP deflators used in this analysis.
For cylinder deactivation costs under the proposed Options 1 or 2, we have used FEV-
conducted teardown-based cylinder deactivation costs as presented in Chapter 3 of this draft
RIA.351 The marginal technology costs for exhaust aftertreatment components—also detailed and
presented in Chapter 3 of this draft RIA—are based on an ICCT methodology with extensive
revision by EPA. As discussed in draft RIA Chapter 3, we also have EAS cost estimates from a
recent FEV-conducted teardown study.352 These teardown-based estimated EAS costs are
similar to the EPA-estimated costs and we may use those FEV-study teardown-based cost
estimates in the final rule.
The cost basis (the year dollars) for many of these costs were also presented in a mixed set of
cost basis years. For the cost analysis presented here, we use 2017 dollars throughout the
analysis for consistency with the AEO 2018 data upon which our MOVES emissions inventory
results are based. The costs presented in Chapter 3 are repeated in Table 7-5 for diesel regulatory
classes, in Table 7-6 for gasoline regulatory classes, and in Table 7-7 for CNG regulatory classes
with the exception that all values presented here are updated to a consistent 2017 dollar
basis.DDDDDD See Chapter 5.2 for a discussion of regulatory classes. Table 7-8 shows the gross-
domestic product price deflators used to adjust to 2017 dollars. Note that we have estimated costs
for regulatory classes that exist in our MOVES runs (see Chapter 5 of this draft RIA) to remain
consistent with the inventory impacts we have estimated. Note also that, throughout this section,
LHD=light heavy-duty, MHD=medium heavy-duty, HHD=heavy heavy-duty, CDPF=catalyzed
diesel particulate filter, DOC=diesel oxidation catalyst, SCR=selective catalytic reduction,
HC=hydrocarbon, CNG=compressed natural gas.
dddddd jjlc MY2019 engine and aftertreatment costs estimates presented in draft RIA Chapters 3.1.5 and 3.2.3 are
used as the MY2027 baseline cost in the tables in this draft RIA Chapter 7.1.1.
317
-------
Table 7-5 Diesel Technology and Package Direct Manufacturing Costs per Engine by Regulatory Class, 2017
dollars*
MOVES
Regulatory Class
Technology
Baseline
Proposed Option 1
(MY2027 increment
to Baseline)
Proposed Option 2
(MY2027 increment to
Baseline)
LHD2b3
LHD Package
$3,788
$1,616
$1,616
Engine hardware
$1,097
$0
$0
Closed crankcase
$0
$0
$0
Cylinder deactivation
$0
$196
$196
CDPF
$504
$0
$0
DOC
$350
$0
$0
SCR
$1,837
$1,174
$1,174
Canning
$0
$30
$30
HC dosing
$0
$216
$216
LHD45
LHD45 Package
$3,806
$1,653
$1,653
Engine hardware
$1,097
$0
$0
Closed crankcase
$18
$37
$37
Cylinder deactivation
$0
$196
$196
CDPF
$504
$0
$0
DOC
$350
$0
$0
SCR
$1,837
$1,174
$1,174
Canning
$0
$30
$30
HC dosing
$0
$216
$216
MHD67
MHD67 Package
$4,032
$1,619
$1,619
Engine hardware
$1,254
$0
$0
Closed crankcase
$18
$37
$37
Cylinder deactivation
$0
$147
$147
CDPF
$570
$0
$0
DOC
$316
$0
$0
SCR
$1,875
$1,183
$1,183
Canning
$0
$36
$36
HC dosing
$0
$216
$216
HHD8
HHD8Package
$6,457
$2,210
$2,210
Engine hardware
$2,038
$0
$0
Closed crankcase
$18
$37
$37
Cylinder deactivation
$0
$206
$206
CDPF
$1,067
$0
$0
DOC
$585
$0
$0
SCR
$2,750
$1,681
$1,681
Canning
$0
$71
$71
HC dosing
$0
$216
$216
Urban bus
Urban bus Package
$4,082
$1,653
$1,653
Engine hardware
$1,254
$0
$0
Closed crankcase
$18
$37
$37
Cylinder deactivation
$0
$147
$147
CDPF
$567
$0
$0
DOC
$314
$0
$0
SCR
$1,929
$1,469
$1,469
Canning
$0
$0
$0
HC dosing
$0
$0
$0
* Note that the analysis uses the baseline plus the alternative cost~i.e., Baseline+Proposed Option 1, or
Baseline+Proposed Option 2 when estimating total costs; the incremental costs are shown here for ease of
understanding the increased costs associated with each option.
318
-------
Table 7-6 Gasoline Technology and Package Direct Manufacturing Costs per Engine by Regulatory Class,
2017 dollars*
MOVES
Regulatory Class
Technology
Baseline
Proposed Option
1
(MY2027
increment to
Baseline)
Proposed Option
2
(MY2027
increment to
Baseline)
LHD45
LHD45 Package
$832
$417
$417
Engine hardware
$523
$0
$0
Aftertreatment
$309
$393
$393
ORVR
$0
$24
$24
MHD67
MHD67 Package
$832
$417
$417
Engine hardware
$523
$0
$0
Aftertreatment
$309
$393
$393
ORVR
$0
$24
$24
HHD8
HHD8 Package
$832
$417
$417
Engine hardware
$523
$0
$0
Aftertreatment
$309
$393
$393
ORVR
$0
$24
$24
* Note that the analysis uses the baseline plus the alternative cost~i.e., Baseline+Proposed Option 1, or
Baseline+Proposed Option 2 when estimating total costs; the incremental costs are shown here for ease of
understanding the increased costs associated with each option.
Table 7-7: CNG Technology and Package Direct Manufacturing Costs per Engine by Regulatory Class, 2017
dollars*
MOVES
Regulatory
Class
HHD8 Package
Baseline
Proposed Option 1
(MY2027
increment to
Baseline)
Proposed Option 1
(MY2030
increment to
MY2027)
Proposed Option 2
(MY2027
increment to
Baseline)
HHD8
HHD8 Package
$4,108
$27
$27
$27
Engine hardware
$896
$0
$0
$0
Aftertreatment
$3,212
$27
$27
$27
Urban bus
Urban bus
Package
$3,081
$19
$19
$19
Engine hardware
$672
$0
$0
$0
Aftertreatment
$2,409
$19
$19
$19
* Note that the analysis uses the baseline plus the alternative cost~i.e., Baseline+Proposed Option 1, or
Baseline+Proposed Option 1 or Baseline+Alternative2 when estimating total costs; the incremental costs are shown
here for ease of understanding the increased costs associated with each option.
Table 7-8: GDP Price Deflators* Used to Adjust Costs to 2017 Dollars
Cost Basis Year
Conversion Factor
2011
1.099
2015
1.030
2017
1.000
2018
0.977
2019
0.959
* Based on the National Income and Product Accounts, Table 1.1.4 Price Indexes for Gross Domestic Product,
Bureau of Economic Analysis, U.S. Department of Commerce, December 20, 2019.
319
-------
The direct costs are then adjusted to account for learning effects going forward from the first
year of implementation. To make that adjustment, the following equation is used.353
Where,
yt = cost (or price) of a given item at time t
xt = cumulative production of a given item at time t
xt+i = cumulative production of a given item at time t+1
b = the learning rate
In this cost analysis, EPA makes an adjustment to this standard formula by inserting a "seed
volume factor" meant to represent the number of years of sales of a technology leading to
learning prior to the year for which the technology's cost estimate is intended (model years 2027
or 2030 as indicated in Table 7-5 and Table 7-6). In other words, manufacturers may sell some of
the systems expected for compliance with the proposed rule in years prior to the first year of the
new standards, thereby learning and realizing cost reductions in that time. The value of this seed
volume factor might be 0 to represent a new, unsold technology, or any value >0 to represent a
relatively new technology having had some sales in prior years. A seed volume factor of 0 places
the technology at the beginning of the learning curve and, therefore, subsequent learning effects
(i.e., cost reductions) will be most rapid in the years immediately following the first year of
implementation. An increasing seed volume factor serves to move the effects of learning further
along the curve making subsequent learning effects less dramatic in the years immediately
following the first year of the analysis. Figure 7-1 shows the effect of the seed volume factor on
the levels of learning applied to the direct manufacturing costs assuming constant sales year-
over-year and a learning rate of _o.245.EEEEEE
eeeeee [n effect; the "seed volume factor" sets the cumulative number of units produced by an organization, which is
a required data point to estimate the conventional form of the learning curve. Throughout this analysis, we have used
a learning rate of -0.245 as developed for EPA by ICF and a foremost Subject Matter Expert in, "Cost Reduction
through Learning in Manufacturing Industries and in the Manufacture of Mobile Sources, Final Report and Peer
Review Report," EPA-420-R-16-018, November 2016.
320
-------
1.200
u 1.000
TO
£ 0.800
o 0.600
4—"
CD
'I 0.400
~ai
% 0.200
l/l
o
u 0.000
yearl year2 year3 year4 year5 year6 year7 year8 year9
— — SeedVolumeFactor=0 SeedVolumeFactor=3
Figure 7-1: Costs Relative to First Year Costs using Different Seed Volume Factors
In the end, the learning effects are calculated using the following formula.
-(
b
yt = l7T^- , , , csw„, yt=0
\Salest=0 + (Salest=0 * SeedVolumeFactor) J
xt + (Salest=0 * SeedVolume Factor) \
Where,
b = the learning rate (-0.245 in this analysis)
yt=o = estimated direct cost in the first year of implementation (e.g., MY2027 or MY2030)
Salesto = sales in the first year of implementation (e.g., MY2027 or MY2030)
SeedVolumeFactor = 0 or greater to represent the number of years of learning already having occurred on a
technology
xt = the cumulative sales of vehicles complying with the new standard (equal to first year sales in the first
year of implementation, first year sales plus second year sales in the second year of implementation, etc.)
To illustrate the seed volume factor, if we assume the learning rate, b, is -0.245 (the value
used in this analysis), the direct cost in the first year of implementation, yt=o, is $100, the sales in
the first year of implementation, Salest=o, is 1000 engines, and the seed volume factor is 0 (i.e.,
no learning having occurred prior to the first year of the analysis), then the cost, yt, would be
$100. This is because the cumulative production in year t, xt, equals Salest=o in the first year of
implementation leaving the formula as:
(1) —0-245 # $10() = $100
If the sales in the following year were an additional 1000 engines, the cost would decrease to
$84 since xt would now be 2000 and the formula would be:
/2000\-° 245
VioooJ
$100 = $84
321
-------
If the seed volume factor is set to 3 (i.e., to approximate 3 years of learning prior to the first
year of the analysis, then the cost in 2027 would again be $100 since the formula would be
/1000 + (3 * 1000)\~0245
- *$100 = $100
V1000 + (3 * 1000)/
However, in 2028, the formula would be
/2000 + (3 * 1000)\~°245
- *$100 = $95
V1000 + (3 * 1000)/
With 3 years of learning estimated to have already occurred, the costs going forward from the
first year reduce at a slower pace than in the previous example. Therefore, increasing the seed
volume factor results in less rapid learning effects going forward. The seed volume factors used
in this analysis are shown in Table 7-9. A factor of 10 has been used for the baseline
technologies since those technologies will have undergone considerable learning by 2027. We
have used a factor of 3 for the proposed Options 1 and 2 to reflect at least 3 years of sales with
technologies very similar to those expected under the proposed options thereby resulting in
conservative learning-based cost reductions moving forward from 2027.
Table 7-9: Seed volume factors used in this analysis for each Option
Fuel
Regulatory Class
Baseline
Proposed Option 1
(MY2027 increment to Baseline)
Proposed Option 2
(MY2027 increment to Baseline)
LHD
10
3
3
LHD45
10
3
3
Diesel
MHD67
10
3
3
HHD8
10
3
3
Urban Bus
10
3
3
LHD45
10
3
3
Gasoline
MHD67
10
3
3
HHD8
10
3
3
CNG
HHD8
10
3
3
Urban Bus
10
3
3
Learning factors were applied on a technology package cost basis, and MOVES projected
sales volumes were used to determine first year sales (Salest=o) and cumulative sales (xt). The
resultant direct manufacturing costs and how those costs reduce over time are presented in both
Chapter 7.1.3 (on a per vehicle basis) and in Chapter 7.3.1 (on a total cost basis).
7.1.2 Indirect 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
322
-------
account for indirect costs allocated to a unit of goods sold. To ensure that regulatory analysis
capture the changes in indirect costs, markup factors, which relate total indirect costs to total
direct costs, have been developed and used by EPA and other stakeholders. These factors are
often referred to as retail price equivalent (RPE) multipliers. RPE multipliers provide, at an
aggregate level, the relative shares of revenues, where:
Revenue = Direct Costs + Indirect Costs
so that:
Revenue/Direct Costs = 1 + Indirect Costs/Direct Costs = RPE
and,
Indirect Costs = Direct Costs x (RPE - 1).
If the relationship between revenues and direct costs (i.e., RPE) can be shown to equal an
average value over time, then an estimate of direct costs can be multiplied by that average value
to estimate revenues, or total costs. Further, that difference between estimated revenues, or total
costs, and estimated direct costs can be taken as the indirect costs. Cost analysts and regulatory
agencies have frequently used these multipliers to predict the resultant impact on costs associated
with manufacturers' responses to regulatory requirements and we are using that approach in this
analysis.
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. In the past, EPA has expressed a concern with using the RPE multiplier for all
technologies, because it is not likely that the indirect costs of vehicle modifications are the same
for all technologies.354 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. EPA developed an alternative indirect cost methodology—the Indirect
Cost Multiplier (ICM)--to address those concerns.355
The cost of different technologies was an important distinction in modeling efforts supporting
EPA's greenhouse gas rulemakings (GHG) since a variety of GHG technologies were available
to manufacturers and EPA wanted to project which combinations of technologies were most cost
effective toward achieving compliance. For this proposed rule, we do not have that consideration
since we are projecting the same technologies for all vehicles within a given regulatory class and
are not attempting to project from a variety of technologies which are most cost effective toward
achieving compliance. For that reason, EPA is using the RPE approach to estimate indirect costs
in this analysis.
RPEs themselves are also 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.356
323
-------
This cost analysis estimates indirect costs by applying markup factors used in past
rulemakings setting new greenhouse gas standards for heavy-duty trucks.357 The markup factors
are based on financial filings with the Securities and Exchange Commission for several engine
and engine/truck manufacturers in the heavy-duty industry as detailed in a study done by RTI
International for EPA.358 The RPE factors developed by RTI for HD engine manufacturers, HD
truck manufacturers and for the HD truck industry as a whole are shown in Table 7_io.ffffff
Also shown in Table 7-10 are the RPE factors developed by RTI for light-duty vehicle
manufacturers.359
Table 7-10: Retail Price Equivalent Factors in the Heavy-Duty and Light-Duty Industries
Cost Contributor
HD Engine
Manufacturer
HD Truck
Manufacturer
HD Truck
Industry
LD Vehicle
Industry
Direct manufacturing cost
1.00
1.00
1.00
1.00
Warranty
0.02
0.04
0.03
0.03
R&D
0.04
0.05
0.05
0.05
Other (admin, retirement, health, etc.)
0.17
0.22
0.29
0.36
Profit (cost of capital)
0.05
0.05
0.05
0.06
RPE
1.28
1.36
1.42
1.50
For this analysis, EPA has based cost estimates for diesel and CNG regulatory classes on the
HD Truck Industry values shown in Table 7-10. Because most of the proposed changes apply to
engines, we first considered using the HD Engine Manufacturer values. However, the industry is
becoming more vertically integrated and the costs we are trying to estimate are those that occur
at the end purchaser, or retail, level. For that reason, we believe that the truck industry values
represent the factors of most interest to this analysis. For gasoline regulatory classes, we have
used the LD Vehicle Industry values shown in Table 7-10. We have chosen those values since
they more closely represent the cost structure of manufacturers in that industry—Ford, General
Motors, and Fiat Chrysler.
Of the cost contributors listed in Table 7-10, Warranty and R&D are the elements of indirect
costs that the proposed requirements are expected to impact. As discussed in Chapter 3 of this
draft RIA, EPA is proposing to lengthen the warranty period, which we expect to increase the
contribution of warranty costs to the RPE. EPA is also proposing to extend the regulatory useful
life, which we expect to result in increased R&D expenses as systems are developed to deal with
the longer life during which compliance with standards would be required. We would expect that
the minor OBD-related R&D efforts being proposed—an expanded list of public OBD parameters
and expanded scan tool tests discussed in Section IV.C of the preamble—could be done given the
R&D estimated here. Profit is listed to highlight that profit is being considered and included in
the analysis. All other indirect cost elements—those encapsulated by the "Other" category,
including General and Administrative Costs, Retirement Costs, Healthcare Costs, and other
ffffff j]le engine manufacturers included were Hino and Cummins; the truck manufacturers included were
PACCAR, Navistar, Daimler and Volvo. Where gaps existed such as specific line items not reported by these
companies due to differing accounting practices, data from the Heavy Duty Truck Manufacturers Industry Report by
Supplier Relations LLC (2009) and Census (2009) data for Other Engine Equipment Manufacturing Industry
(NAICS 333618) and Heavy Duty Truck Manufacturing Industry (NAICS 336120) were used to fill the gaps. This is
detailed in the study report at Appendix A. 1.
324
-------
overhead costs—as well as Profits, are expected to scale according to their historic levels of
contribution.
As mentioned, Warranty and R&D are the elements of indirect costs that the proposed
requirements are expected to impact. Warranty expenses are the costs that a business expects to
or has already incurred for the repair or replacement of goods that it has sold. The total amount
of warranty expense is limited by the warranty period that a business typically allows. After the
warranty period for a product has expired, a business no longer incurs a warranty liability; thus, a
longer warranty period results in a longer period of liability for a product. At the time of sale,
companies are expected to set aside money in a warranty liability account to cover any potential
future warranty claims. If and when warranty claims are made by customers, the warranty
liability account is debited and a warranty claims account is credited to cover warranty claim
expenses.
To address the expected increased indirect cost contributions associated with warranty
(increased funding of the warranty liability account) due to the proposed longer warranty
requirements, we have applied scaling factors commensurate with the changes in proposed
Options 1 or 2 to the number of miles included in the warranty period (i.e., VMT-based scaling
factors). For example, the current required emission warranty periods for Class 8 diesel trucks
are 5 years or 100,000 miles. Proposed Option 1 would extend the required warranty period for a
Class 8 diesel to 7 years or 450,000 miles for MYs 2027 through 2030, and then extend further to
10 years or 600,000 miles for MYs 2031 and beyond. As such, in our analysis of proposed
Option 1 for Class 8 diesel trucks we applied a scaling factor of 4.5 (450/100) to the 0.03
warranty contribution factor for MYs 2027 through 2030, and applied a scaling factor of 6.0
(600/100) for MYs 2031 and later. This same approach is followed for the other regulatory
classes, and for our analysis of proposed Option 2.
Similarly, for R&D on that same Class 8 truck, the proposed Option 1 would extend
regulatory useful life from 10 years or 435,000 miles to 11 years or 600,000 miles beginning in
MY2027, and then extend further to 12 years or 800,000 miles for MYs 2031 and later, we have
applied a scaling factor of 1.38 (600/435) to the 0.05 R&D contribution factor for MYs 2027
through 2029, and then 1.33 (800/600) for MYs 2031 through 2033. Notice the different
treatment of the scaling factors for R&D versus those for warranty. We would expect that once
the development efforts into longer useful life is complete, increased expenditures would return
to their normal levels of contribution. As such, we have implemented the R&D scalars in three-
year increments (2027 through 2029 under proposed Options 1 and 2, and then 2031 through
2033 under proposed Option 1). In MY2034 and later (under the proposed Option 1) and in
MY2029 and later in proposed Options 1 and 2, the R&D scaling factor would no longer be
applied.
The VMT-based scaling factors applied to warranty and R&D cost contributers used in our
cost analysis of proposed Options 1 and 2 are shown in Table 7-11 for diesel and CNG
regulatory classes and in Table 7-12 for gasoline regulatory classes.
325
-------
Table 7-11: Scaling Factors Applied to RPE Contribution Factors to Reflect Changes in their Contributions,
Diesel & CNG Regulatory Classes*
Warranty Scalars
R&D Scalars
Scenario
MOVES
Regulatory Class
MY2027
through
MY2030
MY2031+
MY2027
through
MY2029
MY2031
through
MY2033
MY2034+
Baseline
LHD
1.00
1.00
1.00
1.00
1.00
LHD45
1.00
1.00
1.00
1.00
1.00
MHD67
1.00
1.00
1.00
1.00
1.00
HHD8
1.00
1.00
1.00
1.00
1.00
Urban Bus
1.00
1.00
1.00
1.00
1.00
Option
1
LHD
1.50
2.10
1.73
1.42
1.00
LHD45
1.50
2.10
1.73
1.42
1.00
MHD67
2.20
2.80
1.46
1.30
1.00
HHD8
4.50
6.00
1.38
1.33
1.00
Urban Bus
4.50
6.00
1.38
1.33
1.00
Option
2
LHD
1.10
1.10
2.27
1.00
1.00
LHD45
1.10
1.10
2.27
1.00
1.00
MHD67
1.50
1.50
1.76
1.00
1.00
HHD8
3.50
3.50
1.49
1.00
1.00
Urban Bus
3.50
3.50
1.49
1.00
1.00
Table 7-12: Scaling Factors Applied to RPE Contribution Factors to Reflect Changes in their Contributions,
Gasoline Regulatory Classes
Warranty Scalars
R&D Scalars
Scenario
MOVES
Regulatory Class
MY2027
through
MY2030
MY2031+
MY2027
through
MY2029
MY2031
through
MY2033
MY2034+
Baseline
LHD45
1.00
1.00
1.00
1.00
1.00
MHD67
1.00
1.00
1.00
1.00
1.00
HHD8
1.00
1.00
1.00
1.00
1.00
Option
1
LHD45
2.20
3.20
1.41
1.29
1.00
MHD67
2.20
3.20
1.41
1.29
1.00
HHD8
2.20
3.20
1.41
1.29
1.00
Option
2
LHD45
2.20
2.20
1.82
1.00
1.00
MHD67
2.20
2.20
1.82
1.00
1.00
HHD8
2.20
2.20
1.82
1.00
1.00
Lastly, as mentioned in section 7.1.1, the markups for estimating indirect costs are applied to
our estimates of the absolute direct manufacturing costs for emission-control technology shown
in Table 7-5, Table 7-6 and Table 7-7, not just the incremental costs associated with the proposal
(i.e., the Baseline+Proposed Option 1 and Baseline+Proposed Option 2 costs, not just the
incremental costs of proposed Option 1 or 2).GGGGGG This is an important element of the analysis
as shown by the example in Table 7-13. In the example, which is only for illustration, we assume
that the baseline technology cost is $5000, the proposed incremental cost is $1000, and the
indirect cost warranty contribution is 0.03 with a simple scalar of 1.5 associated with a longer
gggggg [ncreaseci indirect costs are included for the baseline technology because the extended warranty and useful
life provisions would impact those technologies too, not just the new, incremental technology.
326
-------
warranty period. In this case, the costs could be calculated according to two approaches as
shown. By including the baseline costs, we are estimating considerable new warranty costs in the
proposal as illustrated by the example where including baseline costs results in warranty costs of
$270 while excluding baseline costs results in warranty costs of just $45 HHHHHH
Table 7-13: Simplified Example of Indirect Warranty Costs Calculated on an Incremental vs. Absolute
Technology Package Cost (values are not from the analysis and are for presentation only)
Excluding Baseline Costs
Including Baseline Costs
Direct Manufacturing Cost (DMC)
$1000
$5000+ $1000 = $6000
Indirect Warranty Costs
$1000x0.03 x 1.5 = $45
$6000x0.03 x 1.5 = $270
DMC + Warranty
$1000+ $45 = $1045
$1000+ $270 = $1270
We are aware of a recent study conducted by the National Renewable Energy Lab (NREL) for
the California Air Resources Board (CARB).360 In that study, NREL surveyed parts suppliers
and engine/vehicle manufacturers regarding estimated costs associated with the technologies
being considered within the context of CARB's Heavy-Duty Low NOx program. As part of that
study, NREL considered costs associated with increased warranties and increased useful life
periods being considered by CARB. Our understanding is that, while the costs associated with
warranty and useful life are quite high, they were in fact estimates associated with complete
system replacement at some point during the useful life of the engine/vehicle. We have assumed
that manufacturers would not pursue such an approach and would, instead, include upfront (i.e.,
at the point of end user purchase) with the expectation that the parts would last the full useful life
without a mandatory replacement. Such has been the practice in industry by and large since
emission controls were introduced in the 1970s and 1980s. For that reason, we have chosen not
to use the warranty and useful life estimates presented by NREL and have instead used the
approach described here. CARB has more recently published their Omnibus analysis, but we
have not had time to adequately digest the details of that analysis to adequately compare it to the
analysis we have done.
7.1.3 Technology Costs per Vehicle
The following tables present the technology costs estimated for the proposed Options 1 and 2
on a per-vehicle basis for MY2027 and for MY2031. Reflected in these tables are learning
effects on direct manufacturing costs and scaling effects—associated with increased costs due to
proposed program elements—on indirect costs. The sum is also shown and reflects the cost per
vehicle in the specific model year that would be multiplied by sales to estimate the total costs
associated with each proposed option.111111 Also reflected in these tables is the phase-in of
implementation for the proposed Option 1 and the differences between the proposed options due
to different standard levels and different warranty and useful life provisions.
hhhhhh as notecj earlier, we have included baseline costs in this analysis because the proposed emissions warranty
and regulatory useful life provisions would be expected to have some impact on not only the new technology added
to comply with the proposed standards, but also on any existing emission control systems (See Chapter 3 for more
details on proposed Emissions Warranty and Regulatory Useful Life).
111111 Note that we have not estimated sales impacts associated with the proposed rule (see Chapter 10), so sales
projections are equivalent across scenarios.
327
-------
Note that, while we show costs per vehicle here, it is important to remember that these are
costs and not prices. We make no effort at estimating how manufacturers would price their
products. Manufacturers may pass costs along to purchasers via price increases in a manner
consistent with what we show here. However, manufacturers may also price certain products
higher than what we show while pricing others lower—the higher-priced products thereby
subsidizing the lower-priced products. This is true in any market, not just the heavy-duty
highway industry. This is perhaps especially true with respect to the indirect costs we have
estimated because, for example, R&D done to improve emission durability can readily transfer
across different engines but the technology added to an engine is uniquely tied to that engine.
Table 7-14: MY2027 Technology Costs for LHD2b3 Diesel, Average per Vehicle, 2017 Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost
Sum
Baseline
$3,788
$114
$189
$1,099
$189
$5,379
Baseline+Proposed Option 1
$5,404
$243
$467
$1,567
$270
$7,952
Baseline+Proposed Option 2
$5,404
$178
$614
$1,567
$270
$8,034
Option 1 increase from
Baseline
$1,616
$130
$277
$469
$81
$2,572
Option 2 increase from
Baseline
$1,616
$65
$425
$469
$81
$2,655
Table 7-15: MY2031 Technology Costs for LHD2b3 Diesel, Average per Vehicle, 2017 Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost
Sum
Baseline
$3,504
$105
$175
$1,016
$175
$4,976
Baseline+Proposed Option 1
$4,863
$306
$346
$1,410
$243
$7,168
Baseline+Proposed Option 2
$4,863
$160
$243
$1,410
$243
$6,920
Option 1 increase from
Baseline
$1,358
$201
$170
$394
$68
$2,192
Option 2 increase from
Baseline
$1,358
$55
$68
$394
$68
$1,944
Table 7-16: MY2027 Technology Costs for LHD45 Diesel, Average per Vehicle, 2017 Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost
Sum
Baseline
$3,806
$114
$190
$1,104
$190
$5,405
Baseline+Proposed Option 1
$5,459
$246
$471
$1,583
$273
$8,032
Baseline+Proposed Option 2
$5,459
$180
$620
$1,583
$273
$8,115
Option 1 increase from
Baseline
$1,653
$131
$281
$479
$83
$2,627
Option 2 increase from
Baseline
$1,653
$66
$430
$479
$83
$2,710
328
-------
Table 7-17: MY2031 Technology Costs for LHD45 Diesel, Average per Vehicle, 2017 Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost
Sum
Baseline
$3,515
$105
$176
$1,019
$176
$4,991
Baseline+Proposed Option 1
$4,900
$309
$348
$1,421
$245
$7,223
Baseline+Proposed Option 2
$4,900
$162
$245
$1,421
$245
$6,973
Option 1 increase from
Baseline
$1,385
$203
$172
$402
$69
$2,232
Option 2 increase from
Baseline
$1,385
$56
$69
$402
$69
$1,982
Table 7-18: MY2027 Technology Costs for MHD67 Diesel, Average per Vehicle, 2017 Dollars*
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost
Sum
Baseline
$4,032
$121
$202
$1,169
$202
$5,725
Baseline+Proposed Option 1
$5,651
$373
$412
$1,639
$283
$8,358
Baseline+Proposed Option 2
$5,651
$254
$496
$1,639
$283
$8,323
Option 1 increase from
Baseline
$1,619
$252
$211
$470
$81
$2,632
Option 2 increase from
Baseline
$1,619
$133
$295
$470
$81
$2,598
Table 7-19: MY2031 Technology Costs for MHD67 Diesel, Average per Vehicle, 2017 Dollars*
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost
Sum
Baseline
$3,723
$112
$186
$1,080
$186
$5,287
Baseline+Proposed Option 1
$5,080
$427
$329
$1,473
$254
$7,563
Baseline+Proposed Option 2
$5,080
$229
$254
$1,473
$254
$7,290
Option 1 increase from
Baseline
$1,357
$315
$143
$394
$68
$2,276
Option 2 increase from
Baseline
$1,357
$117
$68
$394
$68
$2,003
Table 7-20 MY2027 Technology Costs for HHD8 Diesel, Average per Vehicle, 2017 Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost
Sum
Baseline
$6,457
$194
$323
$1,873
$323
$9,169
Baseline+Proposed
Option 1
$8,668
$1,170
$598
$2,514
$433
$13,382
Baseline+Proposed
Option 2
$8,668
$910
$648
$2,514
$433
$13,172
Option 1 increase from
Baseline
$2,210
$976
$275
$641
$111
$4,213
Option 2 increase from
Baseline
$2,210
$716
$325
$641
$111
$4,003
329
-------
Table 7-21: MY2031 Technology Costs for HHD8 Diesel, Average per Vehicle, 2017 Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost
Sum
Baseline
$5,961
$179
$298
$1,729
$298
$8,465
Baseline+Proposed
Option 1
$7,813
$1,406
$521
$2,266
$391
$12,396
Baseline+Proposed
Option 2
$7,813
$820
$391
$2,266
$391
$11,680
Option 1 increase from
Baseline
$1,851
$1,227
$223
$537
$93
$3,931
Option 2 increase from
Baseline
$1,851
$641
$93
$537
$93
$3,215
Table 7-22: MY2027 Technology Costs for Urban bus Diesel, Average per Vehicle, 2017 Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost
Sum
Baseline
$4,082
$122
$204
$1,184
$204
$5,796
Baseline+Proposed Option 1
$5,734
$774
$395
$1,663
$287
$8,854
Baseline+Proposed Option 2
$5,734
$602
$428
$1,663
$287
$8,715
Option 1 increase from
Baseline
$1,653
$652
$191
$479
$83
$3,058
Option 2 increase from
Baseline
$1,653
$480
$224
$479
$83
$2,918
Table 7-23: MY2031 Technology Costs for Urban bus Diesel, Average per Vehicle, 2017 Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost
Sum
Baseline
$3,769
$113
$188
$1,093
$188
$5,352
Baseline+Proposed Option 1
$5,153
$928
$344
$1,494
$258
$8,177
Baseline+Proposed Option 2
$5,153
$541
$258
$1,494
$258
$7,704
Option 1 increase from
Baseline
$1,385
$815
$155
$402
$69
$2,825
Option 2 increase from
Baseline
$1,385
$428
$69
$402
$69
$2,353
Table 7-24: MY2027 Technology Costs for LHD45, MHD67 & HHD8 Gasoline, Average per Vehicle, 2017
Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost Sum
Baseline
$832
$25
$42
$299
$50
$1,248
Baseline+Proposed Option 1
$1,249
$82
$88
$450
$75
$1,944
Baseline+Proposed Option 2
$1,249
$82
$114
$450
$75
$1,969
Option 1 increase from Baseline
$417
$57
$46
$150
$25
$696
Option 2 increase from Baseline
$417
$57
$72
$150
$25
$722
330
-------
Table 7-25: MY2031 Technology Costs for LHD45, MHD67 & HHD8 Gasoline, Average per Vehicle, 2017
Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost Sum
Baseline
$768
$23
$38
$277
$46
$1,152
Baseline+Proposed Option 1
$1,118
$107
$72
$402
$67
$1,767
Baseline+Proposed Option 2
$1,118
$74
$56
$402
$67
$1,717
Option 1 increase from Baseline
$350
$84
$34
$126
$21
$614
Option 2 increase from Baseline
$350
$51
$17
$126
$21
$565
Table 7-26: MY2027 Technology Costs for HHD8 CNG, Average per Vehicle, 2017 Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost Sum
Baseline
$4,108
$123
$205
$1,191
$205
$5,833
Baseline+Proposed Option 1
$4,135
$558
$285
$1,199
$207
$6,384
Baseline+Proposed Option 2
$4,135
$434
$309
$1,199
$207
$6,284
Option 1 increase from Baseline
$27
$435
$80
$8
$1
$551
Option 2 increase from Baseline
$27
$311
$104
$8
$1
$450
Table 7-27: MY2031 Technology Costs for HHD8 CNG, Average per Vehicle, 2017 Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost Sum
Baseline
$3,793
$114
$190
$1,100
$190
$5,386
Baseline+Proposed Option 1
$3,816
$687
$254
$1,107
$191
$6,054
Baseline+Proposed Option 2
$3,816
$401
$191
$1,107
$191
$5,705
Option 1 increase from Baseline
$23
$573
$65
$7
$1
$668
Option 2 increase from Baseline
$23
$287
$1
$7
$1
$318
Table 7-28: MY2027 Technology Costs for Urban bus CNG, Average per Vehicle, 2017 Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost Sum
Baseline
$3,081
$92
$154
$893
$154
$4,375
Baseline+Proposed Option 1
$3,100
$419
$214
$899
$155
$4,787
Baseline+Proposed Option 2
$3,100
$326
$232
$899
$155
$4,711
Option 1 increase from Baseline
$19
$326
$60
$6
$1
$412
Option 2 increase from Baseline
$19
$233
$78
$6
$1
$336
Table 7-29: MY2031 Technology Costs for Urban bus CNG Average per Vehicle, 2017 Dollars
Scenario
DMC
Warranty
R&D
Other
Profit
Tech Cost Sum
Baseline
$2,845
$85
$142
$825
$142
$4,039
Baseline+Proposed Option 1
$2,861
$515
$191
$830
$143
$4,539
Baseline+Proposed Option 2
$2,861
$300
$143
$830
$143
$4,277
Option 1 increase from Baseline
$16
$430
$48
$5
$1
$500
Option 2 increase from Baseline
$16
$215
$1
$5
$1
$237
7.2 Operating Costs
We have estimated three impacts on operating costs associated with the proposed criteria
pollutant standards: increased diesel exhaust fluid (DEF) consumption by diesel vehicles due to
331
-------
increased DEF dose rates to enable compliance with more stringent NOx standards; decreased
fuel costs by gasoline vehicles due to new onboard refueling vapor recovery systems that allow
burning (in engine) of otherwise evaporated hydrocarbon emissions; and, emission repair
impacts. For the repair impacts we expect that the longer duration useful life period would
increase emission control system durability resulting in fewer failing parts needing repair and
that the longer warranty period would result in fewer repairs paid for by owners/operators.
However, the possibility exists that the higher-cost emission control systems may result in higher
repair costs if and when repairs are needed. We have estimated the net effect on repair costs and
describe our approach, along with increased DEF consumption and reduced gasoline
consumption below.
As noted in the introductory text to this chapter, the operating costs we estimate here are for
the heavy-duty truck operation impacted by the proposal (e.g., repair of emission-related
components). These costs (and savings) are incurred by heavy-duty truck purchasers/owners.
7.2.1 Costs Associated with Increased Diesel Exhaust Fluid (DEF) Consumption in Diesel
Engines
To estimate baseline case DEF consumption in heavy-duty vehicles with diesel engines, this
analysis uses the relationship, shown below, of DEF dose rate relative to the reduction in NOx
over the SCR catalyst.JJJJJJ-361
NOx reduction = —73.679x + 0.0149
where x is equal to the DEF dose rate. This relationship was developed giving consideration to
FTP emissions. By estimating the FTP NOx reduction across the SCR catalyst, the DEF dose
rate can be calculated. NOx reduction is estimated from the difference between estimated engine-
out and FTP tailpipe NOx emissions; these variables along with the calculated DEF dose rate for
the baseline case are shown in Table 7-30.
Table 7-30: Diesel Exhaust Fluid Consumption Rates for Diesel Vehicles in the Baseline Case
Value
Engine-out NOx
(FTP g/hp-hr)
4.0
Tailpipe NOx
(FTP mg/hp-hr)
200
DEF Dose Rate
(% of fuel consumed)
5.18%
Because both proposed options involve changes to not only the FTP emission standard, but
also tightened SET and LLC standards along with new idle standards, the above approach was
considered insufficient to estimate increased DEF consumption. Therefore, we have instead
scaled DEF consumption with the NOx reductions achieved under each of the alternatives. To do
this, we considered the molar mass of NOx, the molar mass of urea, the molar ratio of NO to
NO2, the mass concentration of urea in DEF along with the density of DEF to estimate the
jjjjjj j]lc relationship between DEF dose rate and NOx reduction across the SCR catalyst is based on methodology
presented in the Technical Support Document to the 2012 Non-conformance Penalty rule (the NCP Technical
Support Document, or NCP TSD).
332
-------
theoretical gallons of DEF consumed per ton of NOx reduced at 442 gallons/ton. The theoretical
DEF dosing rates was then compared to the data collected from the Stage 3 test program for the
hot FTP, SET and LLC (see Chapter 3 of the draft RIA). The data from this testing showed that
the NOx specific DEF dosing was 536, 478, and 568 gallons/ton for the hot FTP, SET and LLC,
respectively. Since this data takes into account any over dosing that occurs for part of the cycle,
and NO2/NO ratio being greater than 1 for parts of the cycle, we have adjusted the theoretical
442 gallons/ton NOx to the average of the hot FTP, SET and LLC, which is 527 gallons/ton.
These values are shown in Table 7-31
Table 7-31: Derivation of DEF Consumption per Ton of NOx Reduced
Value
Molar mass of NOx (g/mol)
46.0055
Molar mass of urea (g/mol)
60.07
Molar ratio of NO to NO2
1
Mass concentration of urea in DEF
0.325
Density of DEF (g/mL)
1.09
Theoretical gallons DEF/ton NOx reduced
442
Proposed gallons DEF/ton NOx reduced
527
The final calculation of DEF consumption in each option is to multiply the gallons of diesel
fuel consumed by 5.18%, to account for the baseline DEF consumption and add to that amount
527 gallons of DEF for each ton of NOx reduced from the baseline. Both the gallons of diesel
fuel consumed and the tons of NOx reduced are taken directly from the year-over-year MOVES
results.
The gallons of DEF consumed are then multiplied by the estimated price of DEF per gallon.
This analysis uses the DEF prices presented in the NCP Technical Support Document with
growth beyond 2042 projected at the same 1.3 percent rate as noted in the NCP TSD. Note that
the DEF prices presented in Table 7-32 update the NCP TSD's 2011 prices to 2017 dollars using
the GDP deflator presented in Table 7-8.
333
-------
Table 7-32: Diesel Exhaust Fluid Price per Gallon (2017 dollars)
Calendar Year
DEF Price/Gallon
2027
3.25
2028
3.30
2029
3.33
2030
3.37
2031
3.42
2032
3.46
2033
3.52
2034
3.56
2035
3.60
2036
3.65
2037
3.69
2038
3.75
2039
3.79
2040
3.85
2041
3.89
2042
3.94
2043
4.00
2044
4.04
2045
4.10
The impacts on DEF costs per mile for the first 10 years of the indicated model year lifetime
and under each alternative are shown in Table 7-36 through Table 7-37.
334
-------
Table 7-33: Diesel Exhaust Fluid Average Cost per Mile during the first 10 Years of each MY Lifetime by
MOVES Sourcetypes equipped with LHD Engines (cents/mile in 2017 dollars, Undiscounted) *
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Passenger Trucks
Baseline
0.85
0.90
0.94
Baseline+Proposed Option 1
0.86
0.95
1.00
Baseline+ Proposed Option 2
0.85
0.90
0.95
Option 1 increase from Baseline
0.01
0.05
0.06
Option 2 increase from Baseline
0.00
0.00
0.00
Light Commercial Trucks
Baseline
0.87
0.92
0.96
Baseline+ Proposed Option 1
0.88
0.97
1.02
Baseline+ Proposed Option 2
0.87
0.92
0.97
Option 1 increase from Baseline
0.01
0.05
0.06
Option 2 increase from Baseline
0.00
0.00
0.00
Short-Haul Single Unit
Trucks
Baseline
1.14
1.20
1.26
Baseline+ Proposed Option 1
1.14
1.25
1.31
Baseline+ Proposed Option 2
1.13
1.19
1.26
Option 1 increase from Baseline
0.00
0.05
0.05
Option 2 increase from Baseline
0.00
0.00
0.00
Long-Haul Single Unit
Trucks
Baseline
1.07
1.13
1.19
Baseline+ Proposed Option 1
1.07
1.16
1.23
Baseline+ Proposed Option 2
1.06
1.12
1.18
Option 1 increase from Baseline*
0.00
0.04
0.04
Option 2 increase from Baseline
-0.01
-0.01
-0.01
* As discussed in Section 5.2.2.5, we did not include diesel-fueled, engine-certified LHD2b3 vehicles in the baseline
emission inventory, but did include these vehicles in our emission inventory estimates for the Proposal and
Alternative 1 control scenarios. We decided to include the engine-certified LHD2b3 vehicles in the emissions
inventories of the control scenario to better estimate the projected emissions of heavy-duty vehicles in the control
scenarios, even though it will conservatively underestimate the benefits of controlling these vehicles. As a result,
the NOx emission inventories may demonstrate an increase in emissions from sourcetypes containing diesel-fueled
LHD2b3 vehicles for certain calendar years. We plan to update our baseline modeling to reflect any requirements
for diesel-fueled, engine-certified LHD2b3 vehicles in the final rulemaking. In addition, the MOVES baseline
emission rates for all LHD diesel (not just 2b3 vehicles) were incorrectly developed using a warranty period of 100K
instead of 50K, and thus we underestimated the LHD diesel emissions impact of the longer warranty periods in the
proposal and alternative scenarios. We will update the baseline warranty period for LHD diesel in the FRM analysis.
335
-------
Table 7-34: Diesel Exhaust Fluid Average Cost per Mile during the first 10 Years of each MY Lifetime by
MOVES Sourcetypes equipped with LHD45 Engines (cents/mile in 2017 dollars, Undiscounted)
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Other Buses
Baseline
1.99
2.10
2.21
Baseline+Proposed Option 1
2.21
2.39
2.52
Baseline+Proposed Option 2
2.20
2.32
2.44
Option 1 increase from Baseline
0.22
0.30
0.31
Option 2 increase from Baseline
0.21
0.23
0.24
Transit Buses
Baseline
2.00
2.10
2.21
Baseline+Proposed Option 1
2.20
2.39
2.51
Baseline+Proposed Option 2
2.20
2.32
2.44
Option 1 increase from Baseline
0.21
0.28
0.30
Option 2 increase from Baseline
0.21
0.22
0.23
School Buses
Baseline
1.49
1.57
1.66
Baseline+Proposed Option 1
1.66
1.80
1.90
Baseline+Proposed Option 2
1.66
1.75
1.84
Option 1 increase from Baseline
0.17
0.23
0.24
Option 2 increase from Baseline
0.16
0.17
0.18
Short-Haul Single Unit Trucks
Baseline
1.27
1.33
1.40
Baseline+Proposed Option 1
1.41
1.53
1.62
Baseline+Proposed Option 2
1.41
1.48
1.56
Option 1 increase from Baseline
0.15
0.20
0.21
Option 2 increase from Baseline
0.14
0.15
0.16
Long-Haul Single Unit Trucks
Baseline
1.19
1.25
1.31
Baseline+Proposed Option 1
1.31
1.41
1.49
Baseline+Proposed Option 2
1.30
1.37
1.44
Option 1 increase from Baseline
0.12
0.17
0.18
Option 2 increase from Baseline
0.12
0.12
0.13
336
-------
Table 7-35: Diesel Exhaust Fluid Average Cost per Mile during the first 10 Years of each MY Lifetime by
MOVES Sourcetypes equipped with MHD67 Engines (cents/mile in 2017 dollars, Undiscounted)
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Other Buses
Baseline
2.18
2.29
2.42
Baseline+Proposed Option 1
2.37
2.54
2.68
Baseline+Proposed Option 2
2.35
2.47
2.60
Option 1 increase from Baseline
0.19
0.25
0.26
Option 2 increase from Baseline
0.17
0.18
0.19
Transit Buses
Baseline
2.17
2.29
2.41
Baseline+Proposed Option 1
2.35
2.52
2.65
Baseline+Proposed Option 2
2.33
2.45
2.58
Option 1 increase from Baseline
0.18
0.23
0.24
Option 2 increase from Baseline
0.16
0.17
0.18
School Buses
Baseline
1.69
1.78
1.88
Baseline+Proposed Option 1
1.85
2.00
2.10
Baseline+Proposed Option 2
1.84
1.94
2.04
Option 1 increase from Baseline
0.16
0.21
0.22
Option 2 increase from Baseline
0.15
0.16
0.16
Refuse Trucks
Baseline
2.30
2.42
2.55
Baseline+Proposed Option 1
2.49
2.67
2.82
Baseline+Proposed Option 2
2.47
2.60
2.74
Option 1 increase from Baseline
0.20
0.26
0.27
Option 2 increase from Baseline
0.17
0.18
0.19
Short-Haul Single Unit Trucks
Baseline
1.64
1.73
1.82
Baseline+Proposed Option 1
1.82
1.95
2.06
Baseline+Proposed Option 2
1.80
1.89
1.99
Option 1 increase from Baseline
0.17
0.22
0.23
Option 2 increase from Baseline
0.15
0.16
0.17
Long-Haul Single Unit Trucks
Baseline
1.52
1.60
1.68
Baseline+Proposed Option 1
1.66
1.78
1.88
Baseline+Proposed Option 2
1.64
1.73
1.82
Option 1 increase from Baseline
0.14
0.18
0.19
Option 2 increase from Baseline
0.13
0.13
0.14
Motor Homes
Baseline
1.63
1.72
1.81
Baseline+Proposed Option 1
1.78
1.91
2.01
Baseline+Proposed Option 2
1.77
1.86
1.96
Option 1 increase from Baseline
0.14
0.19
0.20
Option 2 increase from Baseline
0.14
0.14
0.15
Short-Haul Combination Trucks
Baseline
2.12
2.23
2.35
Baseline+Proposed Option 1
2.27
2.43
2.56
Baseline+Proposed Option 2
2.26
2.38
2.50
Option 1 increase from Baseline
0.15
0.20
0.21
Option 2 increase from Baseline
0.14
0.14
0.15
337
-------
Table 7-36: Diesel Exhaust Fluid Average Cost per Mile during the first 10 Years of each MY Lifetime by
MOVES Sourcetypes equipped with HHD8 Engines (cents/mile in 2017 dollars, Undiscounted)
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Other Buses
Baseline
2.46
2.59
2.73
Baseline+Proposed Option 1
2.87
3.09
3.25
Baseline+Proposed Option 2
2.81
2.96
3.12
Option 1 increase from Baseline
0.41
0.50
0.53
Option 2 increase from Baseline
0.35
0.37
0.39
School Buses
Baseline
1.97
2.07
2.18
Baseline+Proposed Option 1
2.29
2.46
2.59
Baseline+Proposed Option 2
2.25
2.37
2.50
Option 1 increase from Baseline
0.32
0.39
0.41
Option 2 increase from Baseline
0.28
0.30
0.31
Refuse Trucks
Baseline
2.60
2.74
2.89
Baseline+Proposed Option 1
3.03
3.27
3.44
Baseline+Proposed Option 2
2.97
3.12
3.29
Option 1 increase from Baseline
0.43
0.52
0.55
Option 2 increase from Baseline
0.36
0.38
0.40
Short-Haul Single Unit Trucks
Baseline
2.17
2.29
2.41
Baseline+Proposed Option 1
2.58
2.77
2.92
Baseline+Proposed Option 2
2.52
2.66
2.80
Option 1 increase from Baseline
0.40
0.49
0.51
Option 2 increase from Baseline
0.35
0.37
0.39
Long-Haul Single Unit Trucks
Baseline
2.01
2.12
2.23
Baseline+Proposed Option 1
2.34
2.52
2.65
Baseline+Proposed Option 2
2.29
2.41
2.54
Option 1 increase from Baseline
0.33
0.40
0.42
Option 2 increase from Baseline
0.28
0.30
0.31
Motor Homes
Baseline
2.12
2.23
2.35
Baseline+Proposed Option 1
2.48
2.66
2.81
Baseline+Proposed Option 2
2.45
2.58
2.71
Option 1 increase from Baseline
0.36
0.44
0.46
Option 2 increase from Baseline
0.33
0.35
0.37
Short-Haul Combination Trucks
Baseline
2.42
2.54
2.68
Baseline+Proposed Option 1
2.75
2.95
3.11
Baseline+Proposed Option 2
2.71
2.85
3.00
Option 1 increase from Baseline
0.33
0.41
0.43
Option 2 increase from Baseline
0.29
0.31
0.32
Long-Haul Combination Trucks
Baseline
2.21
2.33
2.45
Baseline+Proposed Option 1
2.49
2.67
2.81
Baseline+Proposed Option 2
2.46
2.59
2.73
Option 1 increase from Baseline
0.28
0.34
0.36
Option 2 increase from Baseline
0.25
0.27
0.28
338
-------
Table 7-37: Diesel Exhaust Fluid Average Cost per Mile during the first 10 Years of each MY Lifetime by
MOVES Sourcetypes equipped with Urban Bus Engines (cents/mile in 2017 dollars, Undiscounted)
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Baseline
2.46
2.59
2.73
Baseline+Proposed Option 1
2.85
3.03
3.19
Transit Buses
Baseline+Proposed Option 2
2.79
2.93
3.09
Option 1 increase from Baseline
0.39
0.44
0.46
Option 2 increase from Baseline
0.32
0.34
0.36
7.2.2 Costs Associated with ORVR and the Estimated Reduction in Fuel Costs for Gasoline
Engines
This analysis estimates a small decrease in fuel costs, i.e., fuel savings, by vehicles equipped
with gasoline engines because the proposed ORVR system would be expected to capture
previously evaporated fuel and then burn that fuel in the engine (see Table 7-6 for our estimated
ORVR direct manufacturing cost). To estimate these impacts, we first converted grams of
hydrocarbon captured by the ORVR system to milliliters of gasoline that would ultimately be
burned in the engine. Based on that relationship, we estimate that 1.48 milliliters of gasoline
would be consumed for each gram of hydrocarbon emissions reduced under the proposed
Options 1 or 2. We estimated this value, 1.48, by assuming that the ORVR system would
exchange captured butane for gasoline on an energy basis and Tier 3 certification fuel has a
density of 0.7482 g/ml, or 2832 g/gal at 60 degrees F.362 We then used a butane energy density
of of 45.8 MJ/kg, or 19752 Btu/lb,363 and the Tier 3 cert fuel energy density of 17890 Btu/lb,364
giving a ratio of 1.117 grams of gasoline displaced for each gram of butane burned. Using the
density of Tier 3 certification fuel, we get 1.117 / 2832 = 0.0003943 gallons, or 1.48 ml, of
gasoline saved for each gram of butane captured.KKKKKK The effect being that the owner/operator
is no longer paying for evaporated fuel since it would be burned in the engine. Using AEO 2018
reference case gasoline prices, the impacts on gasoline costs per mile for the first 10 years of
each model year lifetime and under each alternative are shown in Table 7-38 through Table 7-40.
kkkkkk \ye estimate that the ORVR requirements in both the proposed Options 1 and 2 would result in a reduction
of approximately 0.3 million (calendar year 2027) to 4.8 million (calendar year 2045) gallons of gasoline,
representing roughly 0.1 percent of gasoline consumption from impacted vehicles.
339
-------
Table 7-38: Gasoline Vehicle Average Retail Fuel Cost per Mile during the first 10 Years of each Model Year
Lifetime by MOVES Sourcetype equipped with LHD45 Engines* (cents/mile in 2017 dollars, Undiscounted)
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Other Buses
Baseline
59.12
60.77
62.17
Baseline+Proposed Option 1
58.99
60.63
62.04
Baseline+Proposed Option 2
58.99
60.64
62.04
Option 1 increase from Baseline
-0.13
-0.13
-0.14
Option 2 increase from Baseline
-0.13
-0.13
-0.13
Transit Buses
Baseline
58.29
59.92
61.30
Baseline+Proposed Option 1
58.18
59.80
61.18
Baseline+Proposed Option 2
58.18
59.80
61.18
Option 1 increase from Baseline
-0.11
-0.12
-0.12
Option 2 increase from Baseline
-0.11
-0.12
-0.12
School Buses
Baseline
43.78
45.00
46.03
Baseline+Proposed Option 1
43.70
44.91
45.94
Baseline+Proposed Option 2
43.70
44.91
45.94
Option 1 increase from Baseline
-0.08
-0.08
-0.08
Option 2 increase from Baseline
-0.08
-0.08
-0.08
Short-Haul Single Unit Trucks
Baseline
38.57
39.64
40.57
Baseline+Proposed Option 1
38.53
39.60
40.53
Baseline+Proposed Option 2
38.53
39.60
40.53
Option 1 increase from Baseline
-0.04
-0.04
-0.04
Option 2 increase from Baseline
-0.04
-0.04
-0.04
Long-Haul Single Unit Trucks
Baseline
36.19
37.20
38.07
Baseline+Proposed Option 1
36.16
37.16
38.04
Baseline+Proposed Option 2
36.15
37.16
38.04
Option 1 increase from Baseline
-0.04
-0.04
-0.04
Option 2 increase from Baseline
-0.04
-0.04
-0.04
Motor Homes
Baseline
37.72
38.77
39.66
Baseline+Proposed Option 1
37.66
38.70
39.59
Baseline+Proposed Option 2
37.66
38.70
39.59
Option 1 increase from Baseline
-0.07
-0.07
-0.07
Option 2 increase from Baseline
-0.07
-0.07
-0.07
* Negative values denote lower costs, i.e., savings in fuel expenditures.
340
-------
Table 7-39: Gasoline Vehicle Average Retail Fuel Cost per Mile during the first 10 Years of each Model Year
Lifetime by MOVES Sourcetype equipped with MHD67 Engines* {cents/mile in 2017 dollars, Undiscounted)
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Short-Haul Single Unit Trucks
Baseline
48.76
50.11
51.28
Baseline+Proposed Option 1
48.70
50.05
51.22
Baseline+Proposed Option 2
48.70
50.05
51.22
Option 1 increase from Baseline
-0.05
-0.06
-0.06
Option 2 increase from Baseline
-0.06
-0.06
-0.06
Long-Haul Single Unit Trucks
Baseline
45.04
46.29
47.38
Baseline+Proposed Option 1
44.99
46.24
47.33
Baseline+Proposed Option 2
44.99
46.24
47.33
Option 1 increase from Baseline
-0.05
-0.05
-0.05
Option 2 increase from Baseline
-0.05
-0.05
-0.05
Motor Homes
Baseline
47.68
49.00
50.13
Baseline+Proposed Option 1
47.59
48.90
50.03
Baseline+Proposed Option 2
47.58
48.90
50.03
Option 1 increase from Baseline
-0.09
-0.09
-0.10
Option 2 increase from Baseline
-0.09
-0.09
-0.10
* Negative values denote lower costs, i.e., savings in fuel expenditures.
Table 7-40: Gasoline Vehicle Average Retail Fuel Cost per Mile during the first 10 Years of each Model Year
Lifetime by MOVES Sourcetype equipped with HHD8 Engines* (cents/mile in 2017 dollars, Undiscounted)
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Short-Haul Single Unit Trucks
Baseline
62.22
63.95
65.44
Baseline+Proposed Option 1
62.14
63.87
65.36
Baseline+Proposed Option 2
62.14
63.87
65.36
Option 1 increase from Baseline
-0.07
-0.08
-0.08
Option 2 increase from Baseline
-0.08
-0.08
-0.08
Long-Haul Single Unit Trucks
Baseline
57.62
59.22
60.61
Baseline+Proposed Option 1
57.55
59.15
60.54
Baseline+Proposed Option 2
57.54
59.15
60.54
Option 1 increase from Baseline
-0.07
-0.07
-0.08
Option 2 increase from Baseline
-0.07
-0.07
-0.08
Motor Homes
Baseline
60.77
62.46
63.89
Baseline+Proposed Option 1
60.65
62.33
63.77
Baseline+Proposed Option 2
60.65
62.33
63.77
Option 1 increase from Baseline
-0.12
-0.13
-0.13
Option 2 increase from Baseline
-0.12
-0.13
-0.13
* Negative values denote lower costs, i.e., savings in fuel expenditures.
7.2.3 Repair Cost Impacts Associated with Longer Warranty and Useful Life Periods
As noted above, we would expect that the longer warranty and useful life requirements being
proposed would have an impact on emission-related repair costs incurred by truck owners.
Researchers have noted the relationships among quality, reliability, and warranty for a variety of
goods.365 Wu, for instance, examines how analyzing warranty data can provide "early warnings"
on product problems that can then be used for design modifications.366 Guajardo et al. describe
one of the motives for warranties to be "incentives for the seller to improve product quality;"
specifically for light-duty vehicles, they find that buyers consider warranties to substitute for
341
-------
product quality, and to complement service quality.367 The other rationales are protection for
consumers against failures, provision of product quality information to consumers, and a means
to distinguish consumers according to their risk preferences. Murthy and Jack, for new products,
and Saidi-Mehrabad et al. for second-hand products, consider the role of warranties in improving
a buyer's confidence in quality of the good.368'369
On the one hand, we would expect that owner incurred emission repair costs would decrease
due to the proposed program because the longer regulatory useful life requirements would
encourage more durable parts to ensure compliance over the longer timeframes (i.e., reduced part
failure rates), and because costs for emission repairs that might be needed during the warranty
period would have to be covered by the OEM (see the discussion in Section 7.1.2 regarding
expected increased funding of warranty liability accounts to cover warranty claims). However,
on the other hand, we would also expect that repair costs might increase marginally in the same
way that direct manufacturing costs are increasing (i.e., more costly emission control systems
might have more costly repairs), which could increase OEM costs during the warranty period
and owner costs outside the warranty period. Overall for OEMs, while the longer warranty
period could be seen as potentially increasing repair costs incurred by OEMs, such costs would
have already been funded as part of the increased indirect costs discussed in Section 7.1.2).
Regarding increased technology costs associated with the proposed warranty and useful life
periods, we have estimated the cost of improved emission durability via increased research and
development (R&D) costs scaled by the longer useful life period, and increased warranty costs
scaled by the longer warranty period. This scaling of warranty and R&D costs is described above
in Section 7.1.2. We also included additional aftertreatment costs in the direct manufacturing
costs to address the increased useful life requirements (e.g., larger catalyst volume; see Chapters
2 and 4 for detailed discussions). We estimate that these efforts would help to reduce emission
repair costs during the emission warranty, regulatory useful life period, and possibly beyond.
Regarding the longer warranty and useful life periods reducing repair costs incurred by
owners, we estimated costs by starting with an operating cost analysis presented in a
whitepaper.370 On page 8 of that whitepaper, a figure shows estimated repair and maintenance
costs as reproduced in Table 7-41. A note on the figure states that the seven-year average would
be 10.48 cents/mile. This figure is considerably lower than a value of 17.1 cents/mile from a
study done by the American Transportation Research Institute (ATRI).371 However, the ATRI
value includes a more diverse range of truck applications and ages, including tractor-trailers
averaging 4.4 years old and straight trucks averaging 8.5 years old. By contrast, the Fleet
Advantage study includes Class-8 truck fleets only which, while not explicitly noted in the study,
we take to mean both short- and long-haul combination trucks (tractor-trailers). This would
suggest a higher weighting of older vehicles with higher repair and maintenance costs/mile
relative to the weighting of costs/mile in the Fleet Advantage study. Since we are estimating
impacts on repair costs associated with the proposed Options 1 or 2, using the lower cost/mile
estimate represents a more conservative approach.
342
-------
Table 7-41: Class-8 Diesel Repair & Maintenance Costs per Mile (100,000 miles per year, 2018 dollars*)372
Year in Operation
Cents/Mile
1
2.07
2
5.53
3
6.63
4
10.52
5
14.23
6
14.56
7
19.82
* While not explicitly stated in the study, the 2018
calendar year publication date implies that the
dollar figures in the study represent 2018 dollars.
Using the costs/mile shown in Table 7-41, we estimated the impacts on repair and
maintenance costs—specifically, emission-related repair costs—as described here. First, we
looked at the estimated ages that short- and long-haul combination truck reach their regulatory
warranty and useful life periods in the MOVES model, as shown in Table 7-42. For a discussion
of how estimated ages are determined, see Section 5.2.2 of this draft RIA; the approach used
here is identical except that here we have extended the approach beyond MOVES regulatory
class to MOVES sourcetypes to align more closely with likely in-use repair costs.LLLLLL Notably,
the current regulated warranty period is 5 years/100,000 miles and the current useful life is 10
years/435,000 miles for the trucks summarized in Table 7-44.
llllll jw0 0f die primary vehicle characteristics with MOVES are "sourcetype" and "regulatory class." The
proposed requirements apply to regulatory classes which refers to engines certified for use in vehicles of a specific
weight class (e.g., light heavy-duty, medium heavy-duty, heavy heavy-duty, etc.). The term "sourcetype" refers to
the type of vehicle (e.g., refuse truck, single unit short-haul truck, long-haul combination truck, etc.). Any given
sourcetype could be equipped with any of several different regulatory class engines. In terms of miles accumulated
on a vehicle, the sourcetype designation aligns more closely with the real world since "refuse trucks" tend to be
driven similarly regardless of their regulatory class engine. In this analysis, the warranty and useful life provisions
are tied to the regulatory class of the engine while the repair costs per mile are tied to the miles accumulated by the
sourcetype. Therefore, a refuse truck with a medium heavy-duty engine could have a different cost/mile curve than
one equipped with a heavy heavy-duty engine given the different warranty and useful life periods of the engine. This
is perhaps most easily understood when considering two different sourcetypes~a refuse truck and a long-haul
combination truck-both equipped with a heavy heavy-duty diesel engine. The former would reach its estimated
warranty period in 3 years versus 1 year for the latter. Similarly, the former would reach its estimated useful life
period in the full 10 years while the latter would reach its within 3. If we tie cost/mile metrics to point where
warranty and useful life are reached, then these two vehicles would have different cost/mile curves despite being
equipped with the same engine.
343
-------
Table 7-42: Estimated Vehicle Age at the End of the Warranty Period and the Useful Life for each Heavy
Heavy-duty Diesel Combination Truck for the Baseline Case
Warranty
Useful Life
HHP Short-
HHP Long-
HHP Short-
HHP Long-
haul Combo
haul Combo
haul Combo
haul Combo
Truck
Truck
Truck
Truck
(A)
Age limit
5
5
10
10
(B)
Mileage limit
100,000
100,000
435,000
435,000
(C)
Typical miles/year driven thru 7 years
75,000
144,000
75,000
144,000
(D)
Calculated age when the mileage limit is
reached
1
1
6
3
(E)
Estimated age (minimum of rows A and D)
1
1
6
3
We then looked at how the Fleet Advantage data lined up with the estimated ages shown in
Table 7-42. This is shown in Table 7-43 for short-haul combination trucks and in Table 7-44 for
long-haul combination trucks.
Table 7-43: Class-8 Diesel Short-haul Combination Truck Repair & Maintenance Costs per Mile (see Table
7-41) with MOVES HHD8 Mileage Accumulation
Year of Operating
Fleet Advantage
Cents/Mile
Notes
1
2.07
Within current warranty estimated age
2
5.53
Beyond current warranty estimated age
but within current useful life estimated age
3
6.63
4
10.52
5
14.23
6
14.56
7
19.82
Beyond current warranty and current useful life estimated ages
Table 7-44: Class-8 Diesel Long-haul Combination Truck Repair & Maintenance Costs per Mile (see Table
7-41) with MOVES HHD8 Mileage Accumulation
Year of Operating
Fleet Advantage
Cents/Mile
Notes
1
2.07
Within current warranty estimated age
2
5.53
Beyond current warranty estimated age
but within current useful life estimated age
3
6.63
4
10.52
Beyond current warranty and current useful life estimated ages
5
14.23
6
14.56
7
19.82
As shown in Table 7-43, trends in the Fleet Advantage data line up well with estimated
warranty and useful life ages of short-haul combination trucks, particularly the sharp increase in
cost/mile beyond the estimated useful life age. This seeming trend is not as apparent in Table
7-44 which compares the Fleet Advantage data against the long-haul combination truck
estimated ages. The docket to this rulemaking contains a memorandum that presents all of the
estimated warranty and useful life ages for each sourcetype and for each alternative.373
344
-------
While not clear, it is possible that the Fleet Advantage data were derived primarily from long-
haul combination trucks rather than short-haul combination trucks. Nonetheless, from these
admittedly limited data, it appears that repair and maintenance costs begin to increase beyond the
estimated warranty age and through the estimated useful life age. At some point beyond the
estimated useful life age, there appears to be a considerable jump in costs/mile. We made use of
these trends in developing our repair cost impacts associated with the proposed longer warranty
and useful life periods. Further, for this analysis, we have taken the "in-warranty" cost/mile to be
the 0.0207 value and the "at-useful-life" cost/mile to be the 14.56 value from the Fleet
Advantage data. Between the estimated warranty and useful life ages, the cost/mile is estimated
to lie on the slope of a line drawn between these two points. Beyond the estimated useful life
age, the cost/mile value increases to the 19.82 value and then, given lack of data, remains at that
value for the duration of the vehicle lifetime. The resultant repair and maintenance cost curve we
have developed is shown in Figure 7-2. We show the cost/mile remaining flat at that "beyond
useful life" value because the Fleet Advantage data did not show values beyond seven years.
25
20
Beyond Useful Life Age
* 15
Estimated Useful Life Age
S 10
-Estimated Warranty Age
4 5 6 7
Year of Operation
10
11
Figure 7-2: Repair & Maintenance Cost/Mile Curve (2018 dollars)
However, these are repair and maintenance costs rather than emission-related repair costs to
which the proposed emissions warranty and useful life provisions are tied. To estimate the
emission repair portion of these costs, we used the figure on page 3 of the Fleet Advantage
whitepaper which showed the percent of total repair and maintenance costs attributable to
different systems on the vehicle.374 The details of that chart are recreated below in Table 7-45
along with EPA's estimates for what portion of the repair and maintenance costs could be
considered to be emission repairs.
As shown, our analysis estimates that 10.8 percent of repair and maintenance costs are
emission repairs that the proposed warranty and useful life provisions are meant to impact. In
general, the maintenance/repair shares are estimated to be 50/50 repair/maintenance, with the
exception of "Preventive maintenance" and "Exhaust system." Preventive maintenance, by
definition, is not repair, and thus 100 percent is considered maintenance. For the exhaust system,
345
-------
we estimate that 80 percent of those costs are repair costs, with maintenance costs limited to DPF
cleaning. The share of emission-related vs. non-emission-related items was broken down, in
general, by those that are clearly emission-related, where we attribute 100 percent to the
emission-related category, versus those that are clearly not emission-related where 100 percent is
attributed to the non-emission-related category. Not shown are the non-emission repair share
which totals 35.0 percent, the emission maintenance share which is 7.2 percent and the non-
emission maintenance share which is 47.0 percent.
Table 7-45: Percentage of Total Repair & Maintenance Costs Attributable to Different Vehicle Systems375
Fleet Advantage
EPA Estimates
System
Percent of Total
Repair &
Maintenance
Cost
Maintenance
Share
Repair
Share
Non-Emission-
Related Share
Emission-
Related
Share
Emission
Repair
Share
(EPA
Estimate)
Tires, tubes,
liners & valves
43%
50%
50%
100%
0%
0.0%
Preventive
maintenance
12%
100%
0%
100%
0%
0%
Brakes
9%
50%
50%
100%
0%
0%
Expendable
items
8%
50%
50%
100%
0%
0%
Lighting
5%
50%
50%
100%
0%
0%
Cranking
5%
50%
50%
100%
0%
0%
Power plant
3%
50%
50%
0%
100%
1.5%
Exhaust system
6%
20%
80%
0%
100%
4.8%
Fuel system
6%
50%
50%
0%
100%
3%
Engine/motor
3%
50%
50%
0%
100%
1.5%
Total
100%
54.2%
45.8%
82.0%
18.0%
10.8%
Applying the 10.8 percent factor to the cost/mile curve shown in Figure 7-2 results in an
emission repair cost/mile curve as shown in Figure 7-3.
346
-------
25
20
Beyond Useful Life Age
CD
= 15
Estimated Useful Life Age
£ 10
u
-Estimated Warranty Age
Year in Service
> Repair & Maintenance Cost Curve
• Emission Repair Cost Curve
Figure 7-3: Emission Repair Cost/Mile Curve (2018 dollars)
With these cost curves established, we then estimated the repair costs under each alternative
assuming that for ages less than the estimated warranty age the in-warranty emission repair cost
to the owner would equal the in-warranty emission repair cost shown in Figure 7-3 (0.22
cents/mile). For the ages between the estimated warranty age and the estimated useful life age,
the owner's emission repair costs would be on the slope of a line between the in-warranty
emission repair cost and the end of useful life emission repair cost shown in Figure 7-3 (1.57
cents/mile). Beyond the estimated useful life age, the owner's emission repair costs would
increase to the maximum emission repair cost shown in Figure 7-3 (2.14 cents/mile).
However, the emission repair cost curve shown in Figure 7-3 represents today's Class-8
trucks, but future trucks will have marginally more costly emission control systems due to the
proposed requirements. To estimate that effect, we have scaled the curve using the direct costs of
every regulatory class within each alternative relative to the direct cost of a HHD8 diesel engine
in the baseline case.
For example, for a short-haul combination HHD8 diesel truck in the baseline case, with a
baseline case 1 year estimated warranty age and a 6-year estimated useful life age, the emission
repair costs would be as shown in Figure 7-4 (dashed line). Under the proposed Option 1, for a
MY2027 short-haul combination HHD8 diesel truck with a 450,000 mile warranty and 600,000
mile useful life resulting in an estimated warranty age of 6 years and an estimated useful life age
of 8 years, the emission repair cost curve would look like the solid line in Figure 7-4. The
cost/mile values during the warranty period have been extended for more years, but the cost/mile
has also been scaled by the relative cost of the proposed Option 1 technology package's direct
costs ($8,668 as shown in Table 7-5, adding $6,457 and $2,210) to the baseline package ($6,457
as shown in Table 7-5), or 1.34. So, while the baseline in-warranty emission repair cost/mile
would be 0.22 cents/mile (the Fleet Advantage 2.07 cents/mile * 10.8 percent * 0.977 GDP
deflator), the proposed Option 1 in-warranty emission repair cost would be 0.30 cents/mile.
Similarly, the maximum cost/mile value shown for the "beyond estimated useful life age" in the
baseline case is shown as 2.09 cent/mile (the Fleet Advantage 19.82 cents/mile value * 10.8
347
-------
percent * 0.977 GDP deflator). In the proposed Option 1, that value has been scaled again by the
1.34 direct cost scalar to arrive at the 2.80 cent/mile value shown.
3.50
3.00
2.50
O)
= 2.00
S 1.50
u
1.00
0.50
0.00
Baseline
¦Proposed Option 1
3 4 5 6
Years in Operation
Figure 7-4: Emission Repair Cost Curve in the Baseline Case (dashed line) and the proposed Option 1 (solid
line) for a Short-haul Combination Truck equipped with a HHD Class-8 Diesel Engine
The emission repair cost/mile curves shown in Figure 7-4 would result in an incremental
cost/mile that is negative for the operating years 2 through 7. During the first year, the
incremental cost/mile would be slightly higher due to the marginal technology costs associated
with the proposed Option 1 requirements. From years 1 through 7, the cost/mile would be lower
on increment due to the longer warranty and useful life periods and the efforts we are estimating
manufacturers would undertake to improve durability and to avoid warranty costs after sale
(efforts paid for in upfront indirect costs as described in Section 7.1.2 of this draft RIA). In the
years of operation beyond the useful life, emission repair costs/mile would then be expected to
be marginally higher, again due to the marginal technology costs associated with the proposed
Option 1 requirements. Importantly, in those later years of operation, miles driven per year tend
to decrease year-over-year which serves to offset somewhat the effect of the higher cost/mile
value. In the end, for most vehicle types (i.e., MOVES sourcetypes) our analysis shows that, in
general, the net emission repair costs over the first 10 years of operation would decrease (see
Table 7-46 through Table 7-55.
To make use of the Fleet Advantage data, and to apply it for all of the MOVES sourcetypes
within the scope of the proposed requirements and for each of the proposed options, we have
used the following methodology. First, we have determined the estimated warranty and useful
life ages analogous to those shown in Table 7-42 for each of the sourcetype and fueltype
combinations and for each of the regulatory alternatives. We then scaled the Fleet Advantage
cost/mile curve by the direct costs estimated here for each regulatory class relative to the
348
-------
baseline direct costs for the HHD8 diesel regulatory class. We then adjusted the Fleet Advantage
cost/mile curve accordingly for each regulatory class/sourcetype combination in each alternative.
Note that the estimated warranty and useful life ages used to determine the in-warranty, beyond-
warranty-but-within-useful-life, and beyond-useful-life segments of the cost/mile curves are tied
to sourcetypes, which can have very different miles driven per year and, therefore, very different
estimated ages.
We believe that it is reasonable to estimate that the emission repair costs would remain at the
cost estimated here during the longer warranty periods in each proposed option because of the
increased warranty and research and development costs we are estimating in our technology
costs. Note that we are also estimating that the emission repair costs beyond the useful life would
increase at a slightly higher rate based on the source data which suggested such a trend. Again,
cost/mile rates are estimated to flatten beyond the useful life since the source data included
operating costs through only 7 years. It is possible that cost/mile rates continue to increase with
age and that those would increase at similar rates in the baseline case and under each of the
proposed options. If true, the net effect would be the same as estimated here and the net effect is
of primary concern in our analysis.
The cost analysis tool developed for this analysis includes the calculations used to determine
the emission repair costs per mile for each MOVES sourcetype.376 The resultant emission repair
costs/mile under each alternative averaged over the first 10 years of operation are shown in Table
7-46 through Table 7-55.
349
-------
Table 7-46: Diesel Emission Repair Costs, Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with LHD2b3 Engines (cents/mile in 2017 dollars, Undiscounted)*
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Passenger Trucks
Baseline
0.54
0.54
0.54
Baseline+Proposed Option 1
0.30
0.18
0.18
Baseline+Proposed Option 2
0.48
0.47
0.47
Option 1 increase from Baseline
-0.24
-0.36
-0.36
Option 2 increase from Baseline
-0.05
-0.07
-0.07
Light Commercial Trucks
Baseline
0.54
0.54
0.54
Baseline+Proposed Option 1
0.30
0.18
0.18
Baseline+Proposed Option 2
0.48
0.47
0.47
Option 1 increase from Baseline
-0.24
-0.36
-0.36
Option 2 increase from Baseline
-0.05
-0.07
-0.07
Short-Haul Single Unit
Trucks
Baseline
0.69
0.69
0.70
Baseline+Proposed Option 1
0.58
0.22
0.22
Baseline+Proposed Option 2
0.46
0.45
0.44
Option 1 increase from Baseline
-0.12
-0.48
-0.48
Option 2 increase from Baseline
-0.23
-0.25
-0.25
Long-Haul Single Unit
Trucks
Baseline
0.93
0.94
0.94
Baseline+Proposed Option 1
0.97
0.64
0.63
Baseline+Proposed Option 2
0.83
0.81
0.80
Option 1 increase from Baseline
0.04
-0.30
-0.31
Option 2 increase from Baseline
-0.11
-0.13
-0.14
* Negative values denote lower costs, i.e., savings in emission repair expenditures. As discussed in Section 5.2.2.5,
we did not include diesel-fueled, engine-certified LHD2b3 vehicles in the baseline emission inventory, but we did
include these vehicles in our emission inventory estimates for the proposed Options 1 and 2 control scenarios. We
decided to include the engine-certified LHD2b3 vehicles in the emissions inventories of the control scenario to better
estimate the projected emissions of heavy-duty vehicles in the control scenarios, even though it will conservatively
underestimate the benefits of controlling these vehicles. As a result, the NOx emission inventories may demonstrate
an increase in emissions from Sourcetypes containing diesel-fueled LHD2b3 vehicles for certain calendar years. We
plan to update our baseline modeling to reflect any requirements for diesel-fueled, engine-certified LHD2b3 vehicles
in the final rulemaking. In addition, the MOVES baseline emission rates for all LHD diesel (not just 2b3 vehicles)
were incorrectly developed using a warranty period of 100K instead of 50K, and thus we underestimated the LHD
diesel emissions impact of the longer warranty periods in the proposal and alternative scenarios. We will update the
baseline warranty period for LHD diesel in the FRM analysis.
350
-------
Table 7-47: Diesel Emission Repair Costs, Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with LHD45 Engines (cents/mile in 2017 dollars, Undiscounted) *
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Other Buses
Baseline
0.95
0.95
0.95
Baseline+Proposed Option 1
1.03
0.69
0.68
Baseline+Proposed Option 2
0.88
0.86
0.85
Option 1 increase from Baseline
0.08
-0.25
-0.26
Option 2 increase from Baseline
-0.07
-0.09
-0.10
Transit Buses
Baseline
0.95
0.95
0.95
Baseline+Proposed Option 1
1.03
0.69
0.68
Baseline+Proposed Option 2
0.88
0.86
0.85
Option 1 increase from Baseline
0.08
-0.25
-0.26
Option 2 increase from Baseline
-0.07
-0.09
-0.10
School Buses
Baseline
0.36
0.36
0.36
Baseline+Proposed Option 1
0.32
0.18
0.18
Baseline+Proposed Option 2
0.52
0.51
0.50
Option 1 increase from Baseline
-0.04
-0.18
-0.18
Option 2 increase from Baseline
0.16
0.14
0.14
Short-Haul Single Unit Trucks
Baseline
0.70
0.70
0.70
Baseline+Proposed Option 1
0.58
0.22
0.22
Baseline+Proposed Option 2
0.46
0.45
0.44
Option 1 increase from Baseline
-0.11
-0.48
-0.48
Option 2 increase from Baseline
-0.23
-0.25
-0.25
Long-Haul Single Unit Trucks
Baseline
0.94
0.94
0.94
Baseline+Proposed Option 1
0.98
0.64
0.63
Baseline+Proposed Option 2
0.84
0.81
0.80
Option 1 increase from Baseline
0.04
-0.30
-0.31
Option 2 increase from Baseline
-0.10
-0.13
-0.14
* Negative values denote lower costs, i.e., savings in emission repair expenditures.
351
-------
Table 7-48: Diesel Emission Repair Costs, Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with MHD67 Engines (cents/mile in 2017 dollars, Undiscounted) *
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Other Buses
Baseline
0.80
0.80
0.80
Baseline+Proposed Option 1
0.74
0.48
0.47
Baseline+Proposed Option 2
0.65
0.64
0.63
Option 1 increase from Baseline
-0.06
-0.33
-0.33
Option 2 increase from Baseline
-0.15
-0.16
-0.17
Transit Buses
Baseline
0.80
0.80
0.80
Baseline+Proposed Option 1
0.74
0.48
0.47
Baseline+Proposed Option 2
0.65
0.64
0.63
Option 1 increase from Baseline
-0.06
-0.33
-0.33
Option 2 increase from Baseline
-0.15
-0.16
-0.17
School Buses
Baseline
0.38
0.38
0.38
Baseline+Proposed Option 1
0.36
0.19
0.18
Baseline+Proposed Option 2
0.54
0.52
0.52
Option 1 increase from Baseline
-0.02
-0.20
-0.20
Option 2 increase from Baseline
0.15
0.14
0.13
Refuse Trucks
Baseline
0.72
0.72
0.72
Baseline+Proposed Option 1
0.48
0.23
0.23
Baseline+Proposed Option 2
0.50
0.48
0.48
Option 1 increase from Baseline
-0.24
-0.49
-0.49
Option 2 increase from Baseline
-0.23
-0.24
-0.25
Short-Haul Single Unit Trucks
Baseline
0.51
0.51
0.51
Baseline+Proposed Option 1
0.33
0.19
0.18
Baseline+Proposed Option 2
0.48
0.47
0.46
Option 1 increase from Baseline
-0.18
-0.33
-0.33
Option 2 increase from Baseline
-0.03
-0.05
-0.05
Long-Haul Single Unit Trucks
Baseline
0.77
0.77
0.77
Baseline+Proposed Option 1
0.68
0.39
0.43
Baseline+Proposed Option 2
0.61
0.60
0.59
Option 1 increase from Baseline
-0.09
-0.39
-0.35
Option 2 increase from Baseline
-0.16
-0.18
-0.19
Motor Homes
Baseline
0.38
0.38
0.38
Baseline+Proposed Option 1
0.36
0.19
0.18
Baseline+Proposed Option 2
0.53
0.51
0.51
Option 1 increase from Baseline
-0.02
-0.19
-0.19
Option 2 increase from Baseline
0.15
0.14
0.13
Short-Haul Combination Trucks
Baseline
1.13
1.13
1.13
Baseline+Proposed Option 1
1.22
0.99
0.98
Baseline+Proposed Option 2
1.28
1.24
1.22
Option 1 increase from Baseline
0.08
-0.14
-0.15
Option 2 increase from Baseline
0.15
0.11
0.09
352
-------
Table 7-49: Diesel Emission Repair Costs, Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with HHD8 Engines (cents/mile in 2017 dollars, Undiscounted) *
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Other Buses
Baseline
0.70
0.70
0.70
Baseline+Proposed Option 1
0.53
0.29
0.28
Baseline+Proposed Option 2
0.77
0.75
0.74
Option 1 increase from Baseline
-0.18
-0.42
-0.42
Option 2 increase from Baseline
0.07
0.05
0.04
School Buses
Baseline
0.62
0.61
0.61
Baseline+Proposed Option 1
0.56
0.29
0.28
Baseline+Proposed Option 2
0.83
0.81
0.80
Option 1 increase from Baseline
-0.06
-0.33
-0.33
Option 2 increase from Baseline
0.21
0.19
0.18
Refuse Trucks
Baseline
0.70
0.70
0.70
Baseline+Proposed Option 1
0.52
0.29
0.28
Baseline+Proposed Option 2
0.76
0.74
0.73
Option 1 increase from Baseline
-0.18
-0.41
-0.41
Option 2 increase from Baseline
0.06
0.04
0.04
Short-Haul Single Unit Trucks
Baseline
0.61
0.61
0.61
Baseline+Proposed Option 1
0.50
0.29
0.28
Baseline+Proposed Option 2
0.73
0.72
0.71
Option 1 increase from Baseline
-0.11
-0.33
-0.33
Option 2 increase from Baseline
0.12
0.11
0.10
Long-Haul Single Unit Trucks
Baseline
0.67
0.67
0.67
Baseline+Proposed Option 1
0.49
0.29
0.28
Baseline+Proposed Option 2
0.72
0.70
0.70
Option 1 increase from Baseline
-0.18
-0.38
-0.39
Option 2 increase from Baseline
0.05
0.03
0.03
Motor Homes
Baseline
0.60
0.60
0.60
Baseline+Proposed Option 1
0.55
0.29
0.28
Baseline+Proposed Option 2
0.81
0.79
0.78
Option 1 increase from Baseline
-0.05
-0.32
-0.32
Option 2 increase from Baseline
0.21
0.19
0.18
Short-Haul Combination Trucks
Baseline
1.27
1.26
1.26
Baseline+Proposed Option 1
0.91
0.48
0.47
Baseline+Proposed Option 2
0.86
0.97
0.95
Option 1 increase from Baseline
-0.36
-0.79
-0.79
Option 2 increase from Baseline
-0.41
-0.30
-0.31
Long-Haul Combination Trucks
Baseline
1.69
1.69
1.68
Baseline+Proposed Option 1
1.90
1.57
1.55
Baseline+Proposed Option 2
1.82
1.94
1.91
Option 1 increase from Baseline
0.20
-0.12
-0.13
Option 2 increase from Baseline
0.12
0.25
0.23
* Negative values denote lower costs, i.e., savings in emission repair expenditures.
353
-------
Table 7-50: Diesel Emission Repair Costs, Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with Urban Bus Engines (cents/mile in 2017 dollars,
Undiscounted) *
MOVES
Sourcetype
Scenario
MY2027
MY2031
MY2035
Transit Buses
Baseline
0.45
0.45
0.44
Baseline+Proposed Option 1
0.35
0.19
0.19
Baseline+Proposed Option 2
0.51
0.50
0.49
Option 1 increase from Baseline
-0.10
-0.26
-0.26
Option 2 increase from Baseline
0.07
0.05
0.05
* Negative values denote lower costs, i.e., savings in emission repair expenditures.
Table 7-51: Gasoline Emission Repair Costs, Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with LHD45 Engines (cents/mile in 2017 dollars, Undiscounted)*
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Other Buses
Baseline
0.22
0.22
0.22
Baseline+Proposed Option 1
0.27
0.23
0.22
Baseline+Proposed Option 2
0.25
0.24
0.24
Option 1 increase from Baseline
0.06
0.01
0.01
Option 2 increase from Baseline
0.03
0.02
0.02
Transit Buses
Baseline
0.22
0.22
0.22
Baseline+Proposed Option 1
0.27
0.23
0.22
Baseline+Proposed Option 2
0.25
0.24
0.24
Option 1 increase from Baseline
0.06
0.01
0.01
Option 2 increase from Baseline
0.03
0.02
0.02
School Buses
Baseline
0.08
0.08
0.08
Baseline+Proposed Option 1
0.07
0.04
0.04
Baseline+Proposed Option 2
0.12
0.12
0.11
Option 1 increase from Baseline
-0.01
-0.04
-0.04
Option 2 increase from Baseline
0.04
0.04
0.03
Short-Haul Single Unit Trucks
Baseline
0.17
0.17
0.17
Baseline+Proposed Option 1
0.17
0.10
0.12
Baseline+Proposed Option 2
0.13
0.12
0.14
Option 1 increase from Baseline
0.00
-0.07
-0.06
Option 2 increase from Baseline
-0.05
-0.05
-0.03
Long-Haul Single Unit Trucks
Baseline
0.22
0.22
0.22
Baseline+Proposed Option 1
0.27
0.22
0.22
Baseline+Proposed Option 2
0.24
0.23
0.23
Option 1 increase from Baseline
0.05
0.00
0.00
Option 2 increase from Baseline
0.02
0.02
0.01
Motor Homes
Baseline
0.08
0.08
0.08
Baseline+Proposed Option 1
0.07
0.04
0.04
Baseline+Proposed Option 2
0.12
0.11
0.11
Option 1 increase from Baseline
-0.01
-0.04
-0.04
Option 2 increase from Baseline
0.04
0.04
0.03
* Negative values denote lower costs, i.e., savings in emission repair expenditures.
354
-------
Table 7-52: Gasoline Emission Repair Costs, Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with MHD67 Engines (cents/mile in 2017 dollars, Undiscounted)*
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Short-Haul Single Unit Trucks
Baseline
0.17
0.17
0.17
Baseline+Proposed Option 1
0.17
0.10
0.12
Baseline+Proposed Option 2
0.13
0.12
0.14
Option 1 increase from Baseline
0.00
-0.07
-0.06
Option 2 increase from Baseline
-0.05
-0.05
-0.03
Long-Haul Single Unit Trucks
Baseline
0.22
0.22
0.22
Baseline+Proposed Option 1
0.27
0.22
0.22
Baseline+Proposed Option 2
0.24
0.23
0.23
Option 1 increase from Baseline
0.05
0.00
0.00
Option 2 increase from Baseline
0.02
0.02
0.01
Motor Homes
Baseline
0.08
0.08
0.08
Baseline+Proposed Option 1
0.07
0.04
0.04
Baseline+Proposed Option 2
0.12
0.11
0.11
Option 1 increase from Baseline
-0.01
-0.04
-0.04
Option 2 increase from Baseline
0.04
0.04
0.03
* Negative values denote lower costs, i.e., savings in emission repair expenditures.
Table 7-53: Gasoline Emission Repair Costs, Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with HHD8 Engines (cents/mile in 2017 dollars, Undiscounted)*
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Short-Haul Single Unit Trucks
Baseline
0.17
0.17
0.17
Baseline+Proposed Option 1
0.17
0.10
0.12
Baseline+Proposed Option 2
0.13
0.12
0.14
Option 1 increase from Baseline
0.00
-0.07
-0.06
Option 2 increase from Baseline
-0.05
-0.05
-0.03
Long-Haul Single Unit Trucks
Baseline
0.22
0.22
0.22
Baseline+Proposed Option 1
0.27
0.22
0.22
Baseline+Proposed Option 2
0.24
0.23
0.23
Option 1 increase from Baseline
0.05
0.00
0.00
Option 2 increase from Baseline
0.02
0.02
0.01
Motor Homes
Baseline
0.08
0.08
0.08
Baseline+Proposed Option 1
0.07
0.04
0.04
Baseline+Proposed Option 2
0.12
0.11
0.11
Option 1 increase from Baseline
-0.01
-0.04
-0.04
Option 2 increase from Baseline
0.04
0.04
0.03
* Negative values denote lower costs, i.e., savings in emission repair expenditures.
355
-------
Table 7-54: CNG Emission Repair Costs, Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with HHD8 Engines (cents/mile in 2017 dollars, Undiscounted)*
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Other Buses
Baseline
0.45
0.45
0.45
Baseline+Proposed Option 1
0.25
0.14
0.14
Baseline+Proposed Option 2
0.37
0.37
0.37
Option 1 increase from Baseline
-0.20
-0.31
-0.31
Option 2 increase from Baseline
-0.08
-0.08
-0.08
School Buses
Baseline
0.39
0.39
0.39
Baseline+Proposed Option 1
0.27
0.14
0.14
Baseline+Proposed Option 2
0.39
0.39
0.39
Option 1 increase from Baseline
-0.12
-0.25
-0.25
Option 2 increase from Baseline
0.00
0.00
0.00
Refuse Trucks
Baseline
0.44
0.44
0.44
Baseline+Proposed Option 1
0.25
0.14
0.14
Baseline+Proposed Option 2
0.36
0.36
0.36
Option 1 increase from Baseline
-0.20
-0.30
-0.30
Option 2 increase from Baseline
-0.08
-0.08
-0.08
Short-Haul Single Unit Trucks
Baseline
0.39
0.39
0.39
Baseline+Proposed Option 1
0.24
0.14
0.14
Baseline+Proposed Option 2
0.35
0.35
0.35
Option 1 increase from Baseline
-0.15
-0.25
-0.25
Option 2 increase from Baseline
-0.04
-0.04
-0.04
Long-Haul Single Unit Trucks
Baseline
0.43
0.43
0.43
Baseline+Proposed Option 1
0.24
0.14
0.14
Baseline+Proposed Option 2
0.34
0.34
0.34
Option 1 increase from Baseline
-0.19
-0.29
-0.29
Option 2 increase from Baseline
-0.08
-0.08
-0.08
Short-Haul Combination Trucks
Baseline
0.81
0.80
0.80
Baseline+Proposed Option 1
0.43
0.23
0.23
Baseline+Proposed Option 2
0.41
0.47
0.47
Option 1 increase from Baseline
-0.37
-0.57
-0.57
Option 2 increase from Baseline
-0.40
-0.33
-0.33
* Negative values denote lower costs, i.e., savings in emission repair expenditures.
Table 7-55: CNG Emission Repair Costs, Average Cost per Mile during the first 10 Years of each MY
Lifetime by MOVES Sourcetypes equipped with Urban Bus Engines (cents/mile in 2017 dollars,
Undiscounted) *
MOVES Sourcetype
Scenario
MY2027
MY2031
MY2035
Baseline
0.34
0.34
0.34
Baseline+Proposed Option 1
0.19
0.10
0.10
Transit Buses
Baseline+Proposed Option 2
0.28
0.28
0.28
Option 1 increase from Baseline
-0.15
-0.23
-0.23
Option 2 increase from Baseline
-0.06
-0.06
-0.06
* Negative values denote lower costs, i.e., savings in emission repair expenditures.
7.3 Program Costs
Using the cost elements outlined above, we have estimated the costs associated with the
proposed criteria pollutant standards as presented in the following tables. Costs are broken into
356
-------
two main categories: Technology Costs and Operating Costs. Technology costs are broken
further into direct costs and the indirect costs elements (warranty costs, R&D costs, other indirect
costs and profits) to arrive at total technology costs. Operating costs are broken into urea/DEF
costs (diesel only), fuel savings (gasoline only) and repair costs. Section 7.3.1 present the total
technology costs for each proposed option relative to the baseline case, broken down by class,
showing diesel and gasoline separately as well as combined. Section 7.3.2 presents the operating
costs, similarly grouped. Section 7.3.3 presents the total program costs relative to the baseline
case for each proposed option. Costs are presented in 2017 dollars in undiscounted annual values
along with present values at both 3 and 7 percent discount rates with values discounted to the
2019 calendar year.
As noted in the introductory text to this chapter, the costs presented here reflect our best
estimate of the costs to society. Total program costs under the two proposed options are
presented in terms of calendar year 2045 costs, present value costs, and annualized costs (see
Table 7-98 and Table 7.99) mmmmmm
We are not including an analysis of the costs of the Alternative (described in Sections III and
IV) because we currently do not have sufficient information to conclude that the Alternative
standards would be feasible in the MY2027 timeframe. Section III presents our current feasibility
analysis for the Alternative.
7.3.1 Total Technology Costs
The series of tables shown here show direct manufacturing, warranty, R&D, profits, other
indirect costs and total technology costs by fuel type, then by regulatory class for each of the
proposed options. Values shown for a given calendar year are undiscounted values while
discounted values are presented at both 3 and 7 percent discount rates. All values are shown in
2017 dollars.
141414141414 The costs presented in Table 7-98 and Table 7-99 are presented again in Table ES .w hich summarizes the
net benefits of the two proposed options.
357
-------
Table 7-56: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, Diesel, Millions
of 2017 dollars *
Calendar Year
Direct
Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other
Indirect
Costs
Profits
Total
Technology
Costs
2027
$950
$290
$130
$280
$48
$1,700
2028
$920
$280
$130
$270
$46
$1,600
2029
$900
$280
$120
$260
$45
$1,600
2030
$880
$280
$44
$260
$44
$1,500
2031
$870
$390
$100
$250
$43
$1,700
2032
$850
$390
$100
$250
$42
$1,600
2033
$830
$390
$98
$240
$42
$1,600
2034
$820
$380
$41
$240
$41
$1,500
2035
$820
$380
$41
$240
$41
$1,500
2036
$810
$380
$41
$240
$41
$1,500
2037
$800
$380
$40
$230
$40
$1,500
2038
$800
$380
$40
$230
$40
$1,500
2039
$800
$380
$40
$230
$40
$1,500
2040
$800
$380
$40
$230
$40
$1,500
2041
$800
$390
$40
$230
$40
$1,500
2042
$800
$390
$40
$230
$40
$1,500
2043
$800
$390
$40
$230
$40
$1,500
2044
$800
$390
$40
$230
$40
$1,500
2045
$810
$390
$40
$230
$40
$1,500
PV, 3%
$12,000
$5,100
$970
$3,500
$600
$22,000
PV, 7%
$8,800
$3,600
$760
$2,600
$440
$16,000
* Values show 2 significant digits. Note that the Information Collection Request costs addressed in Section
XIII of the preamble would fall within the "Other" indirect costs shown here.
358
-------
Table 7-57: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, Diesel, Millions
of 2017 dollars *
Calendar Year
Direct
Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other
Indirect
Costs
Profits
Total
Technology
Costs
2027
$950
$200
$170
$280
$48
$1,600
2028
$920
$200
$170
$270
$46
$1,600
2029
$900
$200
$160
$260
$45
$1,600
2030
$880
$190
$44
$260
$44
$1,400
2031
$870
$190
$43
$250
$43
$1,400
2032
$850
$190
$42
$250
$42
$1,400
2033
$830
$190
$42
$240
$42
$1,300
2034
$820
$190
$41
$240
$41
$1,300
2035
$820
$190
$41
$240
$41
$1,300
2036
$810
$190
$41
$240
$41
$1,300
2037
$800
$190
$40
$230
$40
$1,300
2038
$800
$180
$40
$230
$40
$1,300
2039
$800
$190
$40
$230
$40
$1,300
2040
$800
$190
$40
$230
$40
$1,300
2041
$800
$190
$40
$230
$40
$1,300
2042
$800
$190
$40
$230
$40
$1,300
2043
$800
$190
$40
$230
$40
$1,300
2044
$800
$190
$40
$230
$40
$1,300
2045
$810
$190
$40
$230
$40
$1,300
PV, 3%
$12,000
$2,700
$940
$3,500
$600
$20,000
PV, 7%
$8,800
$2,000
$750
$2,600
$440
$15,000
* Values show 2 significant digits. Note that the Information Collection Request costs addressed in Section
XIII of the preamble would fall within the "Other" indirect costs shown here.
Table 7-58: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, Diesel by
Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other Indirect
Costs
Profits
Total Technology
Costs
LHD
$280
$37
$26
$81
$14
$440
LHD45
$1,700
$230
$150
$500
$86
$2,700
MHD67
$3,400
$740
$270
$990
$170
$5,600
HHD8
$6,600
$4,100
$520
$1,900
$330
$13,000
Urban
bus
$100
$56
$7.6
$29
$5.0
$200
Sum
$12,000
$5,100
$970
$3,500
$600
$22,000
* Values show 2 significant digits.
359
-------
Table 7-59: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, Diesel by
Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other Indirect
Costs
Profits
Total Technology
Costs
LHD
$280
$11
$28
$81
$14
$420
LHD45
$1,700
$70
$170
$500
$86
$2,500
MHD67
$3,400
$300
$270
$990
$170
$5,100
HHD8
$6,600
$2,300
$470
$1,900
$330
$12,000
Urban
bus
$100
$32
$6.9
$29
$5.0
$170
Sum
$12,000
$2,700
$940
$3,500
$600
$20,000
* Values show 2 significant digits.
Table 7-60: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, Diesel by
Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other Indirect
Costs
Profits
Total Technology
Costs
LHD
$210
$26
$21
$60
$10
$320
LHD45
$1,300
$160
$120
$360
$63
$2,000
MHD67
$2,500
$530
$210
$720
$120
$4,100
HHD8
$4,800
$2,900
$410
$1,400
$240
$9,700
Urban
bus
$73
$39
$5.9
$21
$3.7
$140
Sum
$8,800
$3,600
$760
$2,600
$440
$16,000
* Values show 2 significant digits.
Table 7-61: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, Diesel by
Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other Indirect
Costs
Profits
Total Technology
Costs
LHD
$210
$8.4
$23
$60
$10
$310
LHD45
$1,300
$51
$140
$360
$63
$1,900
MHD67
$2,500
$220
$220
$720
$120
$3,800
HHD8
$4,800
$1,700
$370
$1,400
$240
$8,400
Urban
bus
$73
$23
$5.4
$21
$3.7
$130
Sum
$8,800
$2,000
$750
$2,600
$440
$15,000
* Values show 2 significant digits.
360
-------
Table 7-62: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, Gasoline,
Millions of 2017 dollars *
Calendar
Year
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other
Indirect
Costs
Profits
Total Technology
Costs
2027
$41
$5.7
$4.6
$15
$2.5
$69
2028
$40
$5.6
$4.5
$14
$2.4
$67
2029
$39
$5.6
$4.4
$14
$2.3
$65
2030
$38
$5.5
$1.9
$14
$2.3
$62
2031
$37
$9.0
$3.6
$13
$2.2
$66
2032
$37
$8.9
$3.6
$13
$2.2
$65
2033
$36
$8.8
$3.5
$13
$2.2
$63
2034
$36
$8.8
$1.8
$13
$2.1
$61
2035
$35
$8.7
$1.8
$13
$2.1
$61
2036
$35
$8.7
$1.8
$13
$2.1
$60
2037
$35
$8.7
$1.7
$13
$2.1
$60
2038
$35
$8.7
$1.7
$12
$2.1
$60
2039
$35
$8.7
$1.7
$12
$2.1
$60
2040
$35
$8.7
$1.7
$12
$2.1
$60
2041
$35
$8.8
$1.7
$13
$2.1
$60
2042
$35
$8.8
$1.7
$13
$2.1
$60
2043
$35
$8.9
$1.7
$13
$2.1
$60
2044
$35
$8.9
$1.7
$13
$2.1
$60
2045
$35
$8.9
$1.8
$13
$2.1
$60
PV, 3%
$520
$110
$38
$190
$31
$890
PV, 7%
$380
$80
$29
$140
$23
$650
* Values show 2 significant digits. Note that the Information Collection Request costs addressed in Section XIII of
the preamble would fall within the "Other" indirect costs shown here.
361
-------
Table 7-63: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, Gasoline,
Millions of 2017 dollars *
Calendar
Year
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other
Indirect
Costs
Profits
Total Technology
Costs
2027
$41
$5.7
$7.1
$15
$2.5
$71
2028
$40
$5.6
$7.0
$14
$2.4
$69
2029
$39
$5.6
$6.9
$14
$2.3
$68
2030
$38
$5.5
$1.9
$14
$2.3
$62
2031
$37
$5.4
$1.9
$13
$2.2
$61
2032
$37
$5.4
$1.8
$13
$2.2
$59
2033
$36
$5.3
$1.8
$13
$2.2
$58
2034
$36
$5.3
$1.8
$13
$2.1
$58
2035
$35
$5.2
$1.8
$13
$2.1
$57
2036
$35
$5.2
$1.8
$13
$2.1
$57
2037
$35
$5.2
$1.7
$13
$2.1
$56
2038
$35
$5.2
$1.7
$12
$2.1
$56
2039
$35
$5.2
$1.7
$12
$2.1
$56
2040
$35
$5.2
$1.7
$12
$2.1
$56
2041
$35
$5.3
$1.7
$13
$2.1
$56
2042
$35
$5.3
$1.7
$13
$2.1
$56
2043
$35
$5.3
$1.7
$13
$2.1
$57
2044
$35
$5.3
$1.7
$13
$2.1
$57
2045
$35
$5.3
$1.8
$13
$2.1
$57
PV, 3%
$520
$77
$40
$190
$31
$860
PV, 7%
$380
$56
$32
$140
$23
$630
* Values show 2 significant digits. Note that the Information Collection Request costs addressed in Section XIII of
the preamble would fall within the "Other" indirect costs shown here.
Table 7-64: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, Gasoline by
Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other Indirect
Costs
Profits
Total Technology
Costs
LHD45
$340
$74
$24
$120
$20
$580
MHD67
$150
$34
$11
$56
$9.3
$260
HHD8
$29
$6.3
$2.1
$10
$1.7
$49
Sum
$520
$110
$38
$190
$31
$890
* Values show 2 significant digits.
Table 7-65: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, Gasoline by
Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other Indirect
Costs
Profits
Total Technology
Costs
LHD45
$340
$50
$26
$120
$20
$560
MHD67
$150
$23
$12
$56
$9.3
$250
HHD8
$29
$4.2
$2.2
$10
$1.7
$47
Sum
$520
$77
$40
$190
$31
$860
* Values show 2 significant digits.
362
-------
Table 7-66: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, Gasoline by
Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other Indirect
Costs
Profits
Total Technology
Costs
LHD45
$250
$52
$19
$89
$15
$420
MHD67
$110
$24
$8.6
$41
$6.8
$190
HHD8
$21
$4.4
$1.6
$7.5
$1.3
$36
Sum
$380
$80
$29
$140
$23
$650
* Values show 2 significant digits.
Table 7-67: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, Gasoline by
Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other Indirect
Costs
Profits
Total Technology
Costs
LHD45
$250
$36
$21
$89
$15
$410
MHD67
$110
$16
$10
$41
$6.8
$190
HHD8
$21
$3.0
$1.8
$7.5
$1.3
$34
Sum
$380
$56
$32
$140
$23
$630
* Values show 2 significant digits.
Table 7-68: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, CNG, Millions
of 2017 dollars *
Calendar
Year
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other
Indirect
Costs
Profits
Total Technology
Costs
2027
$0.30
$4.9
$0.89
$0,087
$0,015
$6.2
2028
$0.29
$4.9
$0.90
$0,084
$0,015
$6.2
2029
$0.28
$4.9
$0.90
$0,083
$0,014
$6.2
2030
$0.28
$4.9
$0.01
$0,081
$0,014
$5.3
2031
$0.27
$7.0
$0.79
$0,079
$0,014
$8.1
2032
$0.27
$6.9
$0.78
$0,078
$0,013
$8.1
2033
$0.26
$6.9
$0.77
$0,076
$0,013
$8.0
2034
$0.26
$6.8
$0,013
$0,075
$0,013
$7.2
2035
$0.26
$6.8
$0,013
$0,075
$0,013
$7.2
2036
$0.26
$6.8
$0,013
$0,074
$0,013
$7.2
2037
$0.25
$6.8
$0,013
$0,074
$0,013
$7.2
2038
$0.25
$6.8
$0,013
$0,073
$0,013
$7.2
2039
$0.25
$6.9
$0,013
$0,073
$0,013
$7.2
2040
$0.25
$6.9
$0,013
$0,073
$0,013
$7.3
2041
$0.25
$7.0
$0,013
$0,073
$0,013
$7.3
2042
$0.25
$7.0
$0,013
$0,073
$0,013
$7.3
2043
$0.25
$7.0
$0,013
$0,074
$0,013
$7.4
2044
$0.25
$7.1
$0,013
$0,074
$0,013
$7.4
2045
$0.25
$7.1
$0,013
$0,074
$0,013
$7.4
PV, 3%
$3.8
$91
$4.6
$1.1
$0.19
$100
PV, 7%
$2.8
$65
$4.0
$0.80
$0.14
$72
* Values show 2 significant digits. Note that the Information Collection Request costs addressed in Section XIII of
the preamble would fall within the "Other" indirect costs shown here.
363
-------
Table 7-69: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, CNG, Millions
of 2017 dollars *
Calendar
Year
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other
Indirect
Costs
Profits
Total Technology
Costs
2027
$0.30
$3.5
$1.2
$0,087
$0,015
$5.0
2028
$0.29
$3.5
$1.2
$0,084
$0,015
$5.0
2029
$0.28
$3.5
$1.2
$0,083
$0,014
$5.0
2030
$0.28
$3.5
$0,014
$0,081
$0,014
$3.9
2031
$0.27
$3.5
$0,014
$0,079
$0,014
$3.9
2032
$0.27
$3.5
$0,013
$0,078
$0,013
$3.8
2033
$0.26
$3.4
$0,013
$0,076
$0,013
$3.8
2034
$0.26
$3.4
$0,013
$0,075
$0,013
$3.8
2035
$0.26
$3.4
$0,013
$0,075
$0,013
$3.8
2036
$0.26
$3.4
$0,013
$0,074
$0,013
$3.8
2037
$0.25
$3.4
$0,013
$0,074
$0,013
$3.8
2038
$0.25
$3.4
$0,013
$0,073
$0,013
$3.8
2039
$0.25
$3.4
$0,013
$0,073
$0,013
$3.8
2040
$0.25
$3.5
$0,013
$0,073
$0,013
$3.8
2041
$0.25
$3.5
$0,013
$0,073
$0,013
$3.8
2042
$0.25
$3.5
$0,013
$0,073
$0,013
$3.8
2043
$0.25
$3.5
$0,013
$0,074
$0,013
$3.9
2044
$0.25
$3.5
$0,013
$0,074
$0,013
$3.9
2045
$0.25
$3.6
$0,013
$0,074
$0,013
$3.9
PV, 3%
$3.8
$50
$3.4
$1.1
$0.19
$58
PV, 7%
$2.8
$36
$3.1
$0.80
$0.14
$43
* Values show 2 significant digits. Note that the Information Collection Request costs addressed in Section XIII of
the preamble would fall within the "Other" indirect costs shown here.
Table 7-70: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, CNG by
Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other Indirect
Costs
Profits
Total Technology
Costs
HHD8
$3.7
$89
$4.5
$1.1
$0.19
$99
Urban
bus
$0,094
$2.4
$0.12
$0,027
$0.0047
$2.6
Sum
$3.8
$91
$4.6
$1.1
$0.19
$100
* Values show 2 significant digits.
Table 7-71: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, CNG by
Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other Indirect
Costs
Profits
Total Technology
Costs
HHD8
$3.7
$48
$3.3
$1.1
$0.19
$57
Urban
bus
$0,094
$1.3
$0,088
$0,027
$0.0047
$1.5
Sum
$3.8
$50
$3.4
$1.1
$0.19
$58
* Values show 2 significant digits.
364
-------
Table 7-72: Technology Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, CNG by
Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other Indirect
Costs
Profits
Total Technology
Costs
HHD8
$2.7
$63
$3.9
$0.78
$0.14
$70
Urban
bus
$0.1
$1.7
$0.10
$0,020
$0.0034
$1.9
Sum
$2.8
$65
$4.0
$0.80
$0.14
$72
* Values show 2 significant digits.
Table 7-73: Technology Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, CNG by
Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
Direct Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other Indirect
Costs
Profits
Total Technology
Costs
HHD8
$2.7
$35
$3.1
$0.78
$0.14
$42
Urban
bus
$0,068
$0.93
$0,081
$0,020
$0.0034
$1.1
Sum
$2.8
$36
$3.1
$0.80
$0.14
$43
* Values show 2 significant digits.
Table 7-74: Technology Cost Impacts Relative to the Baseline Case, All Fuels by Proposed Option, Millions of
2017 dollars, Present Values at 3% Discounting *
Direct
Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other
Indirect
Costs
Profits
Total
Technology
Costs
Option 1 increase
from Baseline
$13,000
$5,300
$1,000
$3,700
$630
$23,000
Option 2 increase
from Baseline
$13,000
$2,800
$980
$3,700
$630
$21,000
* Values show 2 significant digits. Note that the Information Collection Request costs addressed in Section XIII of
the preamble would fall within the "Other" indirect costs shown here.
Table 7-75: Technology Cost Impacts Relative to the Baseline Case, All Fuels by Proposed Option, Millions of
2017 dollars, Present Values at 7% Discounting *
Direct
Manufacturing
Costs
Warranty
Costs
R&D
Costs
Other
Indirect
Costs
Profits
Total
Technology
Costs
Option 1 increase
from Baseline
$9,200
$3,800
$800
$2,700
$460
$17,000
Option 2 increase
from Baseline
$9,200
$2,100
$790
$2,700
$460
$15,000
* Values show 2 significant digits. Note that the Information Collection Request costs addressed in Section XIII of
the preamble would fall within the "Other" indirect costs shown here.
7.3.2 Total Operating Costs
The series of tables shown here show emission repair costs, urea costs, pre-tax fuel costs and
total operating costs by fuel type, then by regulatory class for each of the proposed options.
365
-------
Values shown for a given calendar year are undiscounted values while discounted values are
presented at both 3 and 7 percent discount rates. All values are shown in 2017 dollars.
Note that the emission repair and fuel costs are shown as negative costs, or savings. This is
expected with respect to the fuel costs since the ORVR requirements are expected to reduce
gasoline consumption due to the capture of previously evaporated emissions. With respect to the
emission repair costs, we expect the baseline emission repair costs to increase year-over-year as
vehicles age, and thus the proposed longer warranty and useful life periods result in the
magnitude of the savings growing in the early years. However, eventually vehicles would begin
to age beyond their longer warranty and useful life periods and, we estimate, incur higher costs
than in the baseline because of the higher technology costs. This explains the lower magnitude of
emission repair savings in the later years shown in the tables. The early implementation vehicles-
-MY2027 through MY2035—are aging into more costly emission-repair rates, which is beginning
to offset the lower emission-repair rates of the MY2036 and later vehicles.
For Table 7-76 through Table 7-81, we note that, as discussed in Section 5.2.2.5, we did not
include diesel-fueled, engine-certified LHD2b3 vehicles in the baseline emission inventory, but
did include these vehicles in our emission inventory estimates for the proposed Options 1 and 2
control scenarios. We decided to include the engine-certified LHD2b3 vehicles in the emissions
inventories of the control scenario to better estimate the projected emissions of heavy-duty
vehicles in the control scenarios, even though it will conservatively underestimate the benefits of
controlling these vehicles. As a result, the NOx emission inventories may demonstrate an
increase in emissions from source types containing diesel-fueled LHD2b3 vehicles for certain
calendar years. We plan to update our baseline modeling to reflect any requirements for diesel-
fueled, engine-certified LHD2b3 vehicles in the final rulemaking. In addition, the MOVES
baseline emission rates for all LHD diesel (not just 2b3 vehicles) were incorrectly developed
using a warranty period of 100K instead of 50K, and thus we underestimated the LHD diesel
emissions impact of the longer warranty periods in the proposal and alternative scenarios. We
will update the baseline warranty period for LHD diesel in the FRM analysis.
366
-------
Table 7-76: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, Diesel, Millions
of 2017 dollars *
Calendar Year
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
2027
$16
$61
$0.0
$77
2028
-$55
$130
$0.0
$72
2029
-$220
$200
$0.0
-$21
2030
-$250
$270
$0.0
$20
2031
-$230
$370
$0.0
$140
2032
-$230
$460
$0.0
$230
2033
-$230
$530
$0.0
$290
2034
-$430
$590
$0.0
$160
2035
-$470
$650
$0.0
$180
2036
-$420
$720
$0.0
$300
2037
-$400
$800
$0.0
$400
2038
-$420
$890
$0.0
$470
2039
-$450
$940
$0.0
$490
2040
-$480
$990
$0.0
$500
2041
-$470
$1,000
$0.0
$560
2042
-$470
$1,100
$0.0
$610
2043
-$460
$1,100
$0.0
$660
2044
-$450
$1,200
$0.0
$710
2045
-$440
$1,200
$0.0
$760
PV, 3%
-$4,600
$9,100
$0.0
$4,400
PV, 7%
-$3,100
$5,800
$0.0
$2,700
* Values show 2 significant digits.
367
-------
Table 7-77: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, Diesel, Millions
of 2017 dollars *
Calendar Year
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
2027
$16
$56
$0.0
$72
2028
-$55
$120
$0.0
$61
2029
-$150
$180
$0.0
$34
2030
-$210
$250
$0.0
$38
2031
-$200
$320
$0.0
$120
2032
-$89
$400
$0.0
$310
2033
-$56
$460
$0.0
$400
2034
-$8.6
$510
$0.0
$500
2035
$47
$560
$0.0
$610
2036
$120
$610
$0.0
$730
2037
$210
$650
$0.0
$860
2038
$290
$690
$0.0
$980
2039
$370
$730
$0.0
$1,100
2040
$410
$770
$0.0
$1,200
2041
$450
$810
$0.0
$1,300
2042
$500
$840
$0.0
$1,300
2043
$530
$870
$0.0
$1,400
2044
$560
$910
$0.0
$1,500
2045
$580
$940
$0.0
$1,500
PV, 3%
$1,900
$7,400
$0.0
$9,400
PV, 7%
$910
$4,800
$0.0
$5,700
* Values show 2 significant digits.
Table 7-78: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, Diesel by
Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
LHD
-$70
$10
$0.0
-$60
LHD45
-$510
$420
$0.0
-$85
MHD67
-$620
$950
$0.0
$330
HHD8
-$3,400
$7,600
$0.0
$4,200
Urban bus
-$21
$96
$0.0
$76
Sum
-$4,600
$9,100
$0.0
$4,400
* Values show 2 significant digits.
Table 7-79: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, Diesel by
Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
LHD
-$32
-$1.3
$0.0
-$33
LHD45
-$330
$340
$0.0
$9.4
MHD67
$150
$750
$0.0
$900
HHD8
$2,100
$6,200
$0.0
$8,400
Urban bus
$21
$80
$0.0
$100
Sum
-$4,600
$9,100
$0.0
$4,400
* Values show 2 significant digits.
368
-------
Table 7-80: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, Diesel by
Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
LHD
-$42
$6.2
$0.0
-$36
LHD45
-$310
$270
$0.0
-$39
MHD67
-$380
$610
$0.0
$220
HHD8
-$2,300
$4,900
$0.0
$2,500
Urban bus
-$12
$61
$0.0
$48
Sum
-$3,100
$5,800
$0.0
$2,700
* Values show 2 significant digits.
Table 7-81: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, Diesel by
Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
LHD
-$21
$0
$0.0
-$21
LHD45
-$220
$220
$0.0
$3.4
MHD67
$75
$490
$0.0
$560
HHD8
$1,100
$4,000
$0.0
$5,100
Urban bus
$12
$51
$0.0
$63
Sum
$910
$4,800
$0.0
$5,700
* Values show 2 significant digits.
Table 7-82: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, Gasoline,
Millions of 2017 dollars
Emission
Pre-tax
Fuel Costs
Total
Calendar Year
Repair
Costs
Urea Costs
Operating
Costs
2027
$0.27
$0.0
-$0.80
-$0.53
2028
$0.34
$0.0
-$1.6
-$1.3
2029
-$0.65
$0.0
-$2.5
-$3.2
2030
-$2.1
$0.0
-$3.4
-$5.5
2031
-$4.0
$0.0
-$4.3
-$8.4
2032
-$5.2
$0.0
-$5.2
-$10
2033
-$4.9
$0.0
-$6.3
-$11
2034
-$4.1
$0.0
-$7.3
-$11
2035
-$3.0
$0.0
-$8.3
-$11
2036
-$3.5
$0.0
-$9.2
-$13
2037
-$5.7
$0.0
-$10
-$16
2038
-$7.6
$0.0
-$11
-$19
2039
-$6.5
$0.0
-$12
-$18
2040
-$4.9
$0.0
-$13
-$18
2041
-$4.6
$0.0
-$13
-$18
2042
-$4.4
$0.0
-$14
-$18
2043
-$4.3
$0.0
-$14
-$19
2044
-$4.3
$0.0
-$15
-$19
2045
-$4.2
$0.0
-$16
-$20
PV, 3%
-$52
$0.0
-$120
-$170
PV, 7%
-$34
$0.0
-$73
-$110
* Values show 2 significant digits.
369
-------
Table 7-83: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, Gasoline,
Millions of 2017 dollars
Emission
Pre-tax
Fuel Costs
Total
Calendar Year
Repair
Costs
Urea Costs
Operating
Costs
2027
$0.27
$0.0
-$0.79
-$0.52
2028
$0.34
$0.0
-$1.6
-$1.3
2029
-$0.65
$0.0
-$2.5
-$3.1
2030
-$2.4
$0.0
-$3.4
-$5.8
2031
-$4.6
$0.0
-$4.3
00
00
&
1
2032
-$6.5
$0.0
-$5.1
-$12
2033
-$7.5
$0.0
-$6.2
-$14
2034
-$7.7
$0.0
-$7.2
-$15
2035
-$7.3
$0.0
-$8.1
-$15
2036
-$6.0
$0.0
-$9.0
-$15
2037
-$4.4
$0.0
-$10
-$14
2038
-$2.8
$0.0
-$11
-$13
2039
-$1.1
$0.0
-$11
-$13
2040
$0.8
$0.0
-$12
-$11
2041
$1.4
$0.0
-$13
-$11
2042
$2.0
$0.0
-$13
-$12
2043
$2.4
$0.0
-$14
-$12
2044
$2.7
$0.0
-$15
-$12
2045
$3.0
$0.0
-$15
-$12
PV, 3%
-$32
$0.0
-$110
-$140
PV, 7%
-$25
$0.0
-$71
-$97
* Values show 2 significant digits.
Table 7-84: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, Gasoline by
Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
LHD45
-$37
$0
-$75
-$110
MHD67
-$11
$0
-$29
-$40
HHD8
-$4.3
$0
-$11
-$15
Sum
-$52
$0
-$120
-$170
* Values show 2 significant digits.
Table 7-85: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, Gasoline by
Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
LHD45
-$26
$0.0
-$74
-$100
MHD67
-$2.1
$0.0
-$28
-$30
HHD8
-$3.8
$0.0
-$10
-$14
Sum
-$32
$0.0
-$110
-$140
* Values show 2 significant digits.
370
-------
Table 7-86: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, Gasoline by
Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
LHD45
-$24
$0
-$48
-$72
MHD67
-$7.1
$0
-$18
-$26
HHD8
-$2.8
$0
-$6.7
-$9.6
Sum
-$34
$0
-$73
-$110
* Values show 2 significant digits.
Table 7-87: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, Gasoline by
Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
LHD45
-$20
$0.0
-$47
-$67
MHD67
-$2.3
$0.0
-$18
-$20
HHD8
-$2.8
$0.0
-$6.5
-$9.2
Sum
-$25
$0.0
-$71
-$97
* Values show 2 significant digits.
Table 7-88: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, CNG, Millions of
2017 dollars
Calendar Year
Emission Repair
Costs, Owner
Urea
Costs
Fuel
Costs
Total
Operating
Costs
2027
$0.0035
$0.0
$0.0
$0.0035
2028
-$0.21
$0.0
$0.0
-$0.21
2029
-$0.66
$0.0
$0.0
-$0.66
2030
-$1.5
$0.0
$0.0
-$1.5
2031
-$2.8
$0.0
$0.0
-$2.8
2032
-$4.6
$0.0
$0.0
-$4.6
2033
-$6.5
$0.0
$0.0
-$6.5
2034
-$7.7
$0.0
$0.0
-$7.7
2035
-$8.4
$0.0
$0.0
-$8.4
2036
-$8.9
$0.0
$0.0
-$8.9
2037
-$10
$0.0
$0.0
-$10
2038
-$11
$0.0
$0.0
-$11
2039
-$13
$0.0
$0.0
-$13
2040
-$15
$0.0
$0.0
-$15
2041
-$16
$0.0
$0.0
-$16
2042
-$16
$0.0
$0.0
-$16
2043
-$16
$0.0
$0.0
-$16
2044
-$17
$0.0
$0.0
-$17
2045
-$17
$0.0
$0.0
-$17
PV, 3%
-$120
$0.0
$0.0
-$120
PV, 7%
-$71
$0.0
$0.0
-$71
* Values show 2 significant digits.
371
-------
Table 7-89: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, CNG, Millions of
2017 dollars
Calendar Year
Emission Repair
Costs, Owner
Urea
Costs
Fuel
Costs
Total
Operating
Costs
2027
$0.0035
$0.0
$0.0
$0.0035
2028
-$0.21
$0.0
$0.0
-$0.21
2029
-$0.66
$0.0
$0.0
-$0.66
2030
-$1.5
$0.0
$0.0
-$1.5
2031
-$2.8
$0.0
$0.0
-$2.8
2032
-$3.9
$0.0
$0.0
-$3.9
2033
-$5.1
$0.0
$0.0
-$5.1
2034
-$5.9
$0.0
$0.0
-$5.9
2035
-$6.4
$0.0
$0.0
-$6.4
2036
-$6.4
$0.0
$0.0
-$6.4
2037
-$6.3
$0.0
$0.0
-$6.3
2038
-$6.1
$0.0
$0.0
-$6.1
2039
-$5.9
$0.0
$0.0
-$5.9
2040
-$5.9
$0.0
$0.0
-$5.9
2041
-$6.0
$0.0
$0.0
-$6.0
2042
-$5.7
$0.0
$0.0
-$5.7
2043
-$5.6
$0.0
$0.0
-$5.6
2044
-$5.6
$0.0
$0.0
-$5.6
2045
-$5.6
$0.0
$0.0
-$5.6
PV, 3%
-$60
$0.0
$0.0
-$60
PV, 7%
-$39
$0.0
$0.0
-$39
* Values show 2 significant digits.
Table 7-90: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, CNG by
Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
HHD8
-$110
$0.0
$0.0
-$110
Urban bus
-$2.8
$0.0
$0.0
-$2.8
Sum
-$120
$0.0
$0.0
-$120
* Values show 2 significant digits.
Table 7-91: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, CNG by
Regulatory Class, Millions of 2017 dollars, Present Values at 3% Discounting *
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
HHD8
-$59
$0.0
$0.0
-$59
Urban bus
-$0.89
$0.0
$0.0
-$0.89
Sum
-$60
$0.0
$0.0
-$60
* Values show 2 significant digits.
372
-------
Table 7-92: Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline Case, CNG by
Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
HHD8
-$70
$0
$0
-$70
Urban bus
-$1.7
$0
$0
-$1.7
Sum
-$71
$0
$0
-$71
* Values show 2 significant digits.
Table 7-93: Operating Cost Impacts of the Proposed Option 2 Relative to the Baseline Case, CNG by
Regulatory Class, Millions of 2017 dollars, Present Values at 7% Discounting *
Emission Repair Costs
Urea Costs
Pre-tax Fuel Costs
Total Operating Costs
HHD8
-$39
$0.0
$0.0
-$39
Urban bus
-$0.57
$0.0
$0.0
-$0.57
Sum
-$39
$0.0
$0.0
-$39
* Values show 2 significant digits.
Table 7-94: Operating Cost Impacts Relative to the Baseline Case, All Fuels by Proposed Option, Millions of
2017 dollars, Present Values at 3% Discounting *
Emission Repair
Costs
Urea
Costs
Pre-tax Fuel
Costs
Total Operating
Costs
Option 1 increase from
Baseline
-$4,800
$9,100
-$120
$4,200
Option 2 increase from
Baseline
$1,800
$7,400
-$110
$9,200
* Values show 2 significant digits.
Table 7-95: Operating Cost Impacts Relative to the Baseline Case, All Fuels by Proposed Option, Millions of
2017 dollars, Present Values at 7% Discounting *
Emission Repair
Costs
Urea
Costs
Pre-tax Fuel
Costs
Total Operating
Costs
Option 1 increase from
Baseline
-$3,200
$5,800
-$73
$2,600
Option 2 increase from
Baseline
$840
$4,800
-$71
$5,500
* Values show 2 significant digits.
Note that the ORVR requirements would result in previously evaporated gasoline being used
by in the engines of gasoline vehicles. We have estimated the cost savings that owner/operators
would experience and present those in Section 7.2.2. In this section, we also show the pre-tax
fuel savings that are ultimately part of the benefit-cost analysis presented in Chapter 9. Table
7-96 and Table 7-97 show the impacts on fuel tax revenues that would be expected from these
changes under the proposed Options 1 and 2, respectively.
373
-------
Table 7-96: Fuel Cost and Transfer Impacts of the Proposed Option 1 Relative to the Baseline Case, Gasoline,
Millions of 2017 dollars
Calendar Year
Retail Fuel Costs
Pre-tax Fuel Costs
Tax Revenues
2027
-$0.95
-$0.80
-$0.15
2028
-$1.9
-$1.6
-$0.31
2029
-$3.0
-$2.5
-$0.47
2030
-$4.0
-$3.4
-$0.62
2031
-$5.1
-$4.3
-$0.77
2032
-$6.1
-$5.2
-$0.92
2033
-$7.4
-$6.3
-$1.1
2034
-$8.6
-$7.3
-$1.2
2035
-$9.7
-$8.3
-$1.4
2036
-$11
-$9.2
-$1.5
2037
-$12
-$10
-$1.7
2038
-$13
-$11
-$1.8
2039
-$14
-$12
-$1.9
2040
-$15
-$13
-$2.0
2041
-$15
-$13
-$2.1
2042
-$16
-$14
-$2.1
2043
-$17
-$14
-$2.2
2044
-$17
-$15
-$2.3
2045
-$18
-$16
-$2.3
PV, 3%
-$130
-$120
-$19
PV, 7%
-$85
-$73
-$12
Table 7-97: Fuel Cost and Transfer Impacts of Proposed Option 2 Relative to the Baseline Case, Gasoline,
Millions of 2017 dollars
Calendar Year
Retail Fuel Costs
Pre-tax Fuel Costs
Tax Revenues
2027
-$0.94
-$0.79
-$0.15
2028
-$1.9
-$1.6
-$0.30
2029
-$2.9
-$2.5
-$0.46
2030
-$4.0
-$3.4
-$0.62
2031
-$5.0
-$4.3
-$0.76
2032
-$6.0
-$5.1
-$0.90
2033
-$7.2
-$6.2
-$1.1
2034
-$8.4
-$7.2
-$1.2
2035
-$9.5
-$8.1
-$1.4
2036
-$11
-$9.0
-$1.5
2037
-$12
-$10
-$1.6
2038
-$12
-$11
-$1.7
2039
-$13
-$11
-$1.8
2040
-$14
-$12
-$1.9
2041
-$15
-$13
-$2.0
2042
-$16
-$13
-$2.1
2043
-$16
-$14
-$2.1
2044
-$17
-$15
-$2.2
2045
-$17
-$15
-$2.2
PV, 3%
-$130
-$110
-$18
PV, 7%
-$83
-$71
-$12
374
-------
7.3.3 Total Program Costs
The series of tables shown here present technology costs, operating costs and the sum of the
two for each proposed option. Values shown for a given calendar year are undiscounted values
while discounted values are presented at both 3 and 7 percent discount rates. All values are
shown in 2017 dollars.
Table 7-98: Total Technology & Operating Cost Impacts of the Proposed Option 1 Relative to the Baseline
Case, All Regulatory Classes and All Fuels, Millions of 2017 dollars
Calendar Year
Total
Technology
Costs
Total
Operating
Costs
Sum
2027
$1,800
$77
$1,800
2028
$1,700
$71
$1,800
2029
$1,700
-$25
$1,700
2030
$1,600
$13
$1,600
2031
$1,700
$120
$1,900
2032
$1,700
$210
$1,900
2033
$1,700
$280
$1,900
2034
$1,600
$140
$1,700
2035
$1,600
$160
$1,700
2036
$1,600
$280
$1,900
2037
$1,600
$380
$1,900
2038
$1,600
$440
$2,000
2039
$1,600
$450
$2,000
2040
$1,600
$470
$2,000
2041
$1,600
$520
$2,100
2042
$1,600
$570
$2,100
2043
$1,600
$620
$2,200
2044
$1,600
$670
$2,200
2045
$1,600
$720
$2,300
PV. 3%
$23,000
$4,200
$27,000
PV. 7%
$17,000
$2,600
$19,000
Annualized, 3%
$1,600
$290
$1,900
Annualized, 7%
$1,600
$250
$1,900
* Values show 2 significant digits.
375
-------
Table 7-99: Total Technology & Operating Cost Impacts of Proposed Option 2 Relative to the Baseline Case,
All Regulatory Classes and All Fuels, Millions of 2017 dollars
Calendar Year
Total
Technology
Costs
Total
Operating
Costs
Sum
2027
$1,700
$72
$1,800
2028
$1,700
$60
$1,700
2029
$1,600
$31
$1,700
2030
$1,500
$31
$1,500
2031
$1,500
$110
$1,600
2032
$1,400
$300
$1,700
2033
$1,400
$380
$1,800
2034
$1,400
$480
$1,900
2035
$1,400
$580
$2,000
2036
$1,400
$710
$2,100
2037
$1,400
$840
$2,200
2038
$1,400
$960
$2,300
2039
$1,400
$1,100
$2,400
2040
$1,400
$1,200
$2,500
2041
$1,400
$1,200
$2,600
2042
$1,400
$1,300
$2,700
2043
$1,400
$1,400
$2,800
2044
$1,400
$1,400
$2,800
2045
$1,400
$1,500
$2,900
PV, 3%
$21,000
$9,200
$30,000
PV, 7%
$15,000
$5,500
$21,000
Annualized, 3%
$1,400
$640
$2,100
Annualized, 7%
$1,500
$540
$2,000
* Values show 2 significant digits.
376
-------
Chapter 8 Estimated Benefits
8.1 Overview
This chapter describes the methods used to estimate health benefits from reducing
concentrations of ozone and PM2.5. For the proposed rulemaking, we have quantified and
monetized health impacts in 2045, representing projected impacts associated with a year when
the program would be fully implemented and when most of the regulated fleet would have turned
over. There are also benefits associated with the standards that, if quantified and monetized,
would increase the total monetized benefits. These unquantified benefits are discussed in Section
8.8 of this chapter. Overall, we estimate that the proposed program would lead to a substantial
decrease in adverse PM2.5- and ozone-related health impacts in 2045.
We adopt an updated analysis approach that was recently used to quantify the benefits of
changes in PM2.5 and ozone in the final Revised Cross-State Air Pollution Rule (CSAPR) Update
RIA.377'nnnnnn While the steps to performing a criteria pollutant benefits analysis remain
unchanged from past mobile source rulemakings (e.g., Tier 3 Motor Vehicle Emission and Fuel
Standards Final Rule)378, the final CSAPR RIA updated the suite of quantified health endpoints
included in the benefits analysis, as well as the data used to quantify each health endpoint, to
reflect more recent scientific evidence. These updates were based on information drawn from the
recent PM2.5 and ozone Integrated Science Assessments (ISAs), which were reviewed by the
Clean Air Science Advisory Committee (CASAC) and the public,379'380 and are summarized in a
technical support document (TSD) originally published for the final Revised CSAPR Update
titled Estimating PM2.5- and Ozone-Attributable Health Benefits 381>000000
8.2 Updates to EPA's Human Health Benefits Methods
When the RIA for the Final Revised CSAPR Update was published in March 2021, the
Agency adopted an updated approach for estimating benefits that incorporated an array of
science-policy and technical changes since the previous reviews of the PM2.5 standards in 2012
and the ozone standards in 2015. As part of this process, the Agency carefully considered: (1) the
extent to which the science supports the existence of a causal relationship between that pollutant
and the adverse effect; (2) whether suitable epidemiologic studies exist to support quantifying
health impacts; and (3) whether robust economic approaches are available for estimating the
value of the impact of reducing human exposure to the pollutant. The Estimating PM2.5- and
Ozone-Attributable Health Benefits TSD provides a full discussion of the Agency's updated
approach for quantifying the number and value of estimated air pollution-related impacts. Below,
we summarize the rationale for selecting health endpoints to quantify; the demographic, health
NNmMN On March 15, 2021, EPA finalized the Revised Cross-State Air Pollution Rule Update for the 2008 ozone
National Ambient Air Quality Standards (NAAQS). Starting in the 2021 ozone season, the rule will require
additional emissions reductions of nitrogen oxides (NOx) from power plants in 12 states.
https://www.epa.gov/csapr/revised-cross-state-air-pollution-rule-update.
oooooo Qn june io, 2021, EPA announced that it will reconsider the previous administration's decision to retain the
PM NAAQS. To the extent that new health information is introduced in the reconsideration, the Agency may update
its benefits methods in future analyses.
377
-------
and economic data used; modeling assumptions; and our techniques for quantifying
uncertainty.pppppp EPA followed a five-step approach:
• Establish criteria for identifying studies and risk estimates most appropriate to inform
a PM2.5 and O3 benefit analysis for an RIA. Study criteria, such as study design,
location, population characteristics, and other attributes, were used to identify the most
suitable estimates. This step precedes health endpoint identification to ensure impartial
health endpoint identification and prevent identification of non-quantifiable endpoints.
• Identify pollutant-attributable health effects for which the ISA reports strong evidence
and that may be quantified in a benefits assessment. EPA considered new evidence
reported in the recent PM and ozone ISAs and clinically significant outcomes (e.g.,
premature mortality and hospital admissions) for which endpoint-specific baseline
incidence data is available. While ISAs form causal determinations for broad endpoint
categories (e.g., respiratory effects), which are generally preferred over specific health
endpoints (e.g., hay fever symptoms) for comprehensive benefits assessments, they do
not make causal determinations for each specific health endpoint. Instead, the ISAs
provide information on the strength and consistency of the evidence supporting more
specific endpoints within each broad category. The strength and consistency of
evidence supporting relationships with specific health endpoints, together with the
broad category causality determinations, are used when identifying specific health
endpoints for inclusion in benefits assessments. New ISA evidence was considered
sufficient for inclusion in the benefits assessment if the ISA determined the broad
heath endpoint category was causally related to pollutant exposure, the ISA
determined that the specific health endpoint is a biologically plausible health effect of
exposure, and the ISA found strong and consistent support relating the specific health
endpoint with pollutant exposure.382
• Collect baseline incidence and prevalence estimates and demographic information.
EPA develops either daily or annual baseline incidence and prevalence rates at the
most geographically- and age-specific levels feasible for each health endpoint
assessed. EPA uses population projections based on economic forecasting models
developed by Woods and Poole, Inc.383 The Woods and Poole database contains
county-level projections of population by age, sex, and race out to 2050, relative to a
baseline using the 2010 Decennial Census.
• Develop economic unit values. To directly compare benefits estimates associated with
a rulemaking to cost estimates, the number of instances of each air pollution-
attributable health impact must be converted to a monetary value. This requires a
valuation estimate for each unique health endpoint, and potentially also discounting if
the benefits are expected to accrue over more than a single year. EPA develops
valuation estimates at the most age-refined level feasible for each health endpoint
assessed.
• Characterize uncertainty associated with quantified benefits estimates. Building on
EPA's current methods for characterizing uncertainty, these approaches will include,
pppppp Updated information has been incorporated into BenMAP-CE version 1.5.8 https://www.epa.gov/benmap.
378
-------
among others, reporting confidence intervals calculated from risk estimates and
separate quantification using multiple studies and risk estimates for particularly
influential endpoints (e.g., mortality https://www.epa.gov/benmap risk), and
approaches for aggregating and representing the results of multiple studies evaluating
a particular health endpoint.QQQQQQ
Since publication of the Final Revised CSAPR Update and the accompanying Estimating
PM2.5- and Ozone-Attributable Health Benefits TSD, the Agency has made two additional, minor
updates to its benefits approach described below.
Use ozone metrics and ozone seasons consistent with underlying health studies. Often, time
and resource limitations constrain an analysis to the use of one ozone metric and/or one ozone
seasonal definition. For example, the Final Revised CSAPR Update converted all ozone risk
estimates into a standard maximum daily 8-hour average (MDA8) metric and applied those
risk estimates to a standard May-September ozone warm season. Both approaches to
standardization introduce some uncertainty into the analysis due to the potential for metric
and seasonal mismatch with the underlying health studies, though we expect this mismatch
only has a limited effect on the magnitude of related health incidence. For the benefits
analysis of the proposed program, we rely on full-form photochemical air quality modeling
concentration surfaces derived using the Community Multi-scale Air Quality Model
(CMAQ).rrrrrr Having run the model, we are able to create ozone concentration surfaces
with exposure metrics and seasons that match as best as possible the same metric and season
used by each study. See draft RIA Chapter 8.2.1 below and Sections 6.5.12 and 6.5.13 of the
TSD for more a more detailed discussion.
Use updated income growth adjustment factors. As discussed in Sections 5.4 and 6.4.3 of the
TSD, evidence and theory suggest that one's willingness-to-pay (WTP) for health and
environmental improvements should increase as real income increases. To account for this,
we adjust WTP-based valuation estimates, such as the Value of a Statistical Life (VSL), to
account for the growth in income over time. This adjustment is a combination of data on
income growth and estimated income elasticity of demand, which measures the
responsiveness of the quantity demanded of a good to the change in the income of the people
demanding the good. In previous analyses, projections of income growth were capped at
2026. For the benefits analysis of the proposed program, we have estimated income growth
adjustment factors out to 2045 using future changes in annual income based on data presented
in the Annual Energy Outlook (AEO) 2020, a report prepared by the U.S. Energy Information
Administration (EIA).384
Estimating the health benefits of reductions in PM2.5 and O3 exposure begins with estimating
the change in exposure for each individual and then estimating the change in each individual's
risks for those health outcomes affected by exposure. The benefit of the reduction in each health
risk is based on the exposed individual's WTP for the risk change, assuming that each outcome
QQQQQQ We consider study quality, inter-study heterogeneity, and redundancy where epidemiologic risk estimates are
combined or aggregated.
111111111111 As noted in Section VII, due to resource constraints we only conducted air quality modeling for the proposed
Option 1. We have also used year-over-year Option 2 NOx emissions reductions to scale the total benefits associated
with Option 1 to derive a best estimate of criteria pollutant benefits associated with Option 2 (see draft RIA Chapter
8.7for details).
379
-------
is independent of one another. The greater the magnitude of the risk reduction from a given
change in concentration, the greater the individual's WTP, all else equal. The social benefit of
the change in health risks equals the sum of the individual WTP estimates across all of the
affected individuals.ssssss We conduct this analysis by adapting primary research - specifically,
air pollution epidemiology studies and economic value studies - from similar contexts. This
approach is sometimes referred to as "benefits transfer." Below we describe the procedure we
follow for quantifying and monetizing the health benefits associated with reduced human
exposure to PM2.5 and ozone.
8.3 Health Impact Assessment for PM2.5 and Ozone
There are four distinct steps the Agency follows when conducting a health impacts
assessment, each of which are described in this section: (1) prepare air quality modeling data for
health impacts analysis; (2) select air pollution health endpoints to quantify; (3) calculate counts
of air pollution effects using a health impact function; (4) specify the health impact function with
concentration-response parameters drawn from the epidemiological literature.
8.3.1 Preparing Air Quality Modeling Data for Health Impacts Analysis
In draft RIA Chapter 5, we present the emissions that would be reduced if the proposed
Options 1 or 2 were in place, including NOx, direct PM, and VOCs, all of which contribute to
ambient concentrations of PM2.5 and ozone. These reduced emissions would benefit public health
and the environment since exposure to ozone and PM2.5 is linked to adverse public health and
environmental effects.TTTTTT In draft RIA Chapter 6, we summarize the air quality modeling
methods and results for the proposed Option 1 uuuuuu These air quality results, measured in
terms of ambient concentrations of PM2.5 and ozone, are in turn associated with human
populations to estimate changes in health effects. This section describes how the CMAQ
modeling output was converted into a format suitable for the health impacts analysis.
The first step was to extract 2016 base year predicted hourly, surface-layer PM2.5 and ozone
concentrations for each grid cell directly from the standard CMAQ output files (at a 12-km by
12-km resolution). For ozone, we generated predicted ozone concentration surfaces for each of
three different warm seasons defined by the underlying health studies used in the analysis: April-
ssssss r[a also reports the change in the sum of the risk, or the change in the total incidence, of a health
outcome across the population. If the benefit per unit of risk is invariant across individuals, the total expected
change in the incidence of the health outcome across the population can be multiplied by the benefit per unit of risk
to estimate the social benefit of the total expected change in the incidence of the health outcome.
tttttt \ye note (iul( detailed county-level emission inventories were generated for the air quality modeling that
supports the benefits analysis. However, national-scale emission inventories, which utilize national average values
for various input variables, were developed to demonstrate the emission impacts of the proposed program over time.
The national-scale approach is simpler but coarser compared to the approach used to develop the detailed
inventories needed for air quality modeling. For this reason, the modeled changes in emissions used to support the
air quality and benefits analyses are slightly smaller relative to the national-scale inventories used to represent the
emissions impacts of the proposed program. We do not expect the magnitude of the differences to materially impact
our cost-benefit conclusions. See draft RIA Chapter 5.4 for more details.
1,111,111,11 As noted in Chapter 5.4 of the draft RIA, while we refer to air quality modeling for the proposed Option 1,
there are differences between the proposed Option 1 standards, emission warranty, and useful life provisions
presented in Sections III and IV of the preamble and those included in the control scenario modeled for the air
quality analysis. Estimates of health benefits are based on our air quality analysis, and thus differences between
proposed Option 1 and modeling are not reflected in the benefits analysis.
380
-------
September, May-September, and June-August. These hourly model predictions were then
combined with monitored observations obtained from the Agency's Air Quality System (AQS)
to interpolate hourly ozone concentrations to 12-km by 12-km grid cells for the contiguous 48
states to create gridded 2016 surfaces informed by observational data. vvvvvv"wwwwww We then
converted these warm-season hourly ozone concentrations to an ozone metric of interest, such as
the daily maximum 8-hour average concentration or the daily maximum 1-hour average
concentration, again consistent with the underlying health studies used in the analysis. Gridded
fields of relative response factors (RRFs) were created for each ozone metric and warm season
definition of interest by dividing unadjusted future year (2045) CMAQ concentrations by
unadjusted 2016 base year CMAQ concentrations. Separate 12-km gridded RRFs were created
for the future year base case and policy cases for each metric/season combination. Then final
future year air quality surfaces were created by multiplying each of the RRF surfaces by the 2016
eVNA surface. These surfaces then served as inputs to the health impact functions of the benefits
analysis, contained within the Environmental Benefits Mapping and Analysis Program -
Community Edition (BenMAP-CE).
For PM2.5, we also used 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. A gridded field of PM2.5
concentrations was created by interpolating Federal Reference Monitor ambient data and
Interagency Monitoring of Protected Visual Environments (IMPROVE) ambient data. Gridded
fields of PM2.5 species concentrations were created by interpolating EPA Chemical Speciation
Network (CSN) ambient data and IMPROVE data. The ambient data were interpolated to the
CMAQ 12-km grid. Future-year estimates of PM2.5 were calculated using gridded RRFs applied
to gridded 2016 ambient PM2.5 and PM2.5 species concentrations.
The procedures for determining the RRFs are similar to those in EPA's Modeling Guidance
for Demonstrating Air Quality Goals for Ozone, PM2.5, and Regional Haze.385 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 proposed Option 1.
Table 8-1 provides ozone and PM2.5 metrics for those 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.
vvvvvv j]le 12-km grid squares contain the population data used in the health benefits analysis model, BenMAP-CE.
wwwwww xhis approach is a generalization of planar interpolation that is technically referred to as enhanced Voronoi
Neighbor Averaging (eVNA) spatial interpolation. See the BenMAP-CE manual for technical details, available for
download at http://www.epa.gov/benmap.
381
-------
Table 8-1: Summary of CMAQ-Derived Population-Weighted Ozone and PM2.5 Air Quality Metrics for
Health Benefits Endpoints Associated with Proposed Option 1
2045
Statistic3
Baseline
Changeb
Ozone Metric: National Population-Weighted Average (ppb)°
Daily Maximum 8-Hour Average
Concentration - May-September
39
0.69
Daily Maximum 8-Hour Average
Concentration - April-September
39
0.64
Daily Maximum 8-Hour Average
Concentration - June-August
38
0.76
Daily Maximum 1-Hour Average
Concentration - April-September
44
0.80
PM2 5 Metric: National Population-Weighted Average (ug/m3)°
Annual Average Concentration
7.3
0.034
a Ozone and PM2 5 metrics were calculated at the CMAQ grid-cell level for use in health effects estimates.
Ozone metrics were calculated over relevant time periods during daylight hours of each "ozone season."
Note that the national, population-weighted PM2 5 and ozone air quality metrics presented in this table
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 draft RIA Chapter 6, which represent the
average for areas with a current air quality monitor.
b The change is defined as the baseline value minus the control-case value.
0 Calculated by summing the product of the projected CMAQ grid-cell population and the estimated CMAQ
grid concentration and then dividing by the total population.
8.3.2 Selecting Air Pollution Health Endpoints to Quantify
As a first step in quantifying ozone and PIVfo.s-related human health impacts, the Agency
consults the Integrated Science Assessment for Ozone and Related Photochemical Oxidants
(Ozone ISA) and the Integrated Science Assessment for Particulate Matter (PM ISA). These two
documents synthesize the toxicological, clinical and epidemiological evidence to determine
whether each pollutant is causally related to an array of adverse human health outcomes
associated with either short-term (i.e., hours to less than one month) or long-term (i.e., one
month to years) exposure; for each outcome, the ISA reports this relationship to be causal, likely
to be causal, suggestive of a causal relationship, inadequate to infer a causal relationship or not
likely to be a causal relationship. The Agency estimates the incidence of air pollution effects for
those health endpoints above where the ISA classified as either causal or likely-to-be-causal.
In brief, the ISA for ozone found short-term exposures to ozone to be causally related to
respiratory effects, a "likely to be causal" relationship with metabolic effects and a "suggestive
of, but not sufficient to infer, a causal relationship" for central nervous system effects,
cardiovascular effects, and total mortality. The ISA reported that long-term exposures to ozone
are "likely to be causal" for respiratory effects including respiratory mortality, and a "suggestive
of, but not sufficient to infer, a causal relationship" for cardiovascular effects, reproductive
effects, central nervous system effects, metabolic effects, and total mortality. The PM ISA found
short-term exposure to PM2.5 to be causally related to cardiovascular effects and mortality (i.e.,
premature death), respiratory effects as likely-to-be-causally related, and a suggestive
relationship for metabolic effects and nervous system effects. The ISA identified cardiovascular
382
-------
effects and total mortality as being causally related to long-term exposure to PM2.5. A likely-to-
be-causal relationship was determined between long-term PM2.5 exposures and respiratory
effects, nervous system effects, and cancer effects and the evidence was suggestive of a causal
relationship for male and female reproduction and fertility effects, pregnancy and birth
outcomes, and metabolic effects.
Table 8-2 reports the effects we quantified and those we did not quantify in this RIA. The list
of benefit categories not quantified is not exhaustive. And, among the effects quantified, it might
not have been possible to quantify completely either the full range of human health impacts or
economic values. The table below omits health effects associated with NO2 exposure, and any
welfare effects such as acidification and nutrient enrichment; these effects are described in the
Ozone and PM NAAQS RIAs and summarized later in this chapter.386'387
383
-------
Table 8-2: Health Effects of Ambient Ozone and PM2.5
Category
Effect
Effect
Quantified
Effect
Monetized
More
Information
Premature
mortality from
exposure to PM2.5
Adult premature mortality from long-term exposure (age 65-
99 or age 30-99)
¦/
¦/
PMISA
Infant mortality (age <1)
¦/
¦/
PMISA
Nonfatal
morbidity from
exposure to PM2.5
Heart attacks (age >18)
¦/
PMISA
Hospital admissions—cardiovascular (ages 65-99)
¦/
¦/
PMISA
Emergency department visits— cardiovascular (age 0-99)
¦/
¦/
PMISA
Hospital admissions—respiratory (ages 0-18 and 65-99)
¦/
¦/
PMISA
Emergency room visits—respiratory (all ages)
¦/
¦/
PMISA
Cardiac arrest (ages 0-99; excludes initial hospital and/or
emergency department visits)
¦/
PMISA
Stroke (ages 65-99)
¦/
PMISA
Asthma onset (ages 0-17)
¦/
¦/
PMISA
Asthma symptoms/exacerbation (6-17)
¦/
¦/
PMISA
Lung cancer (ages 30-99)
¦/
¦/
PMISA
Allergic rhinitis (hay fever) symptoms (ages 3-17)
¦/
¦/
PMISA
Lost work days (age 18-65)
¦/
¦/
PMISA
Minor restricted-activity days (age 18-65)
¦/
¦/
PMISA
Hospital admissions—Alzheimer's disease (ages 65-99)
¦/
¦/
PMISA
Hospital admissions—Parkinson's disease (ages 65-99)
¦/
¦/
PMISA
Other cardiovascular effects (e.g., other ages)
—
—
PMISA2
Other respiratory effects (e.g., pulmonary function, non-
asthma ER visits, non-bronchitis chronic diseases, other ages
and populations)
—
—
PMISA2
Other nervous system effects (e.g., autism, cognitive decline,
dementia)
—
—
PMISA2
Metabolic effects (e.g., diabetes)
—
—
PMISA2
Reproductive and developmental effects (e.g., low birth
weight, pre-term births, etc.)
—
—
PMISA2
Cancer, mutagenicity, and genotoxicity effects
—
—
PMISA2
Mortality from
exposure to
ozone
Premature respiratory mortality from short-term exposure (0-
99)
¦/
¦/
Ozone ISA
Premature respiratory mortality from long-term exposure
(age 30-99)
¦/
¦/
Ozone ISA
Nonfatal
morbidity from
exposure to
ozone
Hospital admissions—respiratory (ages 65-99)
¦/
¦/
Ozone ISA
Emergency department visits—respiratory (ages 0-99)
¦/
¦/
Ozone ISA
Asthma onset (0-17)
¦/
¦/
Ozone ISA
Asthma symptoms/exacerbation (asthmatics age 5-17)
¦/
¦/
Ozone ISA
Allergic rhinitis (hay fever) symptoms (ages 3-17)
¦/
¦/
Ozone ISA
Minor restricted-activity days (age 18-65)
¦/
¦/
Ozone ISA
School absence days (age 5-17)
¦/
¦/
Ozone ISA
Decreased outdoor worker productivity (age 18-65)
—
—
Ozone ISA2
Metabolic effects (e.g., diabetes)
—
—
Ozone ISA2
Other respiratory effects (e.g., premature aging of lungs)
—
—
Ozone ISA2
Cardiovascular and nervous system effects
—
—
Ozone ISA2
Reproductive and developmental effects
—
—
Ozone ISA2
'Valuation estimate excludes initial hospital and/or emergency department visits.
2 Not quantified due to data availability limitations and/or because current evidence is only suggestive of causality.
384
-------
8.3.3 Calculating Counts of Air Pollution Effects Using the Health Impact Function
We use BenMAP-CE to quantify individual risk and counts of estimated premature deaths and
illnesses attributable to photochemical modeled changes in warm season average ozone
concentrations and annual mean PM2.5 for the year 2045 using a health impact function.388 A
health impact function combines information regarding the: concentration-response relationship
between air quality changes and the risk of a given adverse outcome; population exposed to the
air quality change; baseline rate of death or disease in that population; and, air pollution
concentration to which the population is exposed.
The following provides an example of a health impact function, in this case for PM2.5
mortality risk. We estimate counts of PM2.5-related total deaths (yij) during each year i (i=l,... ,1
where I is the total number of years analyzed) among adults aged 30 and older (a) in each county
in the contiguous U.S. j (j=l,... ,J where J is the total number of counties) as
yij= Ea yija
yija = moija x(ep-ACij-l) x Pija, Eq[l]
where moija is the baseline all-cause mortality rate for adults aged a=30-99 in county j in year
i stratified in 10-year age groups, P is the risk coefficient for all-cause mortality for adults
associated with annual average PM2.5 exposure, Cij is the annual mean PM2.5 concentration in
county j in year i, and Pija is the number of county adult residents aged a=30-99 in county j in
year i stratified into 5-year age groups.xxxxxx
The BenMAP-CE tool is pre-loaded with projected population from the Woods & Poole
company; cause-specific and age-stratified death rates from the Centers for Disease Control and
Prevention, projected to future years; recent-year baseline rates of hospital admissions,
emergency department visits and other morbidity outcomes from the Healthcare Cost and
Utilization Program and other sources; concentration-response parameters from the published
epidemiologic literature cited in the ISAs for fine particles and ground-level ozone; and, cost of
illness or WTP unit values for each endpoint.
8.3.4 Quantifying Ozone-Attributable Premature Mortality
In 2008, the National Academies of Science (NAS) issued a series of recommendations to
EPA regarding the procedure for quantifying and valuing ozone-related mortality due to short-
term exposures.389 Chief among these was that "... short-term exposure to ambient ozone is likely
to contribute to premature deaths" and the committee recommended that "ozone-related
mortality be included in future estimates of the health benefits of reducing ozone exposures..."
The NAS also recommended that".. .the greatest emphasis be placed on the multicity and
[National Mortality and Morbidity Air Pollution Studies (NMMAPS)] ... studies without
exclusion of the meta-analyses."
xxxxxx [n illustrative example, the air quality is resolved at the county level. For this RIA, we simulate air
quality concentrations at 12km by 12km grids. The BenMAP-CE tool assigns the rates of baseline death and disease
stored at the county level to the 12km by 12km grid cells using an area-weighted algorithm. This approach is
described in greater detail in the appendices to the BenMAP-CE user manual.
385
-------
Prior to the 2015 Ozone NAAQS RIA, the Agency estimated ozone-attributable premature
deaths using an NMMAPS-based analysis of total mortality, two multi-city studies of
cardiopulmonary and total mortality and effect estimates from three meta-analyses of non-
accidental mortality.390-391-392-393-394-395 Beginning with the 2015 Ozone NAAQS RIA, the
Agency began quantifying ozone-attributable premature deaths using two newer multi-city
studies of non-accidental mortality and one long-term cohort study of respiratory mortality.
396.397.398 2020 Ozone ISA included changes to the causality relationship determinations
between short-term exposures and total mortality, as well as including more recent
epidemiologic analyses of long-term exposure effects on respiratory mortality.399 In the final
2021 CSAPRRIA, mortality from long-term exposures was estimated using the Turner et al.
(2016) study extending and expanding the analysis of the American Cancer Society cohort
(ACS). Mortality for short-term exposures was estimated using the risk estimate parameters from
Zanobetti et al. (2008) and Katsouyanni et al. (2009) were pooled using a consistent ozone
season (May-Sept) and ozone metric (maximum daily 8-hour average).400 401402
In this RIA, ozone-attributable respiratory deaths are also estimated using the risk estimate
parameters described in the final 2021 CSAPRRIA. However, instead of pooling results derived
from different ozone air quality surfaces, we have chosen to use only the risk estimates derived
from the Katsouyanni et al. (2009) study because the study includes more cities across the United
States. Furthermore, this analysis uses modeled ozone concentration data that matches the ozone
metric (maximum daily 1-hour average) and season (April-September) used by Katsouyanni et
al. (2009) rather than using a default ozone season (May-Sept) and metric (maximum daily 8-
hour average) used in the CSAPR RIA.
8.3.5 Quantifying PM2.s-Attributable Premature Mortality
When quantifying PM-attributable cases of adult mortality, we use risk estimates from two
epidemiology studies examining two large population cohorts: the American Cancer Society
cohort and the Medicare cohort.403 404 The 2019 PM ISA concluded that the analyses of the ACS
and Medicare cohorts provide strong evidence of an association between long-term PM2.5
exposure and premature mortality with support from additional cohort studies. Both the ACS and
Medicare cohort studies have separate and distinct attributes that make them well-suited to being
used in a PM benefits assessment, so we present PM2.5 related effects derived using relative risk
estimates from both cohorts.
The PM ISA, which was reviewed by the Clean Air Scientific Advisory Committee of EPA's
Science Advisory Board (SAB-CASAC), concluded that there is a causal relationship between
mortality and both long-term and short-term exposure to PM2.5 based on the entire body of
scientific evidence.405 The PM ISA also concluded that the scientific literature supports the use
of a no-threshold log-linear model to portray the PM-mortality concentration-response
relationship while recognizing potential uncertainty about the exact shape of the concentration-
response relationship. The 2019 PM ISA, which informed the setting of the 2020 PM NAAQS,
reviewed available studies that examined the potential for a population-level threshold to exist in
the concentration-response relationship. Based on such studies, the ISA concluded that "evidence
from recent studies reduce uncertainties related to potential co-pollutant confounding and
continues to provide strong support for a linear, no-threshold concentration-response
relationship".406 Consistent with this evidence, the Agency historically has estimated health
impacts above and below the prevailing NAAQS.407"408"409"410"411"412"413"414"415"416"417"418"419"420"421
386
-------
Following this approach, we report the estimated PIvfc.s-related benefits (in terms of both
health impacts and monetized values) calculated using a log-linear concentration-response
function that quantifies risk from the full range of simulated PM2.5 exposures.422'423 When setting
the 2020 PM NAAQS, the EPA noted that:
. .an important consideration in characterizing the potential for additional public health
improvements associated with changes in PM2.5 exposure is whether concentration- response
relationships are linear across the range of concentrations or if nonlinear relationships exist along
any part of this range. Several recent studies examine this issue, and continue to provide
evidence of linear, no-threshold relationships between long-term PM2.5 exposures and all-cause
and cause-specific mortality.424 However, interpreting the shapes of these relationships,
particularly at PM2.5 concentrations near the lower end of the air quality distribution, can be
complicated by relatively low data density in the lower concentration range, the possible
influence of exposure measurement error, and variability among individuals with respect to air
pollution health effects [85 FR 82696],425
Hence, we are most confident in the size of the risks estimated from simulated PM2.5
concentrations that coincide with the bulk of the observed PM concentrations in the
epidemiological studies that are used to estimate the benefits. Likewise, we are less confident in
the risk we estimate from simulated PM2.5 concentrations that fall below the bulk of the observed
data in these studies.
To give readers insight to the level of uncertainty in the estimated PM2.5 mortality benefits at
lower ambient concentrations, we report the estimated PM benefits as a distribution, identifying
points along this distribution corresponding to the Lowest Reported Levels (LRLs) of each long-
term exposure mortality study and the PM NAAQS (see Figure 8-2 below). In addition to adult
mortality discussed above, we use risk estimates from a multi-city study to estimate PM-related
infant mortality.426
8.4 Economic Valuation Methodology for Health Benefits
We next quantify the economic value of the ozone and PM2.5-related deaths and illnesses
estimated above. Changes in ambient concentrations of air pollution generally yield small
changes in the risk of future adverse health effects for a large number of people. Therefore, the
appropriate economic measure is WTP for changes in risk of a health effect. For some health
effects, such as hospital admissions, WTP estimates are not generally available, so we use the
cost of treating or mitigating the effect. These cost-of-illness (COI) estimates are typically a
lower bound estimate of 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. The
unit values applied in this analysis are provided in Table 21 of the Estimating PM2.5- and Ozone-
Attributable Health Benefits TSD.
The estimated value of avoided premature deaths (PM2.5 plus ozone) account for between 84
percent or 94 percent of total monetized benefits depending on the studies used. The value for
the projected reduction in the risk of premature mortality is the subject of continuing discussion
within the economics and public policy analysis community. Following the advice of the SAB's
Environmental Economics Advisory Committee (SAB-EEAC), EPA currently uses the VSL
approach in calculating estimates of mortality benefits, because we believe this calculation
provides the most reasonable single estimate of an individual's willingness to trade off money
387
-------
for changes in the risk of death.427 The VSL approach is a summary measure for the value of
small changes in the risk of death experienced by a large number of people.
EPA continues work to update its guidance on valuing mortality risk reductions, and the
Agency consulted several times with the SAB-EEAC on this issue. Until updated guidance is
available, the Agency determined that a single, peer-reviewed estimate applied consistently, best
reflects the SAB-EEAC advice it has received. Therefore, EPA applies the VSL that was vetted
and endorsed by the SAB in the Guidelines for Preparing Economic Analyses while the Agency
continues its efforts to update its guidance on this issue.428 This approach calculates a mean
value across VSL estimates derived from 26 labor market and contingent valuation studies
published between 1974 and 1991. The mean VSL across these studies is $4.8 million (1990$).
We then adjust this VSL to account for the currency year and to account for income growth from
1990 to the analysis year. Specifically, the VSL applied in this analysis in 2017$ after adjusting
for income growth is $11 million for 2045.
The Agency is committed to using scientifically sound, appropriately reviewed evidence in
valuing changes in the risk of premature death and continues to engage with the SAB to update
its mortality risk valuation estimates. In 2016, the Agency proposed new meta-analytic
approaches for updating its estimates, which were subsequently reviewed by the SAB-EEAC.429
EPA is taking the SAB's formal recommendations under advisement.
In valuing PIVfo.s-related premature mortality, we discount the value of premature mortality
occurring in future years using rates of 3 percent and 7 percent.430 We assume that there is a
multi-year "cessation" lag between changes in PM exposures and the total realization of changes
in health effects. Although the structure of the lag is uncertain, EPA follows the advice of the
SAB-Health Effects Subcommittee (HES) to use a segmented lag structure that assumes 30
percent of premature deaths are reduced in the first year, 50 percent over years 2 to 5, and 20
percent over the years 6 to 20 after the reduction in PM2.5.431 Changes in the cessation lag
assumptions do not change the total number of estimated deaths but rather the timing of those
deaths.
Because short-term ozone-related premature mortality occurs within the analysis year, the
estimated ozone-related benefits are identical for all discount rates. When valuing changes in
ozone-attributable deaths using the Turner et al. (2016) study, we follow advice provided by the
SAB-HES, which found that ".. .there is no evidence in the literature to support a different
cessation lag between ozone and particulate matter. The HES therefore recommends using the
same cessation lag structure and assumptions as for particulate matter when utilizing cohort
mortality evidence for ozone".432
These estimated health benefits do not account for the influence of future changes in the
climate on ambient concentrations of pollutants.433 For example, recent research suggests that
future changes to climate may create conditions more conducive to forming ozone; the influence
of changes in the climate on PM2.5 concentrations are less clear.434 The estimated health benefits
also do not consider the potential for climate-induced changes in temperature to modify the
relationship between ozone and the risk of premature death.435'436'437
388
-------
8.5 Characterizing Uncertainty in the Estimated Benefits
This analysis includes many data sources as inputs that are each subject to uncertainty. Input
parameters include projected emission inventories, air quality data from models (with their
associated parameters and inputs), population data, population estimates, health effect estimates
from epidemiology studies, economic data, and assumptions regarding the future state of the
world (i.e., regulations, technology, and human behavior). When compounded, even small
uncertainties can greatly influence the size of the total quantified benefits.
Our estimate of the total monetized PM2.5 and ozone-attributable benefits is based on EPA's
interpretation of the best available scientific literature and methods and supported by the SAB-
HES and the National Academies of Science.438 Below are key assumptions underlying the
estimates for PM2.5-related premature mortality, followed by key uncertainties associated with
estimating the number and value of ozone-related premature deaths. Chapter 6 of the Estimating
PM2.5- and Ozone-Attributable Health Benefits TSD presents a thorough quantitative and
qualitative analysis of the sources of uncertainty present in the health benefits analysis.
We assume that all fine particles, regardless of their chemical composition, are equally potent
in causing premature mortality. Support for this assumption comes from the 2019 PM ISA,
which concluded that "many PM2.5 components and sources are associated with many health
effects and that the evidence does not indicate that any one source or component is consistently
more strongly related with health effects than PM2.5 mass".439
As noted above, we assume that the health impact function for fine particles is log-linear
without a threshold. Thus, the estimates include health benefits from reducing fine particles in
areas with different concentrations of PM2.5, including both areas with projected annual mean
concentrations that are above the level of the fine particle standard and areas with projected
concentrations below the level of the standard.
Also, as noted above, we assume that there is a "cessation" lag between the change in PM
exposures and the total realization of changes in mortality effects. Specifically, we assume that
some of the incidences of premature mortality related to PM2.5 exposures occur in a distributed
fashion over the 20 years following exposure based on the advice of the SAB-HES, which affects
the valuation of mortality benefits at different discount rates. The above assumptions are subject
, . • , 440
to uncertainty.
In general, we are more confident in the magnitude of the risks we estimate from simulated
PM2.5 concentrations that coincide with the bulk of the observed PM concentrations in the
epidemiological studies that are used to estimate the benefits. Likewise, we are less confident in
the risk we estimate from simulated PM2.5 concentrations that fall below the bulk of the observed
data in these studies. There are uncertainties inherent in identifying any particular point at which
our confidence in reported associations decreases appreciably, and the scientific evidence
provides no clear dividing line. This relationship between the air quality data and our confidence
in the estimated risk is represented below in Figure 8-1.
389
-------
Less confident
More confident
<
Below LRL of PM2.5 data in
epidemiology study
(value extrapolated)
1 standard deviation below the
mean PM2 5 observed in
epidemiology study
Mean of PM2.5 data in
epidemiology study
Figure 8-1: Stylized Relationship between the PM2.5 Concentrations Considered in Epidemiology Studies and
our Confidence in the Estimated PM-related Premature Deaths
In this analysis, we plot estimated PM-related deaths according to where they occur along the
distribution of baseline PM2.5 annual mean concentrations (Figure 8-2). Displaying the data in
such a way allows readers to visualize the portion of population exposed to annual mean PM2.5
levels at or above different concentrations, which provides some insight into the level of
uncertainty in the estimated PM2.5 mortality benefits. EPA does not view the level of the PM
NAAQS or the lowest concentration levels reported in the mortality studies as concentration
thresholds below which we would not quantify health benefits of air quality
improvements.^ ^ ^ Rather, the PIvfc.s-attributable benefits estimates reported in this draft RIA
are the most appropriate estimates because they reflect the full range of air quality concentrations
associated with the emission reduction program being evaluated. The 2019 PM ISA concluded
that the scientific evidence collectively is sufficient to conclude that there is a causal relationship
between long-term PM2.5 exposures and mortality and that overall the studies support the use of a
no-threshold log-linear model to estimate mortality attributed to long-term PM2.5 exposure.
Figure 8-2 compares the percentage of the population and PM-related deaths to the annual
mean PM2.5 concentrations in the baseline for the year 2045. The figure identifies the LRL for
each of the major cohort studies and the annual mean PM2.5 NAAQS of 12 |ig/m3. For Turner et
al. (2016), the LRL is 2.8 |ig/m3 and for Di et al. (2017), the LRL is 0.02 |ig/m3.ZZZZZZ As
PM-related mortality quantified using risk estimates from the Di et al. (2017) and Turner et al.
(2016) are within 5 percent of one another, in the interest of clarity and simplicity, we present the
results estimated using the risk estimate from Turner et al. (2016) alone in Figure 8-2. Additional
information on low concentration exposures in Turner et al. (2016) and Di et al. (2017) can be
found in section 6.1.2.1 of the Estimating PM2.5- and Ozone-Attributable Health Benefits TSD.
The air quality modeling predicts PM2.5 concentrations to be at or below the level of the annual
PM2.5 NAAQS (12 |ig/m3) in most locations in 2045. As noted in draft RIA Chapter 6.3.2, we are
more confident in the projected changes in annual mean PM2.5 concentrations than we are in the
projected absolute PM2.5 concentrations in 2045.
yyyyyy por a summary 0f the scientific review statements regarding the lack of a threshold in the PM2 5-mortality
relationship, see the TSD entitled Summary of Expert Opinions on the Existence of a Threshold in the
Concentration-Response Function for Phh.s-related Mortality (U.S. EPA, 2010).
zzzzzz Turner et al. (2016) estimated PM2 5 exposures using both a hierarchical Bayesian space-time model (HBM)
and a land use regression model with Bayesian Maximum Entropy kriging of residuals (LURBME). As such, two
LRLs are reported in the paper, 2.8 |ig/m3 and 1.4 |ig/m\ As the HBM risk estimate was used in the final 2021
CSAPR RIA, the HBM LRL is presented here.
390
-------
LRl
D» (2017)
LRL
Turner (2016)
2012
PM;} Annual NAAQS
0.2
Attributable Deaths
0
2 3 4 5 6 7 8 9 10 11 12 13 14 15
2045 Baseline PM2 5 Concentration (ng/m5)
Figure 8-2: Estimated Percentage of PlVh.s-Related Deaths (Turner et al. 2016) and Number of Individuals
Exposed (30+) by Annual Mean PM2.5 Level in 2045
The estimated number and value of avoided ozone-attributable deaths are also subject to
uncertainty. When estimating the economic value of avoided premature mortality from long-term
exposure to ozone, we use a 20-year segment lag (as used for PM2.5) as there is no alternative
empirical estimate of the cessation lag for long-term exposure to ozone. The 20-year segmented
lag accounts for the onset of cardiovascular-related mortality, an outcome which is not relevant
to the long-term respiratory mortality estimated here. We use a log-linear impact function
without a threshold in modeling short-term ozone-related mortality. However, we acknowledge
reduced confidence in specifying the shape of the concentration-response relationship in the
range of < 40ppb and below.441 Thus, the estimates include health benefits from reducing ozone
in areas with varied concentrations of ozone down to the lowest modeled concentrations.
8.6 Estimated Number and Economic Value of Health Benefits
Below we report the estimated number of reduced premature deaths and illnesses in 2045
from proposed Option 1 relative to the baseline along with the 95 percent confidence interval
(Table 8-3 and Table 8-4). The number of reduced estimated deaths and illnesses are calculated
from the sum of individual reduced mortality and illness risk across the population. Table 8-5
reports the estimated individual economic value of avoided premature deaths and illnesses
relative to the baseline along with the 95 percent confidence interval.
391
-------
Table 8-6 reports total benefits associated with proposed Option 1 in 2045, reflecting
alternative combinations of the economic value of PM2.5- and ozone-related premature deaths
summed with the economic value of illnesses for each discount rate.AAAAAAA
aaaaaaa continues to refine its approach for estimating and reporting PM-related effects at lower
concentrations. The Agency acknowledges the additional uncertainty associated with effects estimated at these
lower levels and seeks to develop quantitative approaches for reflecting this uncertainty in the estimated PM
benefits.
392
-------
Table 8-3: Estimated Avoided PM2.5 Mortality and Illnesses in 2045 for Proposed Option 1 (95% Confidence
Interval) a'b
Proposed Option 1
Avoided premature mortality
Turner et al. (2016) - Ages 30+
740
(500 to 980)
Di et al. (2017) - Ages 65+
800
(780 to 830)
Woodruff et al. (2008) - Ages < 1
4.1
(-2.6 to 11)
Non-fatal heart attacks among adults
Short-term exposure
Peters et al. (2001)
790
(180 to 1,400)
Pooled estimate
85
(31 to 230)
Morbidity effects
Long-term exposure
Asthma onset
1,600
(1,500 to 1,600)
Allergic rhinitis symptoms
10,000
(2,500 to 18,000)
Stroke
41
(11 to 70)
Lung cancer
52
(16 to 86)
Hospital Admissions - Alzheimer's disease
400
(300 to 500)
Hospital Admissions - Parkinson's disease
43
(22 to 63)
Short-term exposure
Hospital admissions-cardiovascular
110
(76 to 130)
ED visits- cardiovascular
210
(-82 to 500)
Hospital admissions - respiratory
68
(23 to 110)
ED visits - respiratory
400
(78 to 830)
Asthma symptoms
210,000
(-100,000 to 520,000)
Minor restricted-activity days
460,000
(370,000 to 550,000)
Cardiac arrest
10
(-4.2 to 24)
Lost work days
78,000
(66,000 to 90,000)
a Values rounded to two significant figures.
b PM2.5 exposure metrics are not presented here because all PM health endpoints are based on studies that used
daily 24-hour average concentrations. Annual exposures are estimated using daily 24-hour average concentrations.
393
-------
Table 8-4: Estimated Avoided Ozone Mortality and Illnesses in 2045 for the Proposed Option 1 (95%
Confidence Interval)3
Metric and Seasonb
Proposed Option 1
Avoided premature mortality
Long-term
exposure
Turner et al. (2016)
MDA8
April-September
2,100
(1,400 to 2,700)
Short-term
exposure
Katsouyanni et al (2009)
MDA1
April-September
120
(-69 to 300)
Morbidity effects
Long-term
exposure
Asthma onset0
MDA8
June-August
16,000
(14,000 to 18,000)
Short-term
exposure
Allergic rhinitis symptoms
MDA8
May-September
88,000
(47,000 to 130,000)
Hospital admissions - respiratory
MDA1
April-September
350
(-91 to 770)
ED visits - respiratory
MDA8
May-September
5,100
(1,400 to 11,000)
Asthma symptoms - Coughd
MDA8
May-September
920,000
(-50,000 to 1,800,000)
Asthma symptoms - Chest Tightnessd
MDA8
May-September
770,000
(85,000 to 1,400,000)
Asthma symptoms - Shortness of Breathd
MDA8
May-September
390,000
(-330,000 to 1,100,000)
Asthma symptoms - Wheezed
MDA8
May-September
730,000
(-57,000 to 1,500,000)
Minor restricted-activity daysd
MDA1
May-September
1,600,000
(650,000 to 2,600,000)
School absence days
MDA8
May-September
1,100,000
(-150,000 to 2,200,000)
a Values rounded to two significant figures.
b MDA8 - maximum daily 8-hour average; MDA1 - maximum daily 1-hour average. Studies of ozone vary with
regards to season, limiting analyses to various definitions of summer (e.g., April-September, May-September or
June-August). These differences can reflect state-specific ozone seasons, EPA-defined seasons or another seasonal
definition chosen by the study author. The paucity of ozone monitoring data in winter months complicates the
development of full year projected ozone surfaces and limits our analysis to only warm seasons.
0 The underlying metric associated with this risk estimate is daily 8-hour average from 10am - 6pm (AVG8);
however, we ran the study with a risk estimate converted to MDA8.
d Applied risk estimate derived from full year exposures to estimates of ozone across a May-September ozone
season. When risk estimates based on full-year, long-term ozone exposures are applied to warm season air quality
projections, the resulting benefits assessment may underestimate impacts, due to a shorter timespan for impacts to
accrue.
394
-------
Table 8-5: Estimated Economic Value of PM2.5- and Ozone-Attributable Premature Mortality and Illnesses in
2045 for Proposed Option 1 (95% Confidence Interval; millions of 2017$)a
3% Discount Rate
7% Discount Rate
Avoided premature mortality
1/-1
-------
Table 8-6: Total Ozone and PIVh.s-Attributable Benefits in 2045 for Proposed Option 1 (95% Confidence
Interval; billions of 2017$)a b
Total Annual Benefits in 2045
3% Discount Rate
$12 $33
($0.72 to $31)° ($3.5 to $87)d
7% Discount Rate
$10 , $30
($0.37 to $28)c ($3.0 to $78)d
a The benefits associated with the standards presented here do not include the full complement of health,
environmental, and climate-related benefits that, if quantified and monetized, would increase the total monetized
benefits.
b Values rounded to two significant figures. The two benefits estimates separated by the word "and" signify that they
are two separate estimates. The estimates do not represent lower- and upper-bound estimates though they do reflect
a grouping of estimates that yield more and less conservative benefit totals. They should not be summed.
0 Sum of benefits using the Katsouyanni et al. (2009) short-term exposure ozone respiratory mortality risk estimate
and the Turner et al. (2016) long-term exposure PM2 5 all-cause risk estimate.
d Sum of benefits using the Turner et al. (2016) long-term exposure ozone respiratory mortality risk estimate and the
Di et al. (2017) long-term exposure PM2 5 all-cause risk estimate.
8.7 Present Value of Total Benefits of Proposed Option 1 and 2
The full-scale benefits analysis for proposed Option 1 presented earlier in this Chapter reflects
spatially and temporally allocated emissions inventories generated using SMOKE/MOVES (see
draft RIA Chapter 5), photochemical air quality modeling using CMAQ (see draft RIA Chapter
6), and PM2.5 and ozone benefits generated using BenMAP-CE, all for conditions projected to
occur in calendar year 2045. As we presented in draft RIA Chapter 5 and Chapter 7, national
estimates of year-over-year emissions and program costs were generated for both Option 1 and
Option 2 from proposed implementation to a year when the program would be fully phased-in
and the vehicle fleet would be approaching full turnover (2027-2045). The time and resources
required to conduct air quality modeling to support a full-scale benefits analysis for Option 2 in
2045 and for all Option 1 and Option 2 analysis years from 2027 to 2044 precluded the Agency
from conducting benefits analyses comparable to the calendar year 2045 Option 1 benefits
analysis. Instead, we have used a reduced-form approach to scale total Option 1 benefits in 2045
back to 2027 (including interim years) using projected reductions in year-over-year NOx
emissions so that we can estimate the present value of the stream of estimated benefits for Option
1. We have also used year-over-year Option 2 NOx emissions reductions to scale the total
benefits associated with Option 1 to derive a best estimate of criteria pollutant benefits associated
with Option 2.
This approach is similar to the Agency's method for estimating "benefits-per-ton" values over
time.442 For interim analysis years without air quality modeling, we input the proposed
program's 2045 air quality data into BenMAP-CE to generate benefits that occur in earlier
analysis years. This approach allows us to calculate the benefits for interim years by adjusting for
changes in population, baseline mortality incidence, and income growth over time. Table 8-7
396
-------
displays the data used to generate benefits that reflect input data for years 2027, 2030, 2035,
2040, and 2045 BBBBBBB
Table 8-7: Benefits Inputs that Change Over Time used to Calculate Year-over-Year Estimates
Analysis Year
Air Quality
Modeling &
Emissions Year
Population Year
Baseline
Mortality
Incidence Year
Income Growth
Year
Currency Year
2027
2045
2027
2025
2027
2017
2030
2030
2030
2030
2035
2035
2035
2035
2040
2040
2040
2040
2045
2045
2045
2045
We next calculate the total monetized benefits estimated for each of the analysis years and
divide them by the estimated tons of NOx emissions projected to be controlled by the proposed
Option 1 in 2045 (see draft RIA Chapter 5, Table 5-21) to generate "benefit-per-ton" values that
reflect benefits inputs consistent with the analysis year.ccccccc Because NOx is the dominating
pollutant controlled by the proposed program, we make a simplifying assumption that total PM
and ozone benefits can be scaled by NOx emissions, even though emissions of other pollutants
are controlled in smaller amounts by the proposed Options (see draft RIA Chapter 5, Table 5-
21). By using the 2045 air quality modeling surfaces for the earlier analysis years, we also
assume that the spatial distribution of NOx emissions reductions does not change over time.
While there may be localized differences in the rate of fleet turnover due to state or local
incentive programs, we do not currently have sufficient data to incorporate those differences into
our analyses and believe that they would generally even out over time (as noted in draft RIA
Chapter 5, we use MOVES default vehicle activity data, including data on age of the fleet or
turnover).
To estimate total benefits for the interim years, we multiply the benefit-per-ton values
estimated for each earlier analysis year by the NOx emissions projected to be controlled in that
same year (2027, 2030, 2035, and 2040; see draft RIA Chapter 5, Table 5-33). For intervening
years between the analysis years, we linearly interpolate total benefits.
Table 8-8 and Table 8-9 present the undiscounted stream of scaled annual total benefits of the
proposed Option 1 between 2027 and 2045. We also estimate the present value and annualized
value of the stream of benefits in these tables. Table 8-8 presents total benefits as the sum of
short-term ozone respiratory mortality benefits for all ages, long-term PM2.5 all-cause mortality
benefits for ages 30 and above, and all monetized avoided illnesses.443'444 Table 8-9 presents
total benefits as the sum of long-term ozone respiratory mortality benefits for ages 30 and above,
long-term PM2.5 all-cause mortality benefits for ages 65 and above, and all monetized avoided
illnesses.445'446 The present value of benefits in both tables is discounted back to year 2027 using
both a 3 percent and 7 percent discount rate.
bbbbbbb [n(erjm analysis years chosen for computational efficiency at reasonable intervals.
ccccccc no1c that these "benefit-per-ton" values are internally consistent with the air quality modeling conducted for
Option 1 in 2045. They are appropriate for scaling benefits of the proposed program, but should not be used outside
of the context of this rulemaking analysis.
397
-------
To generate the undiscounted stream of total benefits associated with the proposed Option 2,
and to estimate the present value of those benefits, we also employed a reduced-form scaling
approach.DDDDDDD Beginning with the stream of total benefits associated with the proposed
Option 1, we scaled annual benefits by the ratio of year-over-year NOx emission reductions
projected to occur from proposed Option 2 compared to year-over-year NOx emission reductions
associated with proposed Option 1 (see draft RIA Chapter 5, Table 5-33). The stream of total
benefits for proposed Option 2, along with associated present and annualized values, are also
presented in Table 8-8 and Table 8-9.
Table 8-8: Undiscounted Stream and Present Value of Human Health Benefits from 2027 through 2045:
Monetized Benefits Quantified as Sum of Short-Term Ozone Respiratory Mortality Ages 0-99, and Long-
Term PM2.5 All-Cause Mortality Ages 30+ (Discounted at 3% and 7%; billions of 2017$)a b
Proposed Option 1
Proposed Option 2
3%
7%
3%
7%
2027
$0.57
$0.51
$0.52
$0.47
2028
$1.2
$1.1
$1.1
$0.98
2029
$1.8
$1.7
$1.7
$1.5
2030
$2.5
$2.3
$2.3
$2.1
2031
$3.4
$3.1
$3.1
$2.7
2032
$4.3
$3.9
$3.8
$3.4
2033
$5.0
$4.5
$4.3
$3.9
2034
$5.6
$5.0
$4.9
$4.4
2035
$6.3
$5.7
$5.4
$4.8
2036
$6.9
$6.2
$5.8
$5.3
2037
$7.8
$7.0
$6.3
$5.7
2038
$8.6
$7.7
$6.7
$6.0
2039
$9.1
$8.2
$7.1
$6.4
2040
$9.6
$8.6
$7.5
$6.7
2041
$10
$9.0
$7.8
$7.1
2042
$10
$9.4
$8.2
$7.4
2043
$11
$9.8
$8.5
$7.6
2044
$11
$10
$8.8
$7.9
2045°
$12
$10
$9.1
$8.2
Present Value
$87
$50
$71
$41
Annualized Value
$6.1
$4.9
$5.0
$4.0
a The benefits associated with the standards presented here do not include the full complement of health,
environmental, and climate-related benefits that, if quantified and monetized, would increase the total monetized
benefits.
b Benefits calculated as value of avoided: PM2 5-attributable deaths (quantified using a concentration-response
relationship from the Turner et al. 2016 study); Ozone-attributable deaths (quantified using a concentration-
response relationship from the Katsouyanni et al. 2009 study); and PM2 5 and ozone-related morbidity effects.
0 Year in which PM2 5 and ozone air quality associated with Option 1 was simulated (2045).
ddddddd \ye arc n0( including an analysis of benefits of the Alternative (described in Preamble Sections III and IV)
because we currently do not have sufficient information to conclude that the Alternative standards would be feasible
in the MY2027 timeframe. Preamble Section III presents our current feasibility analysis for the Alternative.
398
-------
Table 8-9: Undiscounted Stream and Present Value of Human Health Benefits from 2027 through 2045:
Monetized Benefits Quantified as Sum of Long-Term Ozone Respiratory Mortality Ages 30+, and Long-Term
PM2.5 All-Cause Mortality Ages 65+ (Discounted at 3% and 7%; billions of 2017$)a b
Proposed Option 1
Proposed Option 2
3%
7%
3%
7%
2027
$1.6
$1.4
$1.4
$1.3
2028
$3.3
$2.9
$3.0
$2.7
2029
$5.1
$4.6
$4.7
$4.2
2030
$7.0
$6.3
$6.4
$5.8
2031
$9.6
$8.6
$8.5
$7.6
2032
$12
$11
$11
$9.5
2033
$14
$13
$12
$11
2034
$16
$14
$14
$12
2035
$18
$16
$15
$14
2036
$20
$18
$17
$15
2037
$22
$20
$18
$16
2038
$24
$22
$19
$17
2039
$26
$23
$20
$18
2040
$28
$25
$21
$19
2041
$29
$26
$23
$20
2042
$30
$27
$24
$21
2043
$31
$28
$24
$22
2044
$32
$29
$25
$23
2045°
$33
$30
$26
$23
Present Value
$250
$140
$200
$120
Annualized Value
$17
$14
$14
$11
a The benefits associated with the standards presented here do not include the full complement of health,
environmental, and climate-related benefits that, if quantified and monetized, would increase the total monetized
benefits.
b Benefits calculated as value of avoided: PM2 5-attributable deaths (quantified using a concentration-response
relationship from the Di et al. 2017 study); Ozone-attributable deaths (quantified using a concentration-response
relationship from the Turner et al. 2016 study); and PM2 5 and ozone-related morbidity effects.
0 Year in which PM2 5 and ozone air quality for Option 1 was simulated (2045).
8.8 Unquantified Benefits
In addition to the PM2.5 and ozone-related health impacts we are unable to quantify or
monetize in Table 8-2, there are additional benefits associated with reductions in exposure to
ambient concentrations of NO2, ecosystem benefits, and visibility improvement that EPA is not
currently able to quantify due to data, resource, or methodological limitations. EPA continues to
pursue data and methods to further improve our assessment of benefits that are currently
unquantified. In particular, we are evaluating the feasibility of assessing impacts on ecosystem
services from reductions in nitrogen deposition and terrestrial acidification. Draft RIA Chapter 4
399
-------
provides a qualitative description of both the health and environmental effects of the criteria
pollutants controlled by the proposed program. These additional unquantified health and welfare
benefit categories are listed in Table 8-10.
There would also be benefits associated with reductions in air toxic pollutant emissions that
result from the proposed program (See draft RIA Chapter 4.1.6 and draft RIA Chapter 5.3.1), but
we did not attempt to monetize those impacts. This is 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 estimation or benefits assessment. While EPA has worked to improve
these tools, there remain critical limitations for estimating incidence and assessing benefits of
reducing mobile source air toxics.
The proposed criteria pollutant standards would also reduce methane (CH4) emissions
due to lower total hydrocarbon emission rates from the tailpipe of heavy-duty gasoline vehicles
(see draft RIA Chapter 5.2.2 for more detail). There would be climate-related benefits associated
with these projected reductions in CH4, but we did not monetize them.EEEEEEE We request
comment on how to address the climate benefits and other categories of non-monetized benefits
of the proposed rule.
17171717171717
The U.S. District Court for the Western District of Louisiana has issued an injunction concerning the
monetization of the benefits of greenhouse gas emission reductions by EPA and other defendants. See Louisiana v.
Biden, No. 21-cv-01074-JDC-KK (W.D. La. Feb. 11, 2022).
400
-------
Table 8-10: Unquantified Criteria Pollutant Health and Welfare Benefits Categories
Category
Effect
Effect
Quantified
Effect
Monetized
Mult'
Information
Improved Human Health
Reduced incidence of
morbidity from
exposure to N02
Asthma hospital admissions
—
—
N02 ISA447'1
Chronic lung disease hospital admissions
—
—
N02 ISA1
Respiratory emergency department visits
—
—
N02 ISA1
Asthma exacerbation
—
—
N02 ISA1
Acute respiratory symptoms
—
—
N02 ISA1
Premature mortality
—
—
N02 ISA1'2'3
Other respiratory effects (e.g., airway
hyperresponsiveness and inflammation, lung
function, other ages and populations)
—
—
N02 ISA2'3
Improved Environment
Reduced visibility
impairment
Visibility in Class 1 areas
—
—
PM ISA1
Visibility in residential areas
—
—
PM ISA1
Reduced effects on
materials
Household soiling
—
—
PMISA1-2
Materials damage (e.g., corrosion, increased wear)
—
—
PM ISA2
Reduced effects from
PM deposition
(metals and organics)
Effects on individual organisms and ecosystems
—
—
PMISA2
Reduced vegetation
and ecosystem
effects from
exposure to ozone
Visible foliar injury on vegetation
—
—
Ozone ISA1
Reduced vegetation growth and reproduction
—
—
Ozone ISA1
Yield and quality of commercial forest products and
crops
—
—
Ozone ISA1
Damage to urban ornamental plants
—
—
Ozone ISA2
Carbon sequestration in terrestrial ecosystems
—
—
Ozone ISA1
Recreational demand associated with forest aesthetics
—
—
Ozone ISA2
Other non-use effects
Ozone ISA2
Ecosystem functions (e.g., water cycling,
biogeochemical cycles, net primary productivity, leaf-
gas exchange, community composition)
—
—
Ozone ISA2
Reduced effects from
acid deposition
Recreational fishing
—
—
NOx SOx
ISA448-1
Tree mortality and decline
—
—
NOx SOx ISA2
Commercial fishing and forestry effects
—
—
NOx SOx ISA2
Recreational demand in terrestrial and aquatic
ecosystems
—
—
NOx SOx ISA2
Other non-use effects
NOx SOx ISA2
Ecosystem functions (e.g., biogeochemical cycles)
—
—
NOx SOx ISA2
Reduced effects from
nutrient enrichment
Species composition and biodiversity in terrestrial
and estuarine ecosystems
—
—
NOx SOx ISA2
Coastal eutrophication
—
—
NOx SOx ISA2
Recreational demand in terrestrial and estuarine
ecosystems
—
—
NOx SOx ISA2
Other non-use effects
NOx SOx ISA2
Ecosystem functions (e.g., biogeochemical cycles,
fire regulation)
—
—
NOx SOx ISA2
Reduced vegetation
effects from ambient
exposure to S02 and
NOx
Injury to vegetation from S02 exposure
—
—
NOx SOx ISA2
Injury to vegetation from NOx exposure
—
NOx SOx ISA2
1 We assess these benefits qualitatively due to data and resource limitations for this draft RIA.
2 We assess these benefits qualitatively because we do not have sufficient confidence in available data or methods.
3 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.
401
-------
Chapter 9 Comparison of Benefits and Costs
This chapter compares the estimated range of total monetized health benefits to total costs
associated with the criteria pollutant program proposed Options 1 and 2. This chapter also
presents the range of monetized net benefits (benefits minus costs) associated with the same
options. Criteria pollutant program costs are detailed and presented in Chapter 7 of this draft
RIA. Those costs include costs for both the new technology and the operating costs associated
with that new technology, as well as costs associated with the proposed warranty and useful life
provisions for Options 1 and 2. Criteria pollutant program benefits are detailed and presented in
draft RIA Chapter 8.FFFFFFF Those benefits are the monetized economic value of the reduction in
PM2.5-and ozone-related premature deaths and illnesses that result from reductions in NOx
emissions and directly emitted PM2.5 attributable to implementation of the proposed options.
9.1 Methods
EPA presents three different benefit-cost comparisons for the proposed Options 1 and
2-ggggggg
1. A future-year snapshot comparison of annual benefits and costs in the year 2045, chosen
to approximate the annual health benefits that would occur in a year when the program
would be fully implemented and when most of the regulated fleet would have turned
over. Benefits, costs and net benefits are presented in year 2017 dollars and are not
discounted. However, 3 percent and 7 percent discount rates were applied in the valuation
of avoided premature deaths from long-term pollution exposure, to account for a twenty-
year segmented cessation lag.
2. The present value (PV) of the stream of benefits, costs and net benefits calculated for the
years 2027-2045, discounted back to the first year of implementation of the proposed rule
(2027) using both a 3 percent and 7 percent discount rate, and presented in year 2017
dollars. Note that year-over-year costs are presented in draft RIA Chapter 7 and year-
over-year benefits can be found in draft RIA Chapter 8.
3. The equivalent annualized value (EAV) of benefits, costs and net benefits, representing a
flow of constant annual values that, had they occurred in each year from 2027 to 2045,
would yield an equivalent present value to those estimated in Method 2 above (using
either a 3 percent or 7 percent discount rate). Each EAV represents a typical benefit, cost
or net benefit for each year of the analysis and is presented in year 2017 dollars.
fffffff nolcc[ jn draft RIA Chapter 5.4, there are differences between the standards, emission warranty, and useful
life provisions of proposed Option 1 presented in Sections III and IV and those included in our control case scenario
modeled for the air quality analysis (as noted in Section VII, due to resource constraints we only conducted air
quality modeling for the proposed Option 1). As detailed in draft RIA Chapter 8, estimates of health benefits are
based on our air quality analysis, and thus differences between proposed Option 1 and modeling are not reflected in
the benefits analysis.
ggggggg We are not including an analysis of costs or benefits of the Alternative (described in Preamble Sections III
and IV) because we currently do not have sufficient information to conclude that the Alternative standards would be
feasible in the MY2027 timeframe. Preamble Section III presents our feasibility analysis for the Alternative.
402
-------
The two estimates of benefits (and net benefits) in each of these benefit-cost comparisons
reflect alternative combinations of the economic value of PM2.5- and ozone-related premature
deaths summed with the economic value of illnesses for each discount rate (see Chapter 8 for
more detail).
9.2 Results
Table ES presents the benefits, costs and net benefits of the proposed Options 1 and 2 in
annual terms for year 2045, in PV terms, and in EAV terms.
Annual benefits of proposed Option 1 are larger than the annual costs in 2045, with annual net
benefits of $8.1 and $28 billion using a 7 percent discount rate, and $9.2 and $31 billion using a
3 percent discount rate.HHHHHHH Benefits also outweigh the costs when expressed in PV terms
(net benefits of $33 and $130 billion using a 7 percent discount rate, and $61 and $220 billion
using a 3 percent discount rate) and EAV terms (net benefits of $2.9 and $12 billion using a 7
percent discount rate, and $4.1 and $15 billion using a 3 percent discount rate).
The benefits also outweigh the costs in annual 2045 terms when looking at proposed Option
2, with annual net benefits of $5.3 and $21 billion using a 7 percent discount rate and $6.2 billion
and $23 billion using a 3 percent discount rate. Benefits also outweigh the costs in PV and EAV
terms for proposed Option 2.
Comparing the proposed Options 1 and 2, our analysis shows that proposed Option 2 has
lower net benefits than proposed Option 1 due to both higher costs and lower emission
reductions relative to proposed Option 1. As discussed further in Preamble Section I.G, we have
considered several other factors, including lead time and technological feasibility, in developing
this proposal and considering possible regulatory alternatives.
Given these results, EPA expects that implementation of the proposed rule would provide
society with a substantial net gain in welfare, notwithstanding the health and other benefits we
were unable to quantify (see draft RIA Chapter 8.8 for more information about unquantified
benefits). EPA does not expect the omission of unquantified benefits to impact the Agency's
evaluation of regulatory options since unquantified benefits generally scale with the emissions
impacts of the proposed Options.
hhhhhhh j]lc range of benefits and net benefits presented in this section reflect a combination of assumed PM2 5 and
ozone mortality risk estimates and selected discount rate.
403
-------
Table 9-1: 2045 Annual Value, Present Value and Equivalent Annualized Value of Costs, Benefits and Net
Benefits of the Proposed Options 1 and 2 (billions, 2017$)a b
Proposed Option 1
Proposed Option 2
3%
7%
3%
7%
Discount
Discount
Discount
Discount
Benefits
$12-$33
$10-$30
$9.1 -$26
$8.2 - $23
2045
Costs
$2.3
$2.3
$2.9
$2.9
Net Benefits
$9.2-$31
$8.1 -$28
$6.2 - $23
$5.3 -$21
Benefits
$88 - $250
$52-$150
$71 -$200
$41-$120
Present Value
Costs
$27
$19
$30
$21
Net Benefits
$61 - $220
$33 -$130
$41-$170
$21 - $96
Equivalent
Annualized Value
Benefits
$6.0-$17
$4.7 -$13
$5.0-$14
$4.0 -$11
Costs
$1.9
$1.9
$2.1
$2.0
Net Benefits
$4.1 -$15
$2.9 -$12
$2.9-$12
$2.0 - $9.3
a All benefits estimates are rounded to two significant figures; numbers may not sum due to independent
rounding. The range of benefits (and net benefits) in this table are two separate estimates and do not
represent lower- and upper-bound estimates, though they do reflect a grouping of estimates that yield more
and less conservative benefits totals. The costs and benefits in 2045 are presented in annual terms and are
not discounted. However, all benefits in the table reflect a 3 percent and 7 percent discount rate used to
account for cessation lag in the valuation of avoided premature deaths associated with long-term exposure.
b The benefits associated with the standards presented here do not include the full complement of health,
environmental, and climate-related benefits that, if quantified and monetized, would increase the total
monetized benefits.
404
-------
Chapter 10 Economic Impact Analysis
This proposed rulemaking is considered economically significant, because it is expected to
have an annual impact on the economy of $100 million or more, and thus an economic analysis
has been completed as part of this draft RIA. This rule is not expected to have measurable
inflationary or recessionary effects.
The benefits to human health and the environment are discussed in Chapter 8, and the costs of
the proposed standards are discussed in Chapter 7. The benefit-cost analysis for this proposal is
presented in Chapter 9. This chapter provides an analysis of the impacts of the proposed
standards on vehicle sales and employment.
10.1 Impact on Sales, Fleet Turnover and Mode Shift
As explained in Chapter 7, this proposed rule is expected to increase the cost of heavy-duty
(HD) vehicles by requiring emissions control technologies capable of controlling NOx at lower
levels than are currently permitted, as well as longer emissions warranty periods for emissions
control technology components. In addition, there is an expected small increase in operating
costs mainly due to an increase in the use of diesel exhaust fluid (DEF).
Three sectors expected to be most immediately affected by this action are: 1) HD vehicle and
engine manufacturers, 2) HD vehicle and engine buyers, and 3) HD engine equipment suppliers
(e.g., suppliers of emissions control components). Effects on industries downstream of these
sectors, such as HD vehicle dealerships or delivery industries, would be relatively smaller due to
the limited role of the cost of a HD vehicle in pricing in those sectors. The three sectors will also
be responding to the 2016 rulemaking, "Greenhouse Gas Emissions and Fuel Efficiency
Standards for Medium- and Heavy-Duty Engines and Vehicles - Phase 2" (the Phase 2 rule). The
proposed standards, if finalized as proposed, would phase in during the same time frame as the
final Phase 2 rule standards. Both this proposed rulemaking, if finalized, and the Phase 2 rule
would require HD engine manufacturers to develop and implement improvements in engine
emissions controls.
As discussed in the Phase 2 rule RIA,449 increases in costs of HD vehicles from improved
emissions controls would be likely to lead to increases in final prices for HD vehicles; the
magnitude of that effect would depend on how much of the cost is passed along to potential
buyers. These price increases may affect HD vehicle sales in several ways. First, as basic
economic supply and demand theory suggests, higher prices are expected to reduce HD vehicle
sales. Second, HD vehicle buyers may strategically seek to avoid increased prices by "pre-
buying," increasing the purchases of new vehicles before the compliance deadline for the new
requirements. This might lead to an associated period immediately afterward of "low-buying,"
during which purchases decrease, and thereby impact the rate of fleet turnover. A third potential
effect is transportation mode shift, changing from on-highway trucking to other modes of
transportation (e.g., shipping via barge or rail instead of by truck). The magnitude of each of
405
-------
these three categories of effects (sales, fleet turnover, mode shift) would depend on the costs.
This section discusses these impacts.1111111
10.1.1 Sales
The effects of the proposed standards on HD vehicle sales depends on the magnitude of the
cost increase associated with implementing improved emissions controls to comply with the
proposed requirements, and on the degree to which the costs get passed through to vehicle
buyers.
As discussed in Chapter 7, an increase in cost of HD vehicles could result from the proposed
standards requiring the use of emissions control technologies capable of controlling NOx at
lower levels, as well as imposing longer useful life and emissions warranty periods for emissions
control technology components. While the proposed requirement for longer emissions warranty
periods would likely increase the purchase price of new HD vehicles, the corresponding
lengthened useful life periods are expected to make emissions control technology components
more durable. More durable components coupled with manufacturers paying for repairs during a
longer warranty period would in turn reduce repair costs, which may increase (or reduce the
decrease in) sales of new HD vehicles due to fleets and independent owner-operators being
inclined to purchase vehicles with lower repair costs.Nevertheless, the exact purchase
behavior of fleet owners and independent owner-operators is challenging to predict, particularly
in the time period immediately after new standards go into effect, when buyers may be waiting to
see how the new vehicles perform relative to manufacturer claims.
If cost increases are small, either purchasers or sellers may absorb the cost increase without
measurable changes in behavior. Significant cost increases passed through to buyers may lead
potential buyers to purchase fewer vehicles than without the higher costs, or to buy vehicles
sooner than they would have otherwise, in advance of the requirements. The National Academies
of Sciences, Engineering, and Medicine comment that both pre-buy and low-buy are likely to be
short-lived phenomena, and potentially unavoidable.450 Phasing in the standards would likely
reduce pre-buy, by phasing in the additional costs associated with the standards. In addition,
allowing early credits for compliance in advance of the standards would be expected to mitigate
pre-buy; instead of early compliance imposing only net costs, early compliance now provides a
benefit in terms of reduced compliance costs in the future.451 The timing of the proposed
standards, including the possibility of a phase-in, and early compliance credits are discussed in
Section IV of the Preamble.
Measuring the existence and magnitude of pre- and low-buy depends on separating those
effects from other factors that affect HD vehicle sales. If, for example, the timing of the
standards coincides with a decrease in HD vehicle sales due to an economic downturn, as likely
happened with standards that went into effect in 2007, the eve of the Great Recession, then the
estimated effects of the standard would somehow have to be disentangled from the effects of the
economic slowdown. Researchers estimating pre- or low-buy may seek to control for underlying
1111111 We recognize that additional external factors, including the current global COVID-19 pandemic, might impact
the heavy-duty vehicle market, however due to data limitations we are unable to include possible effects of such
external factors in our analyses. We request comment on this topic in Section X of this proposal.
jjjjjjj j]lc reduced repair costs may counteract some of the sales effect of increased vehicle purchase cost. As a result,
they may reduce incentives for pre- and low-buy and mitigate adverse sales impacts.
406
-------
sales patterns like this by including other factors, such as diesel price and gross domestic product
(GDP), that also influence new HD vehicle sales. They then look for deviations from these trends
at the time that the standards go into effect.
Using this approach to control for other influential factors, Lam and Bausell found a pre-buy
of around 18,000 to 21,000 HD trucks, about 20 to 25 percent of total production, when looking
at sales in the 6-month period before October 2002, a compliance deadline for HD engine
manufacturers to reduce NOx emissions.452 Similarly, Rittenhouse and Zaragoza-Watkins
(RZW) looked for pre-buy in the seven months preceding EPA's HD criteria pollutant standards'
compliance deadlines in 1998, 2002, 2007 and 2010, as well as for low-buy in the seven months
after the standards.453 For the 2007 standards, they found a sales increase of about 31,000
vehicles in those preceding months compared to the baseline, matched by an "approximately
symmetric" drop in sales in the following months. For 2002, they found a similarly symmetric,
though smaller, result of 14,000 - 18,000 Class 8 vehicles. These results suggest that the
standards induced earlier purchases of vehicles that would have otherwise been bought after the
standards were promulgated. The resulting effect was slower adoption of lower-emissions
vehicles, compared to the assumed rate of sales in the absence of new emissions standards,
although there was essentially no net change in sales (the sum of pre-buy and low-buy was not
statistically different from zero in 2007 or 2002). RZW did not find evidence of pre- or low-buy
for standards that went into effect in 1998 and 2010. They speculated that the standards in those
years were less costly and involved use of less risky, already available technologies.
A limitation of the method used by the researchers discussed above is that they do not suggest
a way to predict how future cost changes may influence sales. This is because they do not
include price impacts in the approach to estimate pre- and low-buy impacts, yet vehicle price is
expected to change when the standard goes into effect due to an increase in cost to the
manufacturers. Both the change in price and the timing of the standard would influence pre- and
low-buy because they occur at the same time, and it is statistically difficult to separate the two
effects. In addition, manufacturers of HD vehicles may affect either the magnitude or timing of
price increases in response to cost increases, confusing the effect of price on sales. Thus, while
these studies suggest that the current rule may lead to increases in sales through pre-buying
behavior, and decreases in sales through low-buying, the estimation approaches used by the
studies do not allow EPA to predict existence or magnitude of potential pre- and low-buy
impacts from future standards. In an effort to improve our analyses, EPA has been working on a
method to estimate these impacts. The approach and an example are explained in Draft RIA
Chapter 10.1.2, below.
An unpublished report attempted to develop a predictive model based on the impact of the
2007 standards. The authors assumed that a change in cost translates directly into a change in
price, which was then converted into a change in sales (Harrison and LeBel).454 The price change
was based on asking manufacturers to estimate the costs of meeting the standards. The study then
applied a price elasticity of demand of-1.9 (that is, a 1 percent increase in price will lead to a 1.9
percent decrease in sales) to estimate sales increases of 104,000 trucks during 2005-2006, and
sales reductions of 149,000 trucks over 2007-2008. (The study did not provide details on the
source of this elasticity estimate.) The study then reported "actual"KKKKKKK results of sales
kkkkkkk Qu0(a(j0n marks around "actual" are included in Harrison and LeBel (2008).
407
-------
increases of 120,000 vehicles in 2005-6 and decreases of 183,000 vehicles in 2007-8, based on
comparing estimated sales to an EPA estimate of baseline sales increased by a constant amount
each year. Unlike the published studies reviewed above, Harrison and LeBel did not control for
GDP, diesel prices, or other factors that might independently affect vehicle sales.454 As a result,
the EPA baseline used for the "actual" results is not likely to reflect actual sales in the absence of
the standards, and the "actual" pre- and low-buy values likely do not reflect changes due only to
the standards. For comparison, RZW's finding of pre-buy of about 31,000, based on controlling
for other factors, is about one-third of Harrison and LeBel's prediction and one-fourth of their
"actual" estimate.453
In sum, existing literature does not provide clear guidance on the relationship between a
change in vehicle cost due to a new standard and sales impacts. Neither Lam and Bausell nor
RZW links a change in vehicle cost to sales impacts (a major interest of the EPA); instead, both
papers focus on the magnitude of sales impacts in the periods surrounding compliance deadlines
of HD emission regulations.452-453 The method proposed by Harrison and LeBel links costs to
sales via a demand elasticity, but omits controls for shifts in baseline conditions as well as
omitting details on the source of the demand elasticity they used.454
For this NPRM, EPA acknowledges that these standards may lead to some pre-buy before the
standards go into effect, and some low-buy after the standards are effective. The estimated
increase in operating costs is not expected to have much effect on pre- or low-buy behavior
because the increase is small, and may be offset by lower expected repair costs due to longer
useful life and warranty periods. Based on the literature previously described, EPA is not able to
quantify these effects. In the following subsection we propose an approach to do so.
10.1.2 EPA's Research to Estimate Sales Effects
In 2020 EPA contractors conducted a review of available peer reviewed literature on the
effects of EPA's HD standards on HD sales (see Draft RIA Chapter 10.1.1 for literature review
results). The contractors then conducted an original analysis of the effects of previous EPA
standards on pre- and low-buy for HD vehicles.455
The analysis uses monthly vehicle sales data from the twelve-month period before and after
previous EPA standards went into effect (2002,LLLLLLL 2007, 2010, and 2014) to estimate pre-
and low-buy due to each standard. The analysis examined controls for the effects of month of
year, GDP, Brent Oil price, total imports and exports, and consumer sentiment and then used
binary indicator variables from 1 through 12 months pre- and post-regulation to identify
deviations from trends in sales specifically around those regulations. Unlike in previous studies,
all other variables (except for the binary variables of interest and the month of year) were
transformed into log-differences in order to address statistical issues associated with time series
data. Independent regressions were estimated for vehicle Classes 6 through 8, and for each of the
four previous HD regulations. Additional details of this analysis are available in the contractors'
report.
lllllll Because court rulings for the 2004 regulation pulled the compliance date for most HD truck manufacturers
forward to 2002, we will refer to that regulation as the 2002 standards, instead of the 2004 standards, in order to
keep the focus on compliance dates.
408
-------
Results show no statistically significant sales effects for Class 6 vehicles. There were a few
statistically significant results for Class 7, but the majority were of the opposite sign than
expected (that is, sales reductions before the standards and sales increases afterwards). For Class
8 vehicles, there were statistically significant results in the expected directions, with evidence of
short-lived pre-buy before the 2010 and 2014 standards and low-buy after the 2002, 2007, and
2010 standards. The rest of this section focuses only on Class 8 vehicles. For more discussion on
Classes 6 and 7, see Appendix 7.2 and Chapter 4.4.3 of the report.
The results provide estimates of the percent deviation in sales from trend for the combined
months leading up to and following the start of new emissions standards. For pre-buy,
statistically significant results range from no change persisting for the eleven months before the
2002 standards, to a 13.2 percent increase in the percent change in sales persisting for one month
before the 2014 standard. Statistically significant effects persist for up to eleven months before
the 2002 standards. For low-buy, statistically significant effects range from no change to a 14.9
percent decrease in the percent change in sales persisting for six months after the 2007 standards.
Statistically significant effects persist for up to twelve months following the 2007 standard.
Importantly, in addition to capturing the effects due to price changes associated with the
regulations, the coefficients also capture unobserved factors, such as concerns over vehicle
reliability and control technology uncertainty. Table 10-1 provides the results for the coefficients
on the pre- and low-buy indicators, along with their length of persistence and the regulation to
which the result is attributed. The significant coefficients are shown in bold face type.
409
-------
Table 10-1: Pre and Low-Buy Sales Effects Coefficients
2002
2007
2010
2014
Combined Months Pre-Regulation
12
0.024
0.004
0.009
0.000
11
-0.0**
-0.006
0.021
0.000
10
0.0**
-0.005
0.041
0.010
9
0.032
-0.008
0.032
0.032
8
0.041
-0.004
0.057**
0.013
7
0.044
-0.006
0.059**
0.019
6
0.037
-0.004
0.043
0.021
5
0.029
0.003
0.054*
0.019
4
0.004
-0.011
0.079***
0.030
3
0.047
-0.013
0.071**
0.014
2
-0.017
-0.012
0.105***
0.003
1
-0.032
-0.01
0.078***
0.132***
Combined Months Post-Regulation
1
0.065***
_0
-0.144***
-0.009
2
-0.051
-0.099***
-0.083*
-0.012
3
-0.115
-0.133***
-0.051
-0.015
4
-0.065
-0.143***
-0.052
0.003
5
-0.066
-0.144***
-0.075**
-0.009
6
-0.076*
-0.149***
-0.052
-0.006
7
-0.017
-0.121***
-0.022
0.001
8
-0.018
-0.114***
-0.034
0.000
9
-0.018
-0.099***
-0.020
0.003
10
-0.007
-0.073**
-0.030
0.006
11
-0.027
-0.07**
-0.010
-0.013
12
-0.014
-0.065**
-0.005
0.000
*** p < 0.01; ** p < 0.05; * p < 0.1
As can be seen in Table 10-1, results vary by regulation. For the purposes of this discussion,
we focus on results for the 2007 and 2010 standards.MMMMMMM For the 2007 standards, there is
no statistically significant pre-buy. There is statistically significant low-buy for all the periods
from the period of one month after the standard through the combined period of 12 months after
the standard, with magnitude increasing up to, and falling after, the combined period of 6 months
post-standard. For the 2010 standards, there is some evidence of both pre- and low-buy.
Statistically significant pre-buy can be seen for the period of 1 month up to the combined period
of 5 months, and again at the combined periods of 7 and 8 months pre-standard. There is
significant low-buy for the periods of 1, 2 and 5 months post-regulation. Results indicate that the
observed effects are short-lived, on the order of months rather than years.
14141414141414 We do not consider the results of the 2002 compliance date to be generalizable for several reasons.
Litigation may have affected purchase plans for many firms resulting from the pulling forward of compliance dates
from 2004 to 2002. In addition, there may have been greater concerns over the reliability of new engines compared
to other regulatory actions, which may have led to more low-buy. Also, the cost of compliance in 2002 was
estimated to be lower than that of other regulations. We do not consider the 2014 standards to be generalizable
either. This rule reduced greenhouse gas (GHG) emissions, which had lower technology costs and fuel savings
relative to other rules. In addition, numerous pathways for compliance leads to difficulty estimating the price change
in HD vehicles due to the regulation. More details and discussion on the 2002 and 2004 standards are available in
the contractors' report.
410
-------
10.1.2.1 Estimating Elasticities
To estimate a change in Class 8 vehicle sales due to future EPA emission standards, we
transform the coefficients on the indicator variables, explained above, into demand elasticities.
These elasticities measure the percent change in vehicle sales due to a percent change in vehicle
prices:
Equation 10-1
% AS ales
£ =
%A Price
The percent change in sales for Class 8 vehicles comes from the coefficients on the indicator
variables from Table 10-1. In estimating elasticities, we only use the significant coefficients,
while noting that no response (an elasticity of 0) is also represented in the results. The percent
change in price is estimated by dividing the estimated cost of compliance published in the EPA
RIAs associated with the relevant standards (2007 and 2010) by the estimated purchase price of a
Class 8 vehicle in that year (adjusted to 2010 dollars).NNNNNNN Table 10-2 shows the regulatory
cost, the HD vehicle prices and the resulting percent change in price we are using to estimate the
elasticities from the 2007 and 2010 standards.
Table 10-2: Regulatory Costs and HD Vehicle Prices Used to Estimate Elasticities
Statutory
Deadline
Regulatory Cost
(2010$)
HD Vehicle
Price
% Change in
Price
2007
$9,741
$98,900
9.8%
2010
$7,662
$108,250
7.1%
From the statistically significant pre- and low-buy sales effects for Class 8 vehicles, we
estimate a set of pre- and low-buy elasticities by dividing the percent change in sales, from Table
10-1, by the percent change in price, from Table 10-2.
Table 10-3 shows the estimated statistically significant coefficients (percent change in sales)
from Table 10-1, their period of effect, and the associated estimated elasticity. We expect pre-
buy elasticities to be positive (more sales before new emission standards) and low-buy
elasticities to be negative (fewer sales after new emission standards). Because the smallest
statistically significant sales effect is zero, and a number of other effects are not statistically
different from zero, the smallest pre- and low-buy elasticities are zero - no effect due to the
standards. Not only the magnitude of the elasticity matters, but also the time period over which
the elasticity applies. A large elasticity for a short period may measure less effect than a small
elasticity over a longer period.
NNmMNN The estimated cost of compliance was based on EPA's cost of compliance in the RIA for each regulation.
The price of a Class 8 HD vehicle for each standard was calculated as an average of a high and low list price from
an online source for HD vehicle sales (Commercial Truck Trader, a site that advertises new and used trucks for
sale).
411
-------
Table 10-3: Elasticity Estimates
Statutory
Deadline
Period of
Effect
(Months)
% Change
in Sales (P)
Estimated
Elasticity
All
Any
0
0
2010
8
0.057
0.805
7
0.059
0.834
Prc-Bu\
5
0.054
0.763
4
0.079
1.116
3
0.071
1.003
2
0.105
1.483
1
0.078
1.102
All
Any
0
0
2007
12
-0.065
-0.660
11
-0.070
-0.711
10
-0.073
-0.741
9
-0.099
-1.005
8
-0.114
-1.157
Low-Buv
7
-0.121
-1.229
6
-0.149
-1.513
5
-0.144
-1.462
4
-0.143
-1.452
3
-0.133
-1.350
2
-0.099
-1.005
1
-0.070
-0.711
2010
5
-0.075
-1.060
2
-0.083
-1.173
1
-0.144
-2.034
There are several limitations to the results presented in Table 10-3. As noted in Chapter 10.1.2,
the sales coefficients used to estimate the elasticities likely capture aspects of the proposed
regulation not solely limited to changes in price (e.g., adverse fuel consumption effects, or
concerns about the reliability of untested control technology). Similarly, base vehicle prices and
estimated regulatory costs are estimates and may not correspond with observed base prices or
increased regulatory costs.
These elasticities are based on monthly data, and so it is appropriate to apply the estimated
elasticities to monthly series. Analysis of the coefficients over time indicates that the observed
effects are short-lived, on the order of months rather than years. As noted above and described
below, the time period is a critical factor for estimating the impacts.
10.1.2.2 Illustrative Example
This subsection outlines how we could apply the pre- and low-buy elasticities presented in
Table 10-3 to this rulemaking. Though the methodology to develop the elasticities has been peer
412
-------
reviewed,456 the application in this rulemaking would be new. Thus, in this subsection, we are
illustrating how we could use this new approach in the FRM, or other future rulemakings.
Expanding Equation 10-1, elasticity measures can be approximated as
Equation 10-2
%A Sales ASales Price
e = = *
%A Price Sales APrice
In this application, we want to estimate how a change in price leads to a change in sales.
Therefore, rearranging Equation 10-2, we get
Equation 10-3
A Price
A Sales = e * * Sales
Price
The elasticity measures come from the estimates explained above.
For this example, APrice is the estimated cost of compliance for a Class 8 HD vehicle to
meet the proposed Option 1 MY 2031 standards, $4,203 (seeDRIA, Chapter 7, Table 7-21). We
assume implementation starts January 1.
Price is set to $130,000.0000000
Sales are the estimated monthly Class 8 vehicle sales in 2030 and 2031. Monthly sales are
derived from Class 8 vehicle population data from AEO 2020 and month-specific effects from
the contractors' report.ppppppp To estimate pre-buy, we use the estimated monthly HD vehicle
sales in the months before January 1, 2031. That is, pre-buy for the 2031 compliance date is
estimated with the calculated monthly sales in 2030. To estimate low-buy, we use the estimated
monthly HD vehicle sales in 2031.
To get the sales effects, the elasticity estimates from Table 10-3 are multiplied by the change
in price divided by the base price. This value is then multiplied by the estimated Class 8 HD
vehicle sales for each month over the period of effect for that elasticity measure. This results in a
change in sales for each month over the period of effect. The monthly results are then summed to
get a total affect for each elasticity estimate.
For pre-buy, for example, the elasticity measurement of 1.10 has a period of effect of one
month, so we use the Class 8 sales from December, 2030 and make the pertinent multiplication
for just December, 2030. For the elasticity measurement of 0.81, the period of effect is 8 months,
so we use the Class 8 sales estimates from May, 2030 (8 months before January, 2031) through
December, 2031 and the pertinent multiplication estimation is made for each affected month.
ooooooo j]le prjce of HD vehicles varies greatly, and in part due to the features or options of the vehicle. The price
we use here comes from the estimated price of a low-end, new, semi-truck from Truckers Bookkeeping Service from
June, 2021; a pdf of the page "How much does a semi truck cost?" can be found in the docket for this rule, Docket
ID EPA-HQ-OAR-2019-0055.
ppppppp Because these populations are annual, and the elasticities are monthly, we have to distribute the annual sales
throughout the year. To do so, we estimate the average monthly sales and then use the monthly sales effect (the
percent change in sales by month) estimated in the contractors' report to better approximate the sales by month.
413
-------
Then, the changes in sales for each affected month are added together to get the total effect. The
results for pre-buy are in Table 10-4 below. As discussed in Chapter 10.1.2.1, for all rules there
are sales effects results that are statistically indistinguishable from zero, and thus note that zero
impact on sales is the lower bound on effects. In this example, total sales of Class 8 vehicles
under the Proposed Option 1 are estimated to increase by between 0 and 1.6 percent on an annual
basis before the 2031 compliance deadline. In addition, the duration of the effects is a critical
component in the calculation of sales impacts. For example, the elasticity of 0.83 for a duration
of 7 months has a larger aggregate impact than the larger elasticity of 1.12 for a duration of 4
months. The result that produces the largest estimate for an aggregate increase in sales is the
elasticity of 0.81 for 8 months.
Table 10-4: Illustrative Pre-Buy Results from the 2031 Implementation Date for Proposed Option 1
Period of Effect
(Months)
Elasticity
Aggregate Sales
Change
Cumulative %
Change in Sales
Any
0
0
0
8
0.81
4,108
1.64%
7
0.83
3,727
1.49%
5
0.76
2,461
0.98%
4
1.12
2,900
1.16%
3
1.00
1,971
0.79%
2
1.48
1,919
0.77%
1
1.10
818
0.33%
Low-buy is estimated the same way, though we use monthly sales estimates for the requisite
number of months following the January 1, 2031 compliance date. Low-buy results for the
Proposed Option 1 are show in Table 10-5. For an elasticity of-1.51 over the course of 6 months
(an estimate from the 2007 standards), we use the monthly sales estimates for January, 2031
through June, 2031, make the pertinent multiplication estimations for each affected month, and
add the monthly results to get the aggregate sales change. As discussed in Chapter 10.1.2.1, for
all rules there are sales effects results that are statistically indistinguishable from zero, and thus
note that zero impact on sales is the lower bound on effects. This example estimates sales of
Class 8 vehicles in the months following the 2031 compliance date under the Proposed Option 1
to fall by between 0 and 2.3 percent on an annual basis. As with pre-buy, both the magnitude of
the elasticity and the duration of effect are important in estimating the total effect. For example,
the elasticity of-1.01 for the duration of 9 months (from the 2007 standards) results in a larger
aggregate effect than the larger elasticity of -1.23 for the duration of 7 months. The result that
produces the largest estimate for an aggregate decrease in sales is the elasticity of -1.16 for a
duration of 8 months.
414
-------
Table 10-5: Illustrative Low-Buy Results from the 2031 Implementation Date for Proposed Option 1 QQQQQQQ
Statutory
Deadline
Period of Effect
(Months)
Elasticity
Aggregate Sales
Change
Cumulative %
Change in Sales
All
Any
0
0
0
12
-0.66
(5,072)
-2.00%
11
-0.71
(4,926)
-1.94%
10
-0.74
(4,723)
-1.86%
9
-1.01
(5,720)
-2.25%
8
-1.16
(5,843)
-2.30%
7
-1.23
(5,420)
-2.13%
6
-1.51
(5,773)
-2.27%
5
-1.46
(4,601)
-1.81%
4
-1.45
(3,640)
-1.43%
3
-1.35
(2,513)
-0.99%
r-
2
-1.01
(1.111)
-0.44%
o
o
-------
subject to these proposed standards, so that the net effect of pre-buy is to slow reductions in
emissions.453
Another potential effect of the standards is a net reduction in new vehicle sales. This could
result from either a smaller pre-buy than the post-standards low-buy,RRRRRRR or some potential
buyers deciding not to purchase at all. In this case, the vehicle miles traveled (VMT) of older
vehicles may increase to make up for the VMT otherwise expected of the newer ("missing")
vehicles. To the extent that the older vehicles emit more than the missing vehicles, emissions
may increase.sssssss However, because the VMT is likely to be shifted to the newer HD vehicles
among the existing fleet, and most of those vehicles are expected to be in compliance with the
existing HD vehicle standards, this effect is expected to be small.
Quantifying these effects would require a robust method to estimate the effects of the
standards on pre-buy and low-buy, as well as a method to estimate shifts in VMT among vehicle
vintages in the case of an expected change in the net sales of newer vehicles. In the absence of
robust methods to estimate these effects, EPA is not quantifying the fleet turnover or emissions
impacts in this proposed rule, though, as with pre-buy and low-buy, we acknowledge these
potential impacts.
The estimated increase in operating costs due to an increase in the use of DEF may lead to a
slower fleet turnover if VMT does not shift from older (lower DEF requirements) vehicles to the
newer (higher DEF requirements) vehicles. This would lead to slower emission reductions than if
those newer vehicles were used. However, as this increase in operating costs is small, and may
be offset in part by reduced repair costs, we expect minimal effects on fleet turnover due to the
change in operating costs.
10.1.4 Potential for Mode Shift
Another possible response to the new emissions standards is shifting freight shipments to
other transportation modes, such as rail or barge. This may happen, for example, if the new
standards were to raise operating costs such that truck transportation becomes more expensive
than rail or marine alternatives.
EPA does not expect that this proposed rule is likely to result in a transportation mode shift.
Generally, shipping cargo via truck is more expensive per ton-mile than barge or rail, and less
expensive than air.458-459 This is due to many factors, not the least of which is labor costs (each
truck has at least one driver). Even though trucking is more expensive than rail or marine on a
ton-mile basis, it is a very attractive transportation alternative for several reasons: shipping via
truck is generally faster and more convenient than rail or marine, trucks can reach more places,
and trucks may be less constrained by available infrastructure than barge or rail. In addition,
shipping via truck does not require trans-shipments (transferring from one mode to another, for
example to deliver cargo to or from the port or rail yard), and it allows partial deliveries at many
locations. This speed, infrastructure availability, and delivery flexibility make trucking the
11111111111111 Though this is a possibility, it should be noted that the RZW (2018) study found that the pre-buy
approximately equaled the low-buy. Harrison and LeBel (2008) found that low-buy exceeded pre-buy.
sssssss effect js sometimes called the "Gruenspecht effect," based on the theory presented in Gruenspecht,
Howard (1982), "Differentiated Regulation: The Case of Auto Emissions Standards.". Inierican Economic Review
72:328-331.
416
-------
transportation solution of choice for many kinds of cargo across most distances. As a result,
smaller shipments of higher-valued goods (e.g., consumer goods) tend to be transported by air or
truck, while larger shipments of lower-valued goods (e.g., raw materials) tend to go via rail or
barge.458-460
Studies of intermodal freight shifts, such as Comer et al. or Bushnell and Hughes, focus on
changes in cost per ton-mile as a potential source of transportation mode shift.458-460 Comer et al.
note, for instance, that fuel consumption "depend[s] on the type of freight being moved, route
characteristics, transport speed, and locomotive/truck characteristics."458 Bushnell and Hughes
estimate that increased fuel prices for truck transportation lead to small substitutions between
truck and rail for small for large shipments, and higher shifts for intermediate-sized shipments.460
The findings from this study suggest that the variation in the kinds and values of goods shipped
by different means likely result in only a small amount of mode shift in response to a change in
operating cost (e.g., fuel prices). However, due to data availability, this study approximates
freight rates with fuel costs, assumes shipping distances using different modes are the same, and
mostly does not consider transportation availability constraints affecting some modes in some
regions. These limitations may distort the effects they estimate.
A mode shift study EPA carried out in 2012 in the context of new sulfur limits for fuel used in
large ships operating on the Great Lakes may help address some of these limitations.459 The
methodology used a combination of geospatial modeling and freight rate analysis to examine the
impact of an increase in ship operating costs. While the focus of the study was transportation
mode shift away from marine and toward land, it noted that truck transportation is far more
expensive than both rail and marine on a ton-mile basis.TTTTTTT It also shows that even a large
percentage increase in marine operating costs did not raise freight rates by a similar percentage,
because fuel costs are only part of total operating costs. In the case of truck transportation,
operating costs are a much smaller portion of total costs. The results of this study combined with
the others cited in this section indicate that changing the cost of truck transportation is unlikely to
create mode shift.
The primary effect of the standards on operating costs is an increase in the use of DEF, which
is expected to have a small effect on operating costs, less than 1 cent per vehicle mile traveled
(see Chapter 7.2.1). Because the cost effect is expected to be small, and substitution between
trucks and other modes is limited by the nature of the goods and their routes, we do not expect
significant effects on mode shift. Finally, given the higher costs of truck transportation, a
relatively small increase in truck freight rates due to the small increase in operating costs are
unlikely to affect the competitive dynamics of the transportation sector.
10.1.5 Effects on Domestic and International Shares of Production
The proposed standards are not expected to provide incentives for manufacturers to shift
between domestic and foreign production. This is because the standards would apply to any
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
ttttttt figure 1-5 in U.S. EPA Office of Transportation and Air Quality. "Economic Impacts of the Category 3
Marine Rule on Great Lakes Shipping." EPA-420-R-12-005. 2012.
417
-------
other markets might increase. To the extent that the requirements of these final rules might lead
to application and use of technologies that other countries may seek now or in the future,
developing this capacity for domestic producers now may provide some additional ability to
serve those markets.
10.1.6 Summary of Sales, Turnover, and Mode Shift Impacts
As discussed in Chapter 7, EPA expects the cost of new HD vehicles to increase (see Chapter
7 for details). This increase may have some impact on new vehicle sales; in particular, it is
possible that truck buyers would increase purchases before the standards become effective (pre-
buy), and reduce purchases afterwards (low-buy). Studies of pre-buy and low-buy suggest that
these phenomena have occurred in the past, but those studies do not provide methodologies for
estimating the impact of new rules on future vehicle sales. For that reason, EPA conducted an
analysis to develop a relationship between estimated changes in vehicle price due to a new
regulation and corresponding changes in vehicle sales (i.e., pre- and low-buy elasticities). We
present the details of this new analysis and provide an illustrative example of applying pre- and
low-buy elasticities to estimate potential sales impacts on Class 8 vehicles. For both pre-buy and
low-buy, the illustrative example shows that sales impacts on Class 8 vehicles are of limited
duration and range from zero impact to about two percent.
Whether shippers might choose a different mode for freight depends not only on the cost per
ton-mile of the shipment, but also the value of the shipment, the time needed for shipment, and
the availability of infrastructure. This proposed rule is expected to affect the cost per ton-mile by
only a small amount. For that reason, EPA expects little transportation mode shift to occur due to
the proposed standards. EPA also does not expect changes in where production happens in
response to the proposed standards.
10.2 Employment Impacts
This section explains the methods and estimates of employment impacts due to this proposed
regulation. Though the rule primarily affects HD vehicles, the employment effects may be felt
more broadly in the motor vehicle and parts sectors due to the effects of the standards on sales.
Thus, we focus our assessment on the motor vehicle manufacturing and the motor vehicle parts
manufacturing sectors, with some assessment of impacts on additional sectors likely to be most
affected by the proposed standards. Chapter 10.2.1 offers a brief, high-level explanation of
employment impacts due to environmental regulation and discusses a selection of the peer-
reviewed literature on this topic. Chapter 10.2.2 discusses EPA's qualitative and quantitative
estimates of the partial employment impacts of this proposed rule on regulated industries.
Chapter 10.2.3 examines employment impacts in some closely related sectors, and Chapter
10.2.4 summarizes expected employment impacts.
10.2.1 Economic Framework for Employment Impact Assessment
Economic theory of labor demand indicates that employers affected by environmental
regulation may increase their demand for some types of labor, decrease demand for other types
of labor, or for still other types, not change it at all. A variety of conditions can affect
employment impacts of environmental regulation, including baseline labor market conditions
and employer and worker characteristics such as industry, and region.
418
-------
A growing literature has investigated employment effects of environmental regulation.
Morgenstern et al. decompose the labor consequences in a regulated industry facing increased
abatement costs. They identify three separate components.461 First, there is a demand effect
caused by higher production costs raising market prices. Higher prices reduce consumption (and
production) reducing demand for labor within the regulated industry. Second, there is a cost
effect where, as production costs increase, plants use more of all inputs including labor to
produce the same level of output. Third, there is a factor-shift effect where post-regulation
production technologies may have different labor intensities. These three effects outlined by
Morgenstern et al. provide the foundation for EPA's analysis of the impacts of the current
regulation on labor.461
Additional papers approach employment effects through similar frameworks. Berman and Bui
model two components that drive changes in firm-level labor demand: output effects and
substitution effects. 462'uuuuuuu 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. This is called the output effect. The substitution effect describes
how, holding output constant, regulation affects labor intensity of production. Deschenes
describes environmental regulations as requiring additional capital equipment for pollution
abatement that does not increase labor productivity.463 These higher production costs induce
regulated firms to reduce output and decrease labor demand (an output effect) while
simultaneously shifting away from the use of more expensive capital towards increased labor
demand (a substitution effect).vvvvvvv At the industry level, labor demand is more likely to be
responsive to regulatory costs if: (1) the elasticity of labor demand is high relative to the
elasticity of labor supply, and (2) labor costs are a large share of total production costs.464 Labor
demand might also respond to regulation if compliance activities change labor intensity in
production.
To study labor demand impacts empirically, researchers have compared employment levels at
facilities subject to an environmental regulation to employment levels at similar facilities not
subject to that environmental regulation; some studies find no employment effects, and others
find statistically significant, usually small differences. For example, see Berman and Bui,
Greenstone, Ferris et al., Walker, and Curtis.465'466'467'468
Workers affected by changes in labor demand due to regulation may experience a variety of
impacts including job gains or involuntary job loss and unemployment. Localized reductions in
employment may adversely impact individuals and communities just as localized increases may
have positive impacts. Workforce adjustments in response to decreases in labor demand can be
costly to firms as well as workers, so employers may choose to adjust their workforce over time
through natural attrition or reduced hiring, rather than incur costs associated with job separations
(see, for instance, Curtis and Hafstead and Williams).468'469
1,111,111,1111 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.
vvvvvvv por an overview of the neoclassical theory of production and factor demand, see Chapter 9 of Layard and
Walters (1978).
419
-------
10.2.2 Employment Impacts in the Motor Vehicle and Parts Manufacturing Sectors
The overall effect of the proposed rule on motor vehicle sector employment depends on the
relative magnitude of demand, cost, and factor-shift effects, as described below. As described in
Chapter 10.1.2.2, we provide an example of how pre- and low-buy elasticities could be applied
in the FRM or other future rulemakings, but do not estimate pre- or low-buy for this NPRM;
further, we do not account for the factor-shift effect. Thus, we cannot estimate an overall
employment effect of this proposed rule in these sectors, nor can we infer whether the total effect
would be positive or negative.
Also, due to a lack of firm- and plant-specific data for HD engines and vehicles, EPA is not
quantifying the cost effect impacts of the proposed regulation on employment at the firm level,
although we acknowledge these potential impacts.
In this section, EPA presents partial estimates of industry-level employment effects of the
proposed rule. We do not identify impacts separately for motor vehicle manufacturing and motor
vehicle parts manufacturing, because we do not have information on the division of costs
between these two sectors. Instead, we use the labor intensity of production for these sectors to
provide a range of potential employment impacts.
Our analysis follows the structure of Morgenstern et al., as described above, to estimate the
impacts of this proposed rule on the regulated sector.461 In their theoretical model, a change in
labor demand arising from a change in regulations is decomposed into three main components:
demand, cost, and factor-shift effects. In what follows, we qualitatively describe the employment
impacts due to the demand and factor-shift effects, and quantitatively estimate the employment
impacts due to the cost effect. Because our quantitative estimates of the demand effect are
illustrative, and we do not estimate factor-shift effects, we cannot estimate the total effects on
employment in the regulated industries.
10.2.2.1 The Demand Effect
The demand effect reflects employment changes due to changes in new vehicle sales. Because
we are in the process of refining and finalizing how we might quantify sales impacts, we do not
estimate the demand-effect impact on employment due to the proposed standards. If HD vehicle
sales decrease, fewer people would be needed to assemble trucks and the components used to
manufacturer them. On the other hand, if pre-buy occurs, HD vehicle sales may increase
temporarily, leading to temporary increases in employment. Draft RIA Chapter 10.1.1 and
Chapter 10.1.2 discuss the factors influencing the effect of proposed requirements on demand for
new HD vehicles, explain why that effect is difficult to quantify, proposes a new method to
quantify the impacts and explains how we might use it to estimate pre- and low-buy sales effects
in the FRM.
10.2.2.2 The Factor-Shift Effect
The factor-shift effect reflects employment changes due to changes in labor intensity of
production resulting from to compliance activities. The labor intensity of manufacturing HD
vehicle engines or HD vehicles might increase or decrease because of the proposed rule. Due to a
lack of information on expected changes in labor intensity, the estimated employment impacts in
this chapter do not include the factor-shift effect on employment impact estimates.
420
-------
10.2.2.3 The Cost Effect
The cost effect reflects the impact on employment due to increased costs from adopting
technologies needed for vehicles to meet the standards. The analysis holds output constant,
meaning that it does not include sales impacts. We estimate the cost effect using the historic
share of labor in the cost of production to extrapolate future estimates of impacts on labor due to
new compliance activities in response to the proposed regulation. Specifically, we multiply the
share of labor in production costs by the production cost increase estimated as an impact of this
rule. This provides a sense of the order of magnitude of expected impacts on employment.
The use of the ratio of the share of labor in production costs to estimate "cost effect"
employment has both advantages and limitations. It is often possible to estimate these ratios for
specific sectors, for example, the average number of workers in the HD vehicle manufacturing
sector per $1 million spent in that sector, rather than using ratios from more aggregated sectors,
such as the motor vehicle manufacturing sector. This means that it is not necessary to extrapolate
employment ratios from possibly unrelated sectors. On the other hand, these estimates are
averages, covering all the activities in these sectors and may not be representative of the labor
effects when expenditures are required for specific activities, or when manufacturing processes
change due to compliance activities in such a way that labor intensity changes. For instance, the
ratio of workers to production cost for the motor vehicles manufacturing sector represents this
ratio for all vehicle manufacturing, not just for the HD sector, and not just for production
processes related to emission reductions compliance activities. In addition, these estimates do not
include changes in sectors that supply these sectors, such as steel or electronics producers. The
effects estimated here can be viewed as effects on employment in the HD motor vehicle sector
due to the changes in expenditure in that sector, rather than as an assessment of all employment
changes due to the proposed standards. In addition, labor intensity is held constant in the face of
increased expenditures; this approach does not take in account changes in labor intensity due to
changes in the nature of production (the factor-shift effect), which could either increase or
decrease the employment impacts estimated here.
Some vehicle parts are made in-house by HD vehicle manufacturers. Other parts are made by
independent suppliers who are not directly regulated but would be affected by the proposed
rulemaking as well. Because EPA does not know whether abatement equipment to comply with
the proposed standards would be produced by the original equipment manufacturers (OEMs) or
by suppliers, we use labor ratios for both sectors (and their subsectors) to provide a range of
estimates for the cost effect impacts.
We include estimates from two sectors that are broadly defined and two that are more
narrowly defined. Specifically, we estimate labor impacts for the aggregated sectors 'motor
vehicle manufacturing' and 'motor vehicle parts manufacturing', and for the more specific sectors
'light truck and utility vehicle manufacturing' and 'heavy-duty truck manufacturing.'wwwwwww
We rely on three different public sources to get a range of estimates of employment per $1
million expenditures: the Annual Survey of Manufactures (ASM) and the Economic Census
wwwwwww j]le 'light truck and utility vehicle manufacturing' sector is included because these estimates include
results for trucks from class 2b/3 through 6. See Preamble Section I of this proposal for more discussion on the HD
engine classes included in this rulemaking.
421
-------
(EC), both provided by the U.S. Census Bureau, and the Employment Requirements Matrix
(ERM) provided by the U.S. Bureau of Labor Statistics (BLS). The EC is conducted every 5
years, most recently in 2017. The ASM is an annual subset of the EC and is based on a sample of
establishments. The latest set of data from the ASM is from 2016. The EC and ASM have more
sectoral detail than the ERM, providing estimates out to the 6-digit North American Industry
Classification System (NAICS) code level. They provide separate estimates of the number of
employees and the value of shipments, which we convert to a ratio in this employment
analysis.xxxxxxx The ERM provides direct estimates of employees per $1 million in
expenditures for a total of 202 aggregated sectors that roughly correspond to the 4-digit NAICS
code level, and provides data through 2018. Table 10-6 below shows the sector definition, the
NAICS code, and the ERM sector number where appropriate that EPA uses to estimate
employment effects in this analysis.
Table 10-6: Sectors Used in this Analysis
Sector Definition
NAICS
ERM Sector Number
Motor vehicle manufacturing
3361
80
Light truck and utility vehicle manufacturing
336112
Heavy-duty truck manufacturing
33612
Motor vehicle parts manufacturing
3363
82
Table 10-7 provides the estimates of employment per $1 million of expenditure for each
sector for each data source, adjusted to 2017 dollars using the U.S. Bureau of Economic Analysis
Gross Domestic Product Implicit Price Deflator retrieved from the Federal Reserve Bank of St.
Louis. The values are adjusted to remove effects of imports through the use of a ratio of domestic
production to domestic sales of o.85.YYYYYYY Although the estimated labor ratios are not the
same across data sources, they each exhibit a similar pattern across sectors. Within the 4-digit
NAICS code level, motor vehicle parts manufacturing seems to be the most labor-intensive
sector, followed by the motor vehicle manufacturing sector. Within motor vehicle
manufacturing, heavy-duty truck manufacturing appears to be more labor-intensive than light
truck and utility vehicle manufacturing.
xxxxxxx j]le total employment across the four sectors used in this analysis (see Table 10-6) as reported in the ASM
and the EC ranges from 213,212 to 226,259 depending on which data source is used; as noted above the most recent
ASM and EC were conducted in 2016 and 2017, respectively.
yyyyyyy j0 estimate the proportion of domestic production affected by the change in sales, we use data from
WardsAuto for total car and truck production in the U.S. compared to total car and truck sales in the U.S. Over the
period 2009-2018, the proportion averages 85 percent. From 2015-2018, the proportion average is slightly lower, at
84.3 percent.
422
-------
Table 10-7: Employment per $1 Million Expenditures (2017$) in the Motor Vehicle Manufacturing Sector3
Source
Sector (NAICS)
Ratio of
Workers per
$1 Million
Expenditures
Ratio of Workers per $1
Million Expenditures,
Adjusted for Domestic vs.
Foreign Production
BLS ERM
2017
Motor vehicle manufacturing (3361)
0.584
0.497
ASM 2016
Motor vehicle manufacturing (3361)
0.563
0.479
EC 2017
Motorvehicle manufacturing (3361)
0.615
0.523
BLS ERM
2017
Motor vehicle parts manufacturing (3363)
1.999
1.700
ASM 2016
Motorvehicle parts manufacturing (3363)
2.032
1.728
EC 2017
Motorvehicle parts manufacturing (3363)
2.231
1.897
ASM 2016
Heavy-duty truck manufacturing (33612)
1.048
0.891
EC 2017
Heavy-duty truck manufacturing (33612)
0.988
0.840
ASM 2016
Light truck and utility vehicle
manufacturing (336112)
0.454
0.386
EC 2017
Light truck and utility vehicle
manufacturing (336112)
0.478
0.407
a BLS ERM refers to the U.S. Bureau of Labor Statistics' Employment Requirement Matrix, 2018 values.
ASM refers to the U.S. Census Bureau's Annual Survey of Manufactures, 2016 values. EC refers to the
U.S. Census Bureau's Economic Census, 2017 values. These are the most recent data available.
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, provides estimates that, in 1997, about 1.4 workers in the Motor Vehicle Manufacturing
sector were needed per $1 million, but only 0.6 workers by 2018 (in 2017$).zzzzzzz Because the
ERM is available annually for 1997-2018, we use these data to estimate productivity
improvements over time. We regress logged ERM values on a year trend for the Motor Vehicle
Manufacturing and Motor Vehicle Parts Manufacturing sectors. We use 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 of the regressions suggest a 4.3
percent per year productivity improvement in the Motor Vehicle Manufacturing Sector and a 4.0
percent per year improvement in the Motor Vehicle Parts Manufacturing Sector.
We then use those estimated percent improvements in productivity to project the number of
workers per $1 million of production expenditures through 2032. Although the costs and benefits
analyses in the preceding chapters go out to 2045, we chose to model the employment effect due
to cost increases only through 2032. This is because our method is an approximate, partial
employment analysis, as well as being dependent on future, uncertain, macro-economic
conditions. The choice of 2032 was also affected by the phase-in of more stringent requirements
starting in 2031 in the proposal. The results provided below represent an order of magnitude
effect, rather than definitive impacts. We calculate separate sets of projections (adjusted to
zzzzzzz https://www.bls.gov/emp/data/emp-requirements.htm; this analysis used data for sectors 80 (Motor Vehicle
Manufacturing) and 82 (Motor Vehicle Parts Manufacturing) from "Chain-weighted (2009 dollars) real domestic
employment requirements tables;" see "Cost Effect Employment Impacts calculation" in the docket.
423
-------
2017$) for each set of data, ERM, EC, and ASM, for all four sectors described above. The ERM
projections are calculated directly from the fitted regression equations used to estimate the
projected productivity growth, since the regressions themselves used ERM data. For the ASM
and EC projections, we use the ERM's ratio of the projected productivity growth value in each
future year to the projected production expenditure value in 2016 for the ASM and 2017 for the
EC (the base years in our data) to determine how many workers would be needed per $1 million
of expenditures, in 2017$. In other words, we apply the projected productivity growth estimated
using the ERM data to the ASM and EC numbers.
To simplify the results, we compare the projected employment across the ten data sources
from the ERM, EC, and ASM, and report only the maximum and minimum effects in each year
across all sectors.AAAAAAAA We provide a range rather than a point estimate because of the
inherent difficulties in estimating employment impacts as well as the uncertainty over how the
costs are expended. The reported ranges provide an estimate of the expected magnitude of the
labor effect. In Table 10-7, the Motor Vehicle Parts Manufacturing Sector value from the EC
provides the maximum employment estimates per $1 million; the Light Truck and Utility
Vehicle Manufacturing Sector value from the ASM provides the minimum estimates.
Cost estimates developed for this proposed rule are provided in Chapter 7. We use the
technology cost estimates from that chapter to estimate the employment impacts of a change in
cost of manufacturing HD vehicles due to this rule. The technology cost estimates (in $ million)
are multiplied by the estimates of workers per $1 million in costs. The projected estimates of
technology costs and corresponding minimum and maximum estimated employment impacts for
each year are shown in Table 10-8, below. The effects are shown in job-years, where a job-year
is, for example, one year of full-time work for one person or two years of half-time work for two
workers. Increased technology costs of vehicles and parts would, by itself and holding labor
intensity and output constant, be expected to increase employment by between 300 and 2,200 per
year between 2027 and 2032 under both Proposed Option 1 and Proposed Option 2. While we
estimate employment impacts, measured in job-years, beginning with program implementation,
some of these employment gains may occur earlier as vehicle manufacturers and parts suppliers
hire staff in anticipation of compliance with the standards.
aaaaaaaa -pQ scc details, as well as results for all sources, see "Cost Effect Employment Impacts calculation" in the
docket.
424
-------
Table 10-8: Estimated Employment Effects Due to Increased Costs of Vehicles and Parts (Cost Effect), in
Job-Years
Year
Proposed Option 1
Proposed Option 2
Minimum
Employment
due to Cost
Effect3
Maximum
Employment
due to Cost
Effectb
Minimum
Employment
due to Cost
Effect3
Maximum
Employment
due to Cost
Effectb
2027
400
2,200
400
2,200
2028
400
2,100
400
2,000
2029
400
2,000
400
1,900
2030
300
1,800
300
1,700
2031
400
1,900
300
1,600
2032
300
1,800
300
1,500
Note: Though costs differ between Proposed Option 1 and Proposed Option 2,
results in this table may be the same due to rounding.
10.2.2.4 Summary of Employment Effects in the Motor Vehicle Sector
As explained above, the overall effect of the proposed standards on motor vehicle sector
employment depends on the relative magnitude of the demand, cost, and factor-shift effects, as
described by Morgenstern, et al 461 EPA proposed a method in draft Chapter 10.1.2.2 to quantify
the demand effect, however we are not estimating pre- or low-buy effects in this NPRM, and
therefore are not using demand effects to estimate employment impacts. Also, due to a lack of
data, we do not estimate factor-shift effects. However, we estimate employment impacts of the
proposed standards due to the cost effect.
For the regulated sector, the partial employment impact due to the effect of increased
manufacturing costs due to compliance activities is expected to range between 400 to 2,200 job
years and 300 to 1,800 job years in 2027 and 2032 respectively under the Proposed Option 1.
The partial employment impact under the Proposed Option 2 is expected to range between 400 to
2,200 job years and 300 to 1,500 job years in 2027 and 2032 respectively. We would expect the
demand effect to reduce these employment increases, although we are unable to predict the
magnitude of that reduction. Finally, we are unable to predict the direction of the factor-shift
effect.
10.2.3 Employment Impacts on Related Sectors
10.2.3.1 Effects on Purchasers of Heavy-Duty Vehicles
Because of the diversity of HD vehicles, entities from a very wide range of transportation
sectors would be purchasing vehicles subject to these proposed standards. As discussed in
Chapter 10.1, vehicles subject to these proposed standards would be likely to be more expensive
to purchase compared to vehicles not subject to the proposed standards. HD vehicles are
typically commercial, and typically provide an "intermediate good:" that is, they are used to
provide a commercial service, rather than being a final consumer good. As a result, the higher
costs of the vehicles may result in higher prices for the services provided by these vehicles, and
potentially reduced demand for those services. In turn, there might be less employment in the
sectors providing those services.
425
-------
The American Trucking Association, in comments on the ANPR, provided data indicating
that, between 2010 and 2018, the up-front costs of HD vehicles were approximately 10 to 16
percent of per-mile costs.470 Given that we expect the potential impacts of the proposed
standards on up-front costs of vehicles to be only a few percent of total vehicle cost, the likely
increase in per-mile costs would likely be less than 1 percent (even a 5 percent increase in
vehicle costs would result in only a 0.8 percent increase in per-mile costs, at the high end of the
range). Therefore, we do not expect large impacts on transportation services demand, and related
employment in transportation services sectors. Per-mile cost increases for some sectors would be
higher than this average, while they would be lower in other sectors. The actual effects on
demand for the services and related employment would depend on cost pass-through, and
responsiveness of demand to transportation cost increases.
Lengthening the warranty period may provide some positive impacts on employment for
vehicle purchasers. Both the National Association of Small Trucking Companies (NASTC) and
the Owner-Operator Independent Drivers Association (OOIDA), in comments on the ANPRM,
expressed concerns over the effects on productivity of downtime due to problems with emission
control systems.471 They requested that EPA pursue ways to reduce those losses, with a NASTC
member specifically requesting warranty extensions. As discussed in Chapter 7.2.3, the extended
warranty provisions proposed here would not only be expected to reduce repair costs for vehicle
purchasers, but may also provide incentives to manufacturers to improve quality and thus reduce
the need for repairs. These effects would be expected to reduce costs, and thus mitigate adverse
impacts on employment for vehicle purchasers.
10.2.3.2 Effects on Heavy-Duty Vehicle Dealers and Service Providers
If sales of HD vehicles decrease, then HD vehicle dealers would have fewer sales, and may
employ fewer people as a result. At the same time, dealers, and other independent service repair
shops, often provide repair and maintenance services. The extended warranty provisions would
be expected to facilitate repair and maintenance of emission control system components, which
could result in increased demand for workers servicing vehicles. On the other hand, as discussed
in Chapter 7.2.3, the extended useful life provisions would also be expected to provide incentives
to OEMs to improve quality and may reduce the need for warranty claims. Thus, effects on
employment for service providers, including dealers, may be positive or negative.
Similarly, as discussed in Preamble Section IV.B.3, this proposal aims to ease the ability of
independent service repair shops and independent owner-operators to service vehicles (i.e.,
serviceability). Comments from OOIDA on the ANPR note that it is currently difficult for
anyone other than dealers to service vehicles. It is possible that improving serviceability will
improve maintenance due to lower costs of conducting it.472 It is also possible that improved
serviceability may shift some of the work, and thus employment, from dealers to independent
service repair shops or to the owners themselves, with an unclear impact on the overall level of
employment.
10.2.4 Summary of Employment Impacts
Employment in the HD manufacturing sector depends on three effects: how the effects of the
standards on vehicle prices affect demand for new vehicles (demand effect); the labor demand
needed to meet the standards (cost effect); and any change in labor intensity of production due to
complying with the standards (factor-shift effect). As discussed in Chapter 10.2.2, EPA is
426
-------
proposing a method to quantify sales effects, which may allow us to quantify demand effects, but
we are not using it to quantify effects in this NPRM. In addition, we are unable to quantify the
factor-shift effect, and therefore we are unable to estimate their impact on net employment in the
HD manufacturing sector. Cost-effect-related employment changes due to the proposed
regulation are estimated to be between a few hundred and a few thousand employees. For
comparison, in May 2019, the Bureau of Labor Statistics reports about 234,000 employees in
Motor Vehicle Manufacturing.473
Other sectors that sell, purchase, or service HD vehicles might also experience employment
impacts due to the standards. The effects on these sectors would depend on the degree of cost
pass-through to prices for HD vehicles and the effects of useful life and warranty requirements
on demand for vehicle repair and maintenance.
427
-------
Chapter 11 Small Business Analysis
This chapter presents our analysis of the potential impacts of the proposal on small entities
that would be subject to the highway heavy-duty engine and vehicle provisions of this proposed
rule. These are: heavy-duty vehicle manufacturers, heavy-duty secondary vehicle manufacturers,
and heavy-duty alternative fuel engine converters. Other entities that would be subject to the rule
are either not small (e.g., engine and incomplete vehicle manufacturers) or are not expected to
incur any burden from the proposed rule (e.g., in sectors other than highway heavy-duty engines
and vehicles).
11.1 Definition and Description of Small Businesses
Under the Regulatory Flexibility Act (5 USC 601 et seq.), a small entity is defined as: (1) a
business that meets the definition for small business based on the Small Business
Administration's (SBA) size standards;474 (2) a small governmental jurisdiction that is a
government of a city, county, town, school district or special district with a population of less
than 50,000; or (3) a small organization that is any not-for-profit enterprise which is
independently owned and operated and is not dominant in its field.
This analysis considers only small business entities. Small governmental jurisdictions and
small not-for-profit organizations are not subject to the proposed rule as they have no
certification or compliance requirements.
11.2 Overview of the Heavy-Duty Program and Type of Entities Covered
As described in the proposal and elsewhere in this draft RIA, this rulemaking sets out a
comprehensive approach to reduce air pollution from highway heavy-duty engines. The key
provisions can be grouped into three broad categories: 1) reducing emissions under a wider range
of engine operating conditions than those covered in existing requirements, including refueling
events (e.g., revised exhaust and refueling emission standards and updated test procedures); 2)
maintaining emission control over a greater portion of an engine's operational life (e.g.,
lengthened useful life and regulatory emission warranty periods), and 3) providing manufacturers
with flexibilities to meet the proposed standards while clarifying our regulations. We are also
proposing revisions to a subset of the MY 2027 GHG standards for heavy-duty vehicles.
While the proposed rule also includes regulatory amendments for sectors other than highway
heavy-duty engines and vehicles, these amendments for other sectors correct, clarify, and
streamline the regulatory provisions, and there is no burden from the proposed rule on small
entities in these other sectors.
There are four categories of highway heavy-duty engine and vehicle entities that are subject to
the proposed rule:
• Heavy-duty engine manufacturers
• Heavy-duty conventional vehicle manufacturers, including incomplete and secondary
vehicle manufacturers
• Heavy-duty electric vehicle manufacturers
• Alternative fuel engine converters
428
-------
Heavy-duty engine manufacturers have been developing, testing, and certifying engines for
many years in compliance with EPA rulemakings adopted under the CAA. Under this proposal,
these companies will be required to produce engines that meet new emission standards and
certify them using revised test procedures. The heavy-duty engine manufacturers that certify
engines to EPA's program include no small entities.
Heavy-duty vehicle manufacturers are one of two types. The first type of company
manufactures both the engine and the associated vehicle and these companies are not small
entities. The second type of vehicle manufacturer, some of which are small, produces a vehicle
of its own design using a certified engine produced by a different company. In a variation on the
second type, the heavy-duty vehicle manufacturer finishes an incomplete vehicle produced and
certified by a different company; these so-called "secondary manufacturers" complete the
vehicle by adding the truck body and other equipment. While these secondary manufacturers are
not required to certify with EPA, they may incur costs to accommodate changes made to the
certified incomplete vehicles to meet certain proposed requirements. Several secondary
manufacturers are small entities under the SBA definition, and the economic impacts of the
proposed rule on these companies are described in Section 11.3. The economic impacts of the
proposed rule on other heavy-duty vehicle manufacturers, including electric heavy-duty vehicle
manufacturers, are described in Section 11.4.
Alternative fuel engine converters are also subject to the proposed rule. Several of these
companies are small entities under the SBA definitions, and the impacts on them are described in
Section 11.5. Finally, Section 11.6 contains a summary table of the estimated economic impacts
on small entities subject to the rule.
11.3 Impacts on Small Entities: Heavy-Duty Secondary Vehicle Manufacturers
A secondary vehicle manufacturer is defined as anyone that produces a vehicle by modifying
a complete vehicle or completing the assembly of a partially complete vehicle, although a
manufacturer controlled by the manufacturer of the base vehicle (or by an entity that also
controls the manufacturer of the base vehicle) is not a secondary vehicle manufacturer; instead
both entities are considered to be one manufacturer (40 CFR 1037.801 definition of secondary
vehicle manufacturer). EPA's heavy-duty vehicle program allows an engine manufacturer to
introduce partially complete vehicles into U.S. commerce to be completed by a secondary
vehicle manufacturer (see 40 CFR 1037.622). These incomplete vehicles will typically be
certified by the engine manufacturer. The program also allows a manufacturer to introduce
complete vehicles into U.S. commerce for modification by a small manufacturer (e.g., a
recreational vehicle manufacturer); these also will typically be certified by the engine
manufacturer. The provisions specify that a secondary vehicle manufacturer may finish assembly
of partially complete vehicles if it obtains a vehicle that is not fully assembled with the intent to
manufacture a complete vehicle in a certified configuration.
Secondary vehicle manufacturers that produce a heavy-duty vehicle using a compression-
ignition incomplete vehicle are not expected to need to modify their manufacturing or other
processes to comply with the proposed rule. As discussed in Chapter 3 of this draft RIA,
compression-ignition engine manufacturers are expected to achieve the proposed criteria
pollutant engine emission standards by modifying the engine and aftertreatment system
technologies already applied to these engines to meet the existing standards (e.g., selective
catalytic reduction). As a result, the engine and aftertreatment systems to meet the proposed
429
-------
criteria pollutant emission standards are expected to be similar to the systems vehicle
manufacturers install today and secondary vehicle manufacturers are not expected to need to
redesign or modify their production processes to accommodate compression-ignition engine-
based certified systems. As a result, we do not expect secondary vehicle manufacturers would
experience any adverse impacts by using incomplete vehicles that comply with the compression-
ignition exhaust criteria pollutant emission standards in this proposal.
Secondary vehicle manufacturers that finish an incomplete vehicle certified to the proposed
spark-ignition engine emission standards would be subject to the provisions of the proposed rule.
Similar to manufacturers of compression-ignition engines, Chapter 3 indicates that spark-ignition
engine manufacturers are expected to achieve the proposed criteria pollutant engine emission
standards by modifying the engine and aftertreatment system technologies already applied to
these engines to meet the existing standards (e.g., three-way catalysts). As a result, the engine
and aftertreatment systems are expected to be similar to the systems vehicle manufacturers install
today and secondary vehicle manufacturers are not expected to need to redesign or modify their
production processes to accommodate spark-ignition engine-based certified systems. As a result,
we do not expect secondary vehicle manufacturers would experience any adverse impacts by
using incomplete vehicles that comply with the spark-ignition exhaust criteria pollutant emission
standards in this proposal.
We are also proposing refueling emission standards for incomplete heavy-duty vehicles fueled
by volatile fuels (e.g., gasoline and ethanol). Compliance with these standards may require some
secondary vehicle manufacturers to change their manufacturing or other processes to
accommodate compliant refueling emission control systems. Historically, an incomplete vehicle
that is sold to a secondary vehicle manufacturer includes the fuel system and its evaporative
emission controls as part of the incomplete vehicle's certified configuration. When
manufacturers of chassis-certified complete heavy-duty vehicles (i.e., Classes 2b and 3) adopted
ORVR technology to meet refueling emission standards (59 FR 16262, April 6, 1994), the design
changes were contained in the fuel system and did not require changes to the vehicle body to
accommodate the new technology (i.e., filler door location and designs remained the same). In
the case of the proposed provisions, manufacturers of incomplete heavy-duty vehicles fueled by
volatile fuels are expected to achieve compliance with the proposed refueling standards by
adding to or replacing existing components of the evaporative emission control systems currently
being installed on incomplete vehicles. We expect incomplete vehicle manufacturers will design
compliant ORVR systems that maintain continuity from previous fuel system designs,
minimizing the need for vehicle body redesign to accommodate any changes to emission control
systems. Our expectation is reinforced by the comments we received on the ANPR to this rule, in
which ORVR suppliers expressed confidence in the relationship between engine OEMs and
delegated assemblers (i.e., secondary vehicle manufacturers) to effectively implement refueling
requirements for incomplete heavy-duty vehicles.BBBBBBBB Secondary manufacturers that finish
the vehicle bodies have many years of experience installing evaporative emission control
systems as delegated assemblers, and the ORVR instructions are expected to add very few, if
any, steps to the evaporative system instructions currently provided by the chassis manufacturers.
bbbbbbbb gee comments from the Manufacturers of Emission Controls Association (EPA-HQ-OAR-2019-0055-
0365) and Ingevity Corporation (EPA-HQ-OAR-2019-0055-0271).
430
-------
Secondary vehicle manufacturers are not required to certify with EPA, and so we do not have
a list of secondary manufacturers that will be subject to the rule. Instead, we used the Hoovers
D&B database to identify small companies engaged in the Motor Vehicle Body Manufacturing
(NAICS Code 336211 with 1,000 employees or fewer) or Motor Home Manufacturing (336213
with 1,250 employees or fewer) sectors.475 We limited our search to companies located in the
United States. This approach is reasonable because it is unlikely that a foreign entity would
purchase a certified incomplete vehicle from a manufacturer located in the United States,
transport that vehicle to a location outside the United States, complete the vehicle, and then
export that completed vehicle back into the United States with the associated transportation
costs. If there were such a company, the cost of the additional transportation to and from the
assembling country would likely exceed the expected costs of compliance with the proposed rule
($2,528 per company per year; see below). Also, the additional transportation costs would likely
make the completed vehicle uncompetitive in the U.S. market (with the exception of luxury
trucks or recreational vehicles, in which case the company would likely have revenue that can
accommodate the costs of the proposed program).
We adjusted the initial list of 1,190 companies to remove those that are subsidiaries of another
company (they have a parent or ultimate parent company). For the Motor Vehicle Body
Manufacturing sector, we further adjusted the list to reflect only companies engaged in truck and
bus body manufacturing (as identified by their Standard Industrial Classification (SIC) code) and
removed companies that do not make truck bodies or that make light-duty trucks (these
companies would not be subject to the rule) cccccccc For the Motor Home Manufacturing
sector, we selected only companies engaged in motor home manufacturing (using their SIC code)
and removed van conversions (those are light-duty vehicles not subject to the rule). Finally, we
removed companies with four or fewer employees because it is not likely that a company with
fewer than five total employees manufactures completed trucks. It should be noted that we also
removed one company from the list that does not appear to have become operational (the $100
annual revenue reported in Hoovers for this company was likely a placeholder). This procedure
yielded a list of 249 small entities engaged in the manufacture of secondary vehicles and thereby
potentially subject to the proposed refueling standards. It should be noted that the final list of
companies does not distinguish between those that produce spark-ignition and compression-
ignition vehicles.
We estimated the impacts of the proposed rule on these small entities using the following
information. We assume each company would have one-time costs associated with reviewing
new instructions for ORVR (10 hours/family), possible vehicle design R&D if the change to the
evaporative control cannister requires different mounting assemblies or the area where it is
mounted must be adjusted (8 hours/family), and training associated with installing the new
ORVR system (1 hour/family). We also assume a recurring production cost for installing the
new ORVR system (1 hour/unit). Assuming 2 families with 20 unit per family for each company
(40 vehicles produced per year) and $43.58/hour, the total cost of the program in the first year is
cccccccc Removed: cranes, overhead traveling; dump truck lifting mechanism; fifth wheel, motor vehicle; truck
beds; truck bodies (motor vehicles); truck bodies and parts; truck cabs, for motor vehicles; truck tops; van bodies.
431
-------
expected to be $2,528 per company. We then compared this to the annual revenue reported in
Hoovers for each of the small entities.DDDDDDDD
For these secondary vehicle manufacturers, the expected costs of $2,528 is less than 1 percent
of revenue for 201 of the companies and between 1 and 3 percent of revenue for 48 companies.
Figure 11-1 contains a graphical representation of the revenue distribution of these companies.
The impacts are summarized in Table 11-1 presented in Section 11.6.
100
90
80
70
60 -
50 -
40 -
30
20
10 H
0
Small Entities - Compliance Costs as Percent of Annual Revenue
(Source: Hoovers D&B, accessed August 2021)
~ NAICS 336211 Motor Vehicle Body
Manufacturing, n=217
89
n
69
14
14
10
I
1
4
5 2
9
1 0
¦ NAICS 336213 Motor Home
Manufacturing, n=32
10
*1
5
I—| 0
r 1 1
0 i
i i
3 1
H1
o\o o\o o\o o\o o\o o\o o\o o\o o\o o\o o\o o\o o\o
# A # ^ ^ y y
* * * ^
o- o- o- o-' v' ,v' V' jv' ,y a/ ,v
S> S> ^ f£> rh A^> r?> rh
-------
11.4.1 Heavv-Dutv Electric Vehicle Manufacturers
Under the proposed revisions to the MY 2027 GHG vehicle standards, the electric vehicle
manufacturers would not incur any additional cost. Manufacturers of HD electric vehicles are
currently subject to the Heavy-Duty Phase 2 GHG standards. However, electric vehicles are
deemed to have zero CCh/ton-mile emissions and would meet the proposed revised standards
without any additional GHG-reducing technologies.
Under the criteria pollutant portion of the proposed rule, electric vehicle manufacturers would
attest to meeting the proposed criteria pollutant standards using the same process as they
currently use for greenhouse gas standards. For this analysis, we calculated an impact of time to
compile materials to certify and to update the labels on their vehicles that we believe
conservatively overestimates the burden of our proposal for these manufacturers.
EPA identified 25 companies that make electric heavy-duty vehicles. These are companies
that certified with EPA as of 2021. We obtained their employment and annual revenue numbers
from Hoovers. Of these, 13 are small entities under the SB A definitions.
We estimated the impacts of the rule on these small entities using the following information.
We assume each company would have one-time costs associated with compiling the materials to
certify to criteria pollutants (1 hour/family) and with updating their vehicle label (1 hour/family).
There are no recurring costs per vehicle expected. Assuming 1 family and $43.58/hour, the total
cost of the program in the first year is expected to be $88 per company.
Our examination of the annual revenues for the 13 small electric heavy-duty vehicle
manufacturers that would be subject to the rule reveals that these costs, $88 per company per
year, would not impose a significant impact on any of them. Even a low significant impact
threshold of 1 percent of revenue would correlate to an annual revenue of less than $8,800. The
13 small electric heavy-duty vehicle manufacturers have annual revenues of $78,000 or more and
would not have a significant impact from the proposed rule. These results are summarized in
Table 11-1 presented in Section 11.6.
11.4.2 Heavv-Dutv Conventional Vehicle Manufacturers
Heavy-duty vehicle manufacturers are currently subject to the greenhouse gas emission
standards in 40 CFR part 1037. As part of the HD GHG Phase 2 rulemaking, consistent with the
recommendations of the Small Business Advocacy Review Panel, we adopted a custom chassis
program as part of the vocational vehicle program that includes less stringent standards and a
simplified GEM process (81 FR at 73526). In addition, we provided a one-year delay in the
Phase 2 vehicle GHG emission standards for small vehicle manufacturers (see 40 CFR
1037.150(c)(2)).
In our review of information provided by vehicle manufacturers during the MY 2021
greenhouse gas emissions certification process, we found one conventional vehicle manufacturer
that is a small business. For this manufacturer, we reviewed the types of vehicles they
manufacture, their most recent production volumes, and their annual revenue per Hoover's.
The small manufacturer produces vehicles that would be considered under the custom chassis
program or "urban" vehicles under the vocational vehicle standards in 40 CFR 1037.105. In
Phase 2, we projected the technology costs to meet the MY 2027 CO2 emission standards at
433
-------
$2,727 for medium heavy-duty urban vehicles and $4,121 for heavy heavy-duty urban vehicles
(81 FR at 73718, Table V-30). As discussed in Section XI.C of the preamble, the proposed
revisions to the MY 2027 emission standards would require manufacturers to apply Phase 2
technologies at the costs presented here to approximately five percent of conventional vehicles
that may not otherwise have applied those technologies absent this proposal. Applying these
technology costs to five percent of the company's MY 2019 production volumes (the most recent
data available) for the two vehicle subcategories, we estimate that the impact of proposed
standards would be less than one percent of the company's revenues. Therefore, we determined
that there would not be a significant impact on this manufacturer due to the proposed revisions to
the MY 2027 CO2 emission standards. These results are summarized in Table 11-1 presented in
Section 11.6.
11.5 Impacts on Small Entities: Heavy-Duty Alternative Fuel Engine Converters
Companies that convert compression-ignition or spark-ignition heavy-duty engines to use
alternative fuels would be subject to the proposed rule. Alternative fuel converters do not always
need to certify the conversions to the proposed emission standards. However, they may need to
perform testing to show that modified engines continue to meet the applicable new standards as
part of the process for meeting 40 CFR part 85, subpart F to be exempt from the tampering
prohibition. For this analysis, we conservatively assumed the process for these fuel converters
would include emission testing using a new test procedure (SET if converting an SI engine or
LLC if converting a CI engine), and preparing the data for submission to EPA to obtain an EPA
Certificate of Compliance.
EPA identified 5 companies who convert heavy-duty engines to run on alternative fuel. These
are companies that certified engines with EPA as of 2020. We obtained their employment and
annual revenue numbers from Hoovers. Two of these companies are small entities under the
SBA definitions based on annual receipts.
To estimate the impacts of the proposed rule on these small entities we assume that each
company will have one-time costs associated with developing and performing the emission test
of 20 hours/family. There are no recurring costs per vehicle expected. Assuming 4 families and
$43.58/hour, the total cost of the program in the first year is expected to be $3,486 per company.
Our examination of the annual revenues for the two small alternative fuel engine converters
reveals that these costs, $3,486 per company per year, would not impose a significant impact on
either of them. Even a low significant impact threshold of 1 percent of revenue would correlate
to an annual revenue of $348,640 or less. Each of the two small alternative fuel engine
converters has annual revenues in excess of that amount and would not have a significant impact
from the proposed rule. These results are summarized in Table 11-1 presented in Section 11.6.
434
-------
11.6 Summary Table of Impacts on Small Businesses Subject to the Rule
Table 11-1: Summary of Impacts on Small Businesses Subject to the Rule
NAICS
Category
Sector description
SBA
Threshold
Number of
small companies
subject to the
proposed rule
Impact as
percent of annual
revenue,
number of small
companies
>3%
1-3%
<1%
336211
Secondary manufacturer:
Motor vehicle body manufacturing
1,250
employees
217
0
41
176
336213
Secondary manufacturer:
Motor home manufacturing
1,250
employees
32
0
7
25
Total secondary manufacturer
249
0
48
201
336120
Heavy-duty vehicle manufacturers
(including electric and conventional)
1,500
employees
14
0
0
14
811198
Alternative fuel engine converters
$8.0 million
annual
receipts
2
0
0
2
TOTAL
265
0
48
217
435
-------
References
1 Khalek, I. A., Blanks, M. G., Merritt, P. M., & Zielinska, B. (2015). Regulated and unregulated
emissions from modern 2010 emissions-compliant heavy-duty on-highway diesel engines.
Journal of the Air & Waste Management Association, 65(8), 987-1001.
2 Rudolph, J., Jacob, C.R. (2019) "Computational Insights into the Mechanism of the Selective
Catalytic Reduction of NOX: Fe- versus Cu-Exchanged Zeolite Catalysts." ACS Omega, 4,
7987-7993.
3 Lambert, C.K. "Perspective on SCR NOX control for diesel vehicles." Reaction Chemistry &
Engineering, 2019, 4, 969.
4 Fan, C., Chen, Z., Pang, L., Ming, S., Zhang, X., Albert, K. B., . . . Li, T. (2018). The influence
of Si/Al ratio on the catalytic property and hydrothermal stability of Cu-SSZ-13 catalysts for
NH3-SCR. Applied Catalysis A: General, 550, 256-265.
5 Fedyko, J. M. and H.-Y. Chen (2015). Zeolite Catalyst Containing Metals. U. S. Patent No.
US20150078989A1, Johnson Matthey Public Limited Company, London.
6 Cui, Y., Wang, Y., Walter, E. D., Szanyi, J., Wang, Y., & Gao, F. (2020). Influences of Na+
co-cation on the structure and performance of Cu/SSZ-13 selective catalytic reduction catalysts.
Catalysis Today, 339, 233-240.
7 Fedyko, J. M. and H.-Y. Chen (2019). Zeolite Catalyst Coating Containing Metals. U. S. Patent
No. US 20190224657A1, Johnson Matthey Public Limited Company, London, UK.
8Wang, A., Wang, Y., Walter, E. D., Washton, N. M., Guo, Y., Lu, G., et al. (2019). NH3-SCR
on Cu, Fe and Cu+ Fe exchanged beta and SSZ-13 catalysts: Hydrothermal aging and propylene
poisoning effects. Catalysis Today, 320, 91-99.
9 McDonald, J. F., Schenk, C., Sanchez, L. J., & Nelson, B. J. (2011). Testing of catalytic
exhaust emission control systems under simulated locomotive exhaust conditions. SAE
Technical Paper No. 2011-01-1313.
10 Pereira, M. V. L., Nicolle, A., & Berthout, D. (2015). Hydrothermal aging effects on Cu-
zeolite NH3-SCR catalyst. Catalysis Today, 258, 424-431.
11 Luo, J., Wang, D., Kumar, A., Li, J., Kamasamudram, K, Currier, N., & Yezerets, A. (2016).
Identification of two types of Cu sites in Cu/SSZ-13 and their unique responses to hydrothermal
aging and sulfur poisoning. Catalysis Today, 267, 3-9.
12 Kumar, A., Kamasamudram, K, & Yezerets, A. (2013). Hydrocarbon Storage on Small-Pore
Cu-Zeolite SCR Catalyst. SAE International Journal of Engines, 6(2), 680-687.
13 Kumar, A., Smith, M. A., Kamasamudram, K, Currier, N. W., & Yezerets, A. (2016).
Chemical deSOx: An effective way to recover Cu-zeolite SCR catalysts from sulfur poisoning.
Catalysis Today, 267, 10-16.
14 Detroit. "DETROIT DD8" Available online: https://demanddetroit.com/engines/dd8/
436
-------
15 Ding, C., Roberts, L., Fain, D., Ramesh, A.K., Shaver, G.M., McCarthy, J., et al.(2015). "Fuel
efficient exhaust thermal management for compression ignition engines via cylinder deactivation
and flexible valve actuation." Int.J.Eng. Res. doi:10.1177/1468087415597413.
16 Neely, G., Sharp, C., Pieczko, M., and McCarthy, J., (2020) "Simultaneous NOX and C02
Reduction for Meeting Future CARB Standards Using a Heavy-Duty Diesel CDA-NVH
Strategy," SAE Int. J. Engines 13(2).
17 McDonald, Joseph. "Engine Modeling of LIVC for Heavy-duty Diesel Exhaust Thermal
Management at Light-load Conditions" Memorandum to Docket EPA-HQ-OAR-2019-0055.
November 21, 2019.
18 Farrell, L., Frazier, T., Younkins, M., Fuerst, J. (2020) Diesel Dynamic Skip Fire (dDSFTM):
Simultaneous C02 and NOx Reduction. Proceedings of the of the 41st International Vienna
Motor Symposium
19 Hamedi, M., Tsolakis, A., and Herreros, J., "Thermal Performance of Diesel Aftertreatment:
Material and Insulation CFD Analysis," SAE Technical Paper 2014-01-2818, 2014,
doi:l 0.4271/2014-01-2818.
20 U.S. EPA (2014) Chapter 1 - Vehicle Program Technological Feasibility. Control of Air
Pollution from Motor Vehicles: Tier 3 Motor Vehicle Emission and Fuel Standards Final Rule -
Regulatory Impact Analysis. EPA Document No. EPA-420-R-14-005.
21 Harris, T. M., Mc Pherson, K., Rezaei, R., Kovacs, D., Rauch, H., & Huang, Y. (2019).
Modeling of close-coupled SCR concepts to meet future cold start requirements for heavy-duty
engines. SAE Technical Paper No. 2019-01-0984.
22 Singh, N., Adelman, B., Malagari, S., & Hickey, K. (2019). Ultra-Low NOX Emission
Prediction for Heavy Duty Diesel Applications Using a Map-Based Approach. SAE Technical
Paper No. 2019-01-0987.
23 Zavala, B., Sharp, C., Neely, G., and Rao, S. (2020) CARB Low NOX Stage 3 Program -
Aftertreatment Evaluation and Down Selection, SAE Technical Paper 2020-01-1402,
doi: 10.4271/2020-01-1402.
24 U.S. EPA. "Crankcase Emissions for 2010+ Model Year Heavy-Duty Diesel Trucks - Test
Report." Docket EPA-HQ-OAR-2019-0055. April 2020.
25 Khalek, I., Bougher, T., Merritt, P. "Phase 1 of the Advanced Collaborative Emissions
Study." CRC Report: ACES Phase 1. Alpharetta, GA: Coordinating Research Council. 2009.
26 CARB; On-Road Advanced Technology Demonstration Project, Page 2, September 26, 2017.
https://www.arb.ca.gov/msprog/aqip/solicitations/092617onroadatdapplicationsfyl617.pdf
27 South Coast Air Quality Management District; Execute Contract to Develop and Demonstrate
Near-Zero Emission Opposed Piston Engine, December 2, 2017.
http://www.aqmd.gov/docs/default-source/Agendas/Governing-Board/2017/2017-decl-006.pdf
437
-------
28 Achates Power; Opposed Piston Engine HD Demo Program, September 26, 2019.
https://ww3.arb.ca.gov/msprog/hdlownox/files/workgroup_20190926/guest/10_achates_hd_dem
o_program.pdf
29 Achates Power Cost Study White Paper, March 2020, Achates Power Cost Study White
Paper_March 2020.pdf
30 CARB; Opposed Piston Engine Class 8 Heavy-Duty On-Road Demonstration, January 26,
2018. https://www.arb.ca.gov/msprog/lct/pdfs/opposedpiston.pdf
31 Samenfink, W., Albrodt, H., Frank, M., Gesk, M., Melsheimer, A., Thurso, J., Matt, M.
"Strategies to Reduce HC-Emissions During the Cold Starting of a Port Fuel Injected Engine."
SAE Technical Paper 2003-01-0627.
32 Yi, J., Wooldridge, S., Coulson, G., Hilditch, J., Iver, C., Moilanen, P., Papaioannou, G.,
Reiche, D., Shelby, M., VanDerWege, B., Weaver, C., Xu, Z., Davis, G., Hinds, B., Schamel, A.
"Development and Optimization of the Ford 3.5L V6 EcoBoost Combustion System." SAE
Technical Paper 2009-01-1494.
33 Choi, M., Sun, H., Lee, C., Myung, C., Kim, W., Choi, J. "The Study of HC Emission
Characteristics and Combustion Stability with Spark Timing Retard at Cold Start in Gasoline
Engine Vehicle." SAE Technical Paper 2000-01-0182.
34 Eng, James A. "The Effect of Spark Retard on Engine-out Hydrocarbon Emissions." SAE
Technical Paper 2005- 01-3867.
35 Hattori, M., Inoue, T., Mashiki, Z., Takenaka, A., Urushihata, H., Morino, S., Inohara, T.
"Development of variable Valve Timing System Controlled by Electric Motor." SAE Technical
Paper 2008-01-1358.
36 Ball, D., Zammit, M., Wuttke, J., Buitrago, C. "Investigation of LEV-III Aftertreatment
Designs." SAE Technical Paper 2011-01-0301.
37 Serrano, D., Lavy, J., Kleeman, A., Zinola, S., Dumas, J., Le Mirronet, S., Heitz, D. "Post
Oxidation Study During Secondary Exhaust Air Injection for Fast Catalyst Light-Off" SAE
Technical Paper 2009-01-2706.
38 Lee, D., Heywood, J. "Effects of Secondary Air Injection During Cold Start of SI Engines."
SAE Technical Paper 2010-01-2124.
39 Schell, T., Holle, C., Wunsch, R. et al. (2020) "M 254 - the Mercedes-Benz 4-Cylinder
Gasoline Engine of the Future." Proceedings of the 41st Vienna Motor Symposium.
40 81 FR 73478, October 2, 2016.
41 U.S. EPA. "Heavy-Duty Highway Gasoline and Diesel Certification Data (Model Years:
2015 - Present)". Available online: https://www.epa.gov/sites/production/files/2020-01/heavy-
duty-gas-and-diesel-engines-2015-present.xlsx. Accessed June 2020.
42 SGS-Aurora, Eastern Research Group, "Light Heavy-Duty Gasoline Vehicle Evaporative
Emissions Testing." EPA-420-R-19-017. December 2019.
438
-------
43 U.S. Environmental Protection Agency. "Summary of "Light Heavy-Duty Gasoline Vehicle
Evaporative Emissions Test Program"" EPA-420-S-19-002. December 2019.
44 Graboski, M.S. and McCormick, R.L., "Combustion of Fat and Vegetable Oil Derived Fuels
in Diesel Engines," Progress in Energy and Combustion Science, 24 125-163, 1998.
45 Chaves, Eduardo S. et al; Metals and phosphorus determination in vegetable seeds used in the
production of biodiesel by ICP OES and ICP-MS; Microchemical Journal 96 (2010) 71 - 76.
46 Alleman, T.L., McCormick, R.L., "Results of the 2007 B100 Quality Survey," NREL/TP-
540-42787, Golden, CO: National Renewable Energy Laboratory, March 2008.
47 Alleman, Theresa L., Quality Parameters and Chemical Analysis for Biodiesel Produced in
the United States in 2011; NREL/TP-5400-57662; March 2013.
48 CSN EN 14538 Fat and oil derivatives - Fatty acid methyl ester (FAME) - Determination of
Ca, K, Mg and Na content by optical emission spectral analysis with inductively coupled plasma
(ICP OES).
49 ASTM, 2018. "Standard Specification for Biodiesel Fuel (B100) Blend Stock for Distillate
Fuels", American Society for Testing and Materials, D6751-18.
50 ASTM, 2019. "Standard Specification for Diesel Fuel Oil, Biodiesel Blend (B6 to B20)",
American Society for Testing and Materials, D7467-19.
51 ASTM, 2019. "Standard Specification for Diesel Fuel", American Society for Testing and
Materials, D975 - 19b.
52 Jaaskelainen, H. Biodiesel Standards & Properties.
https://www.dieselnet.com/tech/fuel_biodiesel_std.php
53 Sappok, A., Wong, V., 2007. "Impact of Biodiesel on Ash Emissions and Lubricant
Properties Affecting Fuel Economy and Engine Wear Comparison with Conventional Diesel
Fuel", US DOE, 13th Diesel Engine-Efficiency and Emissions Research (DEER) Conference,
Detroit, MI, August 2007,
https://wwwl.eere.energy.gov/vehiclesandfuels/pdfs/deer_2007/session5/deer07_sappok.pdf
54 Williams, Aaron; Impact of Fuel Metal Impurities on the Durability of a Light-Duty Diesel
Aftertreatment System; SAE 2013-01-0513; 4/08/2013.
55 Dou, D. and Balland, J., "Impact of Alkali Metals on the Performance and Mechanical
Properties of NOX Adsorber Catalysts," SAE Technical Paper 2002-01-0734, 2002,
doi: 10.4271/2002-01-0734.
56 Cavataio, G., Jen, H., Dobson, D., and Warner, J., "Laboratory Study to Determine Impact of
Na and K Exposure on the Durability of DOC and SCR Catalyst Formulations," SAE Technical
Paper 2009-01-2823, 2009, doi: 10.4271/2009-01-2823.
57 McCormick, Robert; Biodiesel Performance with Modern Engines, CRADA #: CRD-05-153;
March 2016
439
-------
58 Schroder, Jorg; Accelerated performance and durability test of the exhaust aftertreatment
system by contaminated biodiesel; International Journal of Engine Research; Vol. 18 (10) 1067-
1076, 2017
59 Brookshear, Daniel W.; Investigation of the effects of biodiesel-based Na on emission control
components; Catalysis Today 184 (2012) 205-2018
60 Iojoiu, E.; Biofuel Impact on Diesel Engine After-Treatment: Deactivation Mechanisms and
Soot Reactivity; Emission Control Science and Technology; (2018) 4:15-32; December 18 2017.
61 Lance, Michael; Evaluation of Fuel-Borne Sodium Effects on a DOC-DPF-SCR Heavy-Duty
Engine Emission Control System: Simulation of Full-Useful Life; SAE 2016-01-2322.
62 Kienkas, Liene; Effect of Biofuel Impurities on the Diesel Oxidation Catalyst; Degree Project
Stockholm 2017.
63 Guo, Jing; Effects of Biodiesel Blending on Exhaust Emissions; Dissertation; June 2, 2011.
64 Williams, Aaron; Impact of Biodiesel Impurities on the Performance and Durability of DOC,
DPF and SCR Technologies; SAE 2011-01-1136; 04/12/2011.
65 Alleman, T.L., 2006 B100 Quality Survey Results; Milestone Report NREL/TP-540-41549;
May 2007.
66 Alleman, Teresa L., Analysis of Biodiesel Blend Samples Collected in the United States in
2008; NREL/TP540-46592.
67 Alleman, Teresa L., Metals Analysis of Biodiesel Blends; Technical Report NREL/TP-5400-
72341; May 2019
68 Alleman, Teresa L. Assessment of BQ-9000 Biodiesel Properties for 2017; NREL/TP-5400-
75795; January 2020.
69 Alleman, Teresa L. Assessment of BQ-9000 Biodiesel Properties for 2018; NREL/TP-5400-
75796; January 2020.
70 ASTM, 2015. "Trace Metals in Organics by ICP-OES", UOP-389-15.
71 Recker, Alissa, "Fuel Quality Impacts on Aftertreatment and Engine;" Daimler Trucks, July
29, 2019.
72 ASTM, 2016. " Standard Test Method for Determination of Trace Elements in Middle
Distillate Fuels by Inductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES)",
American Society for Testing and Materials, D7111-16.
73 CARB response to comments to ANPRM February 24, 2020.
74 NACFE (2019) "Guidance Report: Viable Class 7/8 Electric, Hybrid and Alternative Fuel
Tractors", available online at: https://nacfe.org/downloads/viable-class-7-8-alternative-vehicles/
75 National Academies of Sciences, Engineering, and Medicine 2020. Reducing Fuel
Consumption and Greenhouse Gas Emissions of Medium- and Heavy-Duty Vehicles, Phase
440
-------
Two: Final Report. Washington, DC: The National Academies Press.
https://doi.org/10.17226/25542.
76 Thorton et. al (2015) Data Collection, Testing, and Analysis of Hybrid Electric Trucks and
Buses Operating in California Fleets Final Report", available online at:
https://www.nrel.gov/docs/fyl5osti/62009.pdf
77 Hino Trucks. "Hinol95h Hybrid." Accessed July 24, 2020. https://www.hino.com/hino-
trucks-hino-195h. html
78 NTEA 2019 Fleet Purchasing Outlook Survey
79 MECA 2020, "Technology Feasibility for Heavy-Duty Diesel Trucks in Achieving 90%
Lower NOX Standards in 2027", available online at:
http://www.meca.org/resources/MECA_2027_Low_NOX_White_Paper_FINAL.pdf
80 Yuan, Xiao-Zi, and Haijiang Wang. "PEM fuel cell fundamentals." In PEM fuel cell
electrocatalysts and catalyst layers, pp. 1-87. Springer, London, 2008.
81 Sharer, P., & Rousseau, A. (2013). Benefits of fuel cell range extender for medium-duty
vehicle applications. World Electric Vehicle Journal, 6(2), 452-463.
82 Bergmann D. et al. (2017) Range extender systems for electric drivetrains in medium-duty
distribution vehicles. In: Liebl J., Beidl C. (eds) Internationaler Motorenkongress 2017.
Proceedings. Springer Vieweg, Wiesbaden
83 Sim, K., Vijayagopal, R., Kim, N., & Rousseau, A. (2019). Optimization of Component
Sizing for a Fuel Cell-Powered Truck to Minimize Ownership Cost. Energies, 12(6), 1125.
84 Smith, D. et al. (2019) "Medium- and Heavy-Duty Electrification An Assessment of
Technology and Knowledge Gaps". Oak Ridge National Laboratory and National Renewable
Energy Laboratory. ORNL/SPR-2020/7
85 Sharpe, B. (2019) "Zero-Emission Tractor-Trailers in Canada". ICCT Working Paper, April.
Available online at: https://theicct.org/publications/zero-emission-tractor-trailers-canada.
86 Electrify America, "Charging With Us", accessed 2-19-2020; available online at:
https://www.electrifyamerica.com/charging-with-us
87 U.S. Department of Energy Alternative Fuels Data Center (AFDC), "Developing
Infrastructure to Charge Plug-In Electric Vehicles",
https://afdc.energy.gov/fuels/electricity_infrastructure.html (accessed 2-27-20)
88 European Automobile Manufacturers Association (ACEA) (2017). Charging of Electric
Buses - ACEA Recommendations. Accessed on the internet on 5/27/2020 at the following URL:
https://www.oppcharge.org/dok/ACEA_charging_electric_buses.pdf
89 SAE (2020). J3105_202001 - Electric Vehicle Power Transfer System Using Conductive
Automated Connection Devices. Accessed on the internet on 5/27/2020 at the following URL:
https://www. sae.org/standards/content/j 3105_202001/
441
-------
90 Monios, J., & Bergqvist, R. (2019). The transport geography of electric and autonomous
vehicles in road freight networks. Journal of Transport Geography, 80, 102500.
91 McKinsey (2017) "New reality: electric trucks and their implications on energy demand";
available online at: https://www.mckinsey.com/industries/oil-and-gas/our-insights/a-new-reality-
electric-trucks.
92 NACFE (2018) Guidance Report: Electric Trucks - Where They Make Sense; available
online at: https://nacfe.org/report-library/guidance-reports/.
93 NACFE (2019) "More Regional Haul: An Opportunity for Trucking?", available online at
http s ://nacfe. org/regi onal -haul /
94 Energy Information Association (2018) "Annual Energy Outlook; Table 50: Freight
Transportation Energy Use", available at:
https://www.eia.gov/outlooks/aeo/data/browser/#/?id=58-AE02018&sourcekey=0
95 Jadun, et al. (2017) "Electrification Futures Study: End-Use Electric Technology Cost and
Performance Projections through 2050". Golden, CO: National Renewable Energy Laboratory.
NREL/TP-6A20-70485. https://www.nrel.gov/docs/fyl8osti/70485.pdf.
96 Sonntag, Darrell. Population and Activity of Onroad Vehicles in MOVES CTI NPRM.
Attachment to Memorandum to Docket EPA-HQ-OAR-2019-0055: "Updates to MOVES for
Emissions Analysis of the HD 2027 NPRM." May 2021.
97 NACFE (2018) "Guidance Report: Medium-Duty Electric Trucks Cost of Ownership",
available online at: https://nacfe.org/future-technology/medium-duty-electric-trucks-cost-of-
ownership/
98 Tigue, K. (2019) "U.S. Electric Bus Demand Outpaces Production as Cities Add to Their
Fleets" Inside Climate News, November 14.
https://insideclimatenews.org/news/14112019/electric-bus-cost-savings-health-fuel-charging
99 USC (2019) "Ready for Work: Now Is the Time for Heavy-Duty Electric Vehicles";
www.ucsusa.org/resources/ready-work
100 CALSTART (2020): Drive to Zero's Zero-emission Technology Inventory (ZETI) Tool
Version 5.5. Available online at https://globaldrivetozero.org/tools/zero-emission-technology-
inventory/
101 ICCT (2020) "Race to Zero: How manufacturers are positioned for zero emission
commercial trucks and buses in North America"; available online at:
https://theicct.org/publications/canada-race-to-zero-oct2020.
102 ICCT (2019) "Estimating the infrastructure needs and costs for the launch of zero-emissions
trucks"; available online at: https://theicct.org/publications/zero-emission-truck-infrastructure.
103 Phadke, A., et. al. (2021) " Why Regional and Long-Haul Trucks are Primed for
Electrification Now"; available online at: https://eta-
publications.lbl.gov/sites/default/files/updated_5_final_ehdv_report_033121.pdf.
442
-------
104 Nadel, S. and Junga, E. (2020) "Electrifying Trucks: From Delivery Vans to Buses to 18-
Wheelers". American Council for an Energy-Efficient Economy White Paper, available online
at: https://aceee.org/white-paper/electrifying-trucks-delivery-vans-buses-18.
105 Fisher, J. (2019) "Volvo's First Electric VNR Ready for the Road." Fleet Owner, September
17. www.fleetowner.com/blue-fleets/volvo-s-first-electric-vnr-ready-road.
106 Gnaticov, C. 2018. "Nikola One Hydrogen Electric Semi Hits the Road in Official Film."
Carscoops, Jan. 26. www.carscoops.com/2018/Ol/nikola-one-hydrogen-electric-semi-hits-road-
official-film/.
107 U.S. Department of Energy Alternative Fuels Data Center (AFDC), "Alternative Fueling
Stations Locator," https://afdc.energy.g0v/stations#/f1nd/nearest (accessed 2-19-20)
108 O'Dell, J. (2019) "Nikola Details Ambitious Plan for Hydrogen and Service Network"
Trucks.com, April 18. https://www.trucks.com/2019/04/22/nikola-ambitious-plan-hydrogen-
service-network/
109 ICCT. "Transitioning to Zero-Emission Heavy-Duty Freight Vehicles". September 2017.
Available online: https://theicct.org/publications/transitioning-zero-emission-heavy-duty-freight-
vehicles
110 ICCT. "Zero-emission tractor-trailers in Canada". April 2019. Available online:
https://theicct.org/publications/zero-emission-tractor-trailers-canada
111 CARB. "Notice of Decision: Advanced Clean Truck Regulation." June 2020. Available
online at: https://ww3.arb.ca.gov/regact/2019/act2019/nod.pdf
For more information on this proposed rulemaking in California see:
https://ww2.arb. ca.gov/rulemaking/2019/advancedcleantrucks?utm_medium=email&utm_source
=govdelivery.
112 CARB. "Appendix A Proposed Regulation Order" Advanced Clean Truck Regulation. May
2020. Available online at: https://ww3.arb.ca.gov/regact/2019/act2019/30dayatta.pdf (accessed
July 24, 2020.
113 "Multi-state Medium- and Heavy-Duty Zero Emission Vehicle Memorandum of
Understanding" July, 13 2020. Available online at: https://www.nescaum.org/topics/zero-
emission-vehicles.
114 81 FR 73478, October 25, 2016.
115 See 40 CFR 1036.630 and 1037.550.
116 86 FR 34321, June 29, 2021.
117 "OEM perspective - Meeting EPA/NHTSA GHG/Efficiency Standards", 7th Integer
Emissions Summit USA 2014, Volvo Group North America.
118 Duran, A., Kotz, A., Thorton, M., and Kelly, K. NREL Low Load Engine Test Cycle
Development Update Presentation. February 16,2018.
443
-------
119 Heavy-Duty Low NOx Program Workshop - Low Load Cycle Development Presentation.
January 23, 2019 https://www.arb.ca.gov/msprog/hdlownox/hdlownox.htm
120 Heavy-Duty Low NOx Program Workshop - Low Load Cycle Presentation. September 26,
2019 https://www.arb.ca.gov/msprog/hdlownox/hdlownox.htm
121 Sharp, C. Heavy-Duty Engine Low-Load Emission Control Calibration, Low-Load Test
Cycle Development, and Evaluation of Engine Broadcast Torque and Fueling Accuracy During
Low-Load Operation Low-Load Operation Low NOx Demonstration Program - Stage 2,
prepared for the California Air Resources Board, December 20, 2019.
122 Pondicherry, R., Besch, M., Thiruvengadam, A., and Carder, A. A Vehicle Activity-based
Windowing approach to evaluate real-world NOx emissions from Modern Heavy-duty Diesel
Trucks, Atmospheric Environment, Vol. 247, 2021, 118169,
http://www.sciencedirect.com/science/article/pii/S1352231020308992
123 Hamady, Fakhri, Duncan, Alan. "A Comprehensive Study of Manufacturers In-Use Testing
Data Collected from Heavy-Duty Diesel Engines Using Portable Emissions Measurement
System (PEMS)". 29th CRC Real World Emissions Workshop, March 10 -13, 2019.
124 Sandhu, Gurdas; Sonntag, Darrell; Sanchez, James. 2018. Identifying Areas of High NOx
Operation in Heavy-Duty Vehicles, 28th CRC Real-World Emissions Workshop, March 18-21,
2018, Garden Grove, California, USA
125 Sonntag, Darrell. Exhaust Emission Rates for Heavy-Duty Onroad Vehicles in
MOVES CTI NPRM. Attachment to Memorandum to Docket EPA-HQ-OAR-2019-0055:
"Updates to MOVES for Emissions Analysis of the HD 2027 NPRM." May 2021.
126 COMMISSION REGULATION (EU) No 582/2011, May 25, 2011. Available online:
https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:02011R0582-
20180118&from=EN.
127 COMMISSION REGULATION (EU) 2018/932, June 29, 2018. Available online:
https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32018R0932&from=EN.
128 Rodriguez, F.; Posada, F. "Future Heavy-Duty Emission Standards An Opportunity for
International Harmonization". The International Council on Clean Transportation. November
2019. Available online:
https ://theicct. org/ sites/default/files/publications/Future%20_HD V_standards_opportunity_2019
1125.pdf
129 SGS-Aurora, Eastern Research Group, "Light Heavy-Duty Gasoline Vehicle Evaporative
Emissions Testing." EPA-420-R-19-017. December 2019.
130 U.S. Environmental Protection Agency. "Summary of "Light Heavy-Duty Gasoline Vehicle
Evaporative Emissions Test Program"" EPA-420-S-19-002. December 2019.
131 40 CFR 86.1829-01(b)(5).
132 ISO/SAE 21434, "Road Vehicles - Cybersecurity Engineering", SAE International, February
12, 2020.
444
-------
133 Neely, G., Sharp, C., Pieczko, M., and McCarthy, J., (2020) Simultaneous NOX and C02
Reduction for Meeting Future CARB Standards Using a Heavy-Duty Diesel CDA-NVH
Strategy. SAE Int. J. Engines 13(2).
134 Morris, A., & McCarthy, J. (2020). The Effect of Heavy-Duty Diesel Cylinder Deactivation
on Exhaust Temperature, Fuel Consumption, and Turbocharger Performance up to 3 bar BMEP.
SAE Technical Paper No. 2020-01-1407.
135 Zavala, B., Sharp, C., Neely, G., & Rao, S. (2020). CARB Low NO X Stage 3 Program-
Aftertreatment Evaluation and Down Selection. SAE Technical Paper No. 2020-01-1402.
136 Rao, S., Sarlashkar, J., Rengarajan, S., Sharp, C., & Neely, G. (2020). A Controls Overview
on Achieving Ultra-Low NOx SAE Technical Paper No. 2020-01-1404.
137 Neely, G. D., Sharp, C., & Rao, S. (2020). CARB Low NOx Stage 3 Program-Modified
Engine Calibration and Hardware Evaluations. SAE Technical Paper No. 2020-01-0318.
138 Sharp, C. (2020). Further Development and Validation of Technologies to Lower Oxides of
Nitrogen Emissions from Heavy-Duty Vehicles. Southwest Research Institute. Final Report
CARB Contract 16MSC-010.
139 Eakle, S., & Bartley, G. (2014). The DAAAC Protocol™ for Diesel Aftertreatment System
Accelerated Aging. In Proceedings of the Emissions 2014 Conference.
140 Nagar, P., Szailer, T., & Webb, C. (2013). Comparison of SCR Catalyst Performance on
RMC SET Emission Cycle between an Engine and a High Flow Burner Rig. SAE Technical
Paper No. 2013-01-1070.
141 Sanchez, James. Memorandum to Docket: EPA-HQ-OAR-2019-0055. Test Results from
EPA Diesel Demonstration. February 10, 2022.
142 Dallman, T., Posada, F., Bandivadekar, A. 2018. "Costs of Emission Reduction Technology
for Diesel Engines Used in Non-Road Vehicles and Equipment." International Council on Clean
Transportation.
143 McDonald, Joseph. Memorandum to Docket: EPA-HQ-OAR-2019-0055. Diesel Exhaust
Aftertreatment Costs Derived from MY2019 Engine Certification Data. July 1, 2021.
144 Johnson Matthey Precious Metals Management Price Charts, last accessed on the internet on
March 3, 2020 at the following URL: http://www.platinum.matthey.com/prices/price-charts
145 Mamidanna, S. 2021. Heavy-Duty Engine Valvetrain Technology Cost Assessment. U.S.
EPA Contract with FEV North America, Inc., Contract No. 68HERC19D0008, Task Order No.
68HERH20F0041. Submitted to the Docket.
146 Mamidanna, S. 2021. Heavy-Duty Engine Valvetrain Technology Cost Assessment - Peer
Review Responses. Contract No. 68HERC19D0008, Task Order No. 68HERH20F0041,
Submitted to the Docket.
445
-------
147 Eastern Research Group, Inc. 2021. "External Peer Review of Heavy-Duty Engine
Valvetrain Technology Cost Assessment". Contract No. 68HE0C18C0001, Work Assignment 2-
05. Submitted to the Docket
148 Mamidanna, S. 2021. Heavy-Duty Vehicles Aftertreatment Systems Cost Assessment.
Submitted to the Docket.
149 69 Federal Register at 39126. June 29, 2004. See https://www.govinfo.gov/content/pkg/FR-
2004-06-29/pdf/04-l 1293.pdf.
150 Mitchell, George, "EPA's Medium Heavy-Duty Gasoline Vehicle Emissions Investigation".
February 2019.
151 U.S. EPA. "Heavy-Duty Highway Gasoline and Diesel Certification Data (Model Years:
2015 - Present)". Available online: https://www.epa.gov/sites/production/files/2020-01/heavy-
duty-gas-and-diesel-engines-2015-present.xlsx. Accessed June 2020.
152 Dallman, T., Posada, F., Bandivadekar, A. 2018. "Costs of Emission Reduction Technology
for Diesel Engines Used in Non-Road Vehicles and Equipment." International Council on Clean
Transportation.
153 Pasoda Sanchez, F., Bandivadekar, A., German, J. 2012. "Estimated Cost of Emissions
Reduction Technolgies for Liight-Duty Vehicles." International Council on Clean
Transportation.
154 Dallman, T., Posada, F., Bandivadekar, A. 2018. "Costs of Emission Reduction Technology
for Diesel Engines Used in Non-Road Vehicles and Equipment." International Council on Clean
Transportation.
155 Pasoda Sanchez, F., Bandivadekar, A., German, J. 2012. "Estimated Cost of Emissions
Reduction Technolgies for Liight-Duty Vehicles." International Council on Clean
Transportation.
156 U.S. EPA. Integrated Science Assessment (ISA) for Ozone and Related Photochemical
Oxidants (Final Report). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-
20/012, 2020.
157 U.S. EPA. Policy Assessment (PA) for the Review of the National Ambient Air Quality
Standards for Particulate Matter (Final Report, 2020). U.S. Environmental Protection Agency,
Washington, DC, EPA/452/R-20/002, 2020.
158 U.S. EPA. Integrated Science Assessment (ISA) for Particulate Matter (Final Report, 2019).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-19/188, 2019. Table 2-1.
159 U.S. EPA. Integrated Science Assessment (ISA) for Particulate Matter (Final Report, 2019).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-19/188, 2019. Table 2-1.
160 U.S. EPA. Integrated Science Assessment (ISA) for Particulate Matter (Final Report, 2019).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
446
-------
161 U.S. EPA. Policy Assessment (PA) for the Review of the National Ambient Air Quality
Standards for Particulate Matter (Final Report, 2020). U.S. Environmental Protection Agency,
Washington, DC, EPA/452/R-20/002, 2020.
162 U.S. EPA. (2009). Integrated Science Assessment for Particulate Matter (Final Report).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-08/139F.
163 U.S. EPA. Integrated Science Assessment (ISA) for Particulate Matter (Final Report, 2019).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
164 U.S. EPA. Integrated Science Assessment (ISA) for Particulate Matter (Final Report, 2019).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
165 U.S. EPA. Integrated Science Assessment (ISA) for Particulate Matter (Final Report, 2019).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
166 U.S. EPA. Policy Assessment (PA) for the Review of the National Ambient Air Quality
Standards for Particulate Matter (Final Report, 2020). U.S. Environmental Protection Agency,
Washington, DC, EPA/452/R-20/002, 2020.
167 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.
168 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 / recordi spl ay. cfm? dei d=218686.
169 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://cfpub.epa.gov/ncea/cfm/recordisplay. cfm?deid=54932.
170 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://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=29060. pp. 1-1 1-2.
171 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.
172 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.
173 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.
174 IARC [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]
447
-------
175 U.S. EPA. (2015) Summary of Results for the 2011 National-Scale Assessment.
http://www.epa.gov/sites/production/files/2015-12/documents/2011-nata-summary-results.pdf.
176 U.S. EPA (2018) Technical Support Document EPA's 2014 National Air Toxics
Assessment, https://www.epa.gov/national-air-toxics-assessment/2014-nata-assessment-results
177 U.S. EPA (2018) 2014 NAT A Summary of Results.
https://www.epa.gov/sites/production/files/2020-
07/documents/nata_2014_summary _of_results.pdf
178 U.S. EPA. (2000). Integrated Risk Information System File for Benzene. This material is
available electronically at:
https://cfpub. epa.gov/ncea/iris2/chemicalLanding. cfm?substance_nmbr=276.
179 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.
180 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.
181 U.S. EPA. (2000). Integrated Risk Information System File for Benzene. This material is
available electronically at:
https://cfpub. epa.gov/ncea/iris2/chemicalLanding. cfm?substance_nmbr=276.
182 International Agency for Research on Cancer (IARC, 2018. Monographs on the evaluation
of carcinogenic risks to humans, volume 120. World Health Organization - Lyon, France.
http://publications.iarc.fr/Book-And-Report-Series/Iarc-Monographs-On-The-Identification-Of-
Carcinogenic-Hazards-To-Humans/Benzene-2018.
183 NTP (National Toxicology Program). 2016. Report on Carcinogens, Fourteenth Edition.;
Research Triangle Park, NC: U.S. Department of Health and Human Services, Public Health
Service, https://ntp.niehs.nih.gov/go/rocl4
184 Aksoy, M. (1989). Hematotoxicity and carcinogenicity of benzene. Environ. Health
Perspect. 82: 193-197. EPA-HQ-OAR-2011-0135.
185 Goldstein, B.D. (1988). Benzene toxicity. Occupational medicine. State of the Art
Reviews. 3: 541-554.
186 Rothman, N., G.L. Li, M. Dosemeci, W.E. Bechtold, G.E. Marti, Y.Z. Wang, M. Linet, L.Q.
Xi, W. Lu, M.T. Smith, N. Titenko-Holland, L.P. Zhang, W. Blot, S.N. Yin, and R.B. Hayes.
(1996). Hematotoxicity among Chinese workers heavily exposed to benzene. Am. J. Ind. Med.
29: 236-246.
187 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
448
-------
electronically at
https://cfpub.epa.gov/ncea/iris/iris_documents/documents/toxreviews/0276tr.pdf.
188 Qu, O.; Shore, R.; Li, G.; Jin, X.; Chen, C.L.; Cohen, B.; Melikian, A.; Eastmond, D.;
Rappaport, S.; Li, H.; Rupa, D.; Suramaya, R.; Songnian, W.; Huifant, Y.; Meng, M.; Winnik,
M.; Kwok, E.; Li, Y.; Mu, R.; Xu, B.; Zhang, X.; Li, K. (2003). HEI Report 115, Validation &
Evaluation of Biomarkers in Workers Exposed to Benzene in China.
189 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.
190 Lan, Qing, Zhang, L., Li, G., Vermeulen, R., et al. (2004). Hematotoxically in Workers
Exposed to Low Levels of Benzene. Science 306: 1774-1776.
191 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.l 13.
192 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.
193 EPA. Integrated Risk Information System. Formaldehyde (CASRN 50-00-0)
https://cfpub.epa.gov/ncea/iris2/chemicalLanding.cfm?substance_nmbr=419.
194 NTP (National Toxicology Program). 2016. Report on Carcinogens, Fourteenth Edition.;
Research Triangle Park, NC: U.S. Department of Health and Human Services, Public Health
Service, https://ntp.niehs.nih.gov/go/rocl4.
195 IARC Monographs on the Evaluation of Carcinogenic Risks to Humans Volume 88 (2006):
Formaldehyde, 2-Butoxyethanol and l-tert-Butoxypropan-2-ol.
196 IARC Monographs on the Evaluation of Carcinogenic Risks to Humans Volume 100F
(2012): Formaldehyde.
197 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.
198 Hauptmann, M.; Lubin, J. H.; Stewart, P. A.; Hayes, R. B.; Blair, A. 2004. Mortality from
solid cancers among workers in formaldehyde industries. American Journal of Epidemiology
159: 1117-1130.
199 Beane Freeman, L. E.; Blair, A.; Lubin, J. H.; Stewart, P. A.; Hayes, R. B.; Hoover, R. N.;
Hauptmann, M. 2009. Mortality from lymphohematopoietic malignancies among workers in
formaldehyde industries: The National Cancer Institute cohort. J. National Cancer Inst. 101: 751-
761.
200 Pinkerton, L. E. 2004. Mortality among a cohort of garment workers exposed to
formaldehyde: an update. Occup. Environ. Med. 61: 193-200.
449
-------
201 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.
202 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.
203 ATSDR. 1999. Toxicological Profile for Formaldehyde, U.S. Department of Health and
Human Services (HHS), July 1999.
204 ATSDR. 2010. Addendum to the Toxicological Profile for Formaldehyde. U.S. Department
of Health and Human Services (HHS), October 2010.
205 IPCS. 2002. Concise International Chemical Assessment Document 40. Formaldehyde.
World Health Organization.
206 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://cfpub. epa.gov/ncea/iris_drafts/recordisplay. cfm?deid=223614.
207 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.
208 U.S. EPA (2018). See
https://cfpub.epa.gov/ncea/iris2/chemicalLanding.cfm?substance_nmbr=419.
209 U.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
https://cfpub. epa.gov/ncea/iris2/chemicalLanding. cfm?substance_nmbr=290.
210 U.S. EPA (1991). Integrated Risk Information System File of Acetaldehyde. This material
is available electronically at
https://cfpub. epa.gov/ncea/iris2/chemicalLanding. cfm?substance_nmbr=290.
211 NTP (National Toxicology Program). 2016. Report on Carcinogens, Fourteenth Edition.;
Research Triangle Park, NC: U.S. Department of Health and Human Services, Public Health
Service, https://ntp.niehs.nih.gov/go/rocl4.
212 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.
213 U.S. EPA (1991). Integrated Risk Information System File of Acetaldehyde. This material
is available electronically at
https://cfpub. epa.gov/ncea/iris2/chemicalLanding. cfm?substance_nmbr=290.
450
-------
214 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
https://cfpub.epa.gov/ncea/iris2/chemicalLanding.cfm?substance_nmbr=364.
215 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.
216 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.
217 California OEHHA, 2014. TSD forNoncancer RELs: Appendix D. Individual, Acute, 8-
Hour, and Chronic Reference Exposure Level Summaries. December 2008 (updated July 2014).
https://oehha.ca.gov/media/downloads/crnr/appendixdlfinal.pdf
218 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
https://cfpub. epa.gov/ncea/iris_drafts/recordisplay. cfm?deid=56434.
219 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
https://cfpub. epa.gov/ncea/iris_drafts/recordisplay. cfm?deid=56434.
220 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
https://cfpub. epa.gov/ncea/iris_drafts/recordisplay. cfm?deid=56434.
221 Oak Ridge Institute for Science and Education. (2004). External Peer Review for the IRIS
Reassessment of the Inhalation Carcinogenicity of Naphthalene. August 2004.
http://cfpub. epa.gov/ncea/cfm/recordisplay. cfm?deid=84403.
222 U.S. EPA. (2018) See:
https://cfpub. epa.gov/ncea/iris2/chemicalLanding. cfm?substance_nmbr=436.
223 NTP (National Toxicology Program). 2016. Report on Carcinogens, Fourteenth Edition.;
Research Triangle Park, NC: U.S. Department of Health and Human Services, Public Health
Service, https://ntp.niehs.nih.gov/go/rocl4.
224 International Agency for Research on Cancer (IARC). (2002). Monographs on the
Evaluation of the Carcinogenic Risk of Chemicals for Humans. Vol. 82. Lyon, France.
225 U. S. EPA. (1998). Toxicological Review of Naphthalene, Environmental Protection
Agency, Integrated Risk Information System, Research and Development, National Center for
451
-------
Environmental Assessment, Washington, DC. This material is available electronically at
https://cfpub. epa.gov/ncea/iris_drafts/recordisplay. cfm?deid=56434.
226 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
https://cfpub. epa.gov/ncea/iris_drafts/recordisplay. cfm?deid=56434.
227 U.S. EPA Integrated Risk Information System (IRIS) database is available at:
www.epa.gov/iris.
228 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],
229 Clark, L.P.; Millet, D.B.; Marshall, J.D. (2017) Changes in Transportation-Related Air
Pollution Exposures by Race-Ethnicity and Socioeconomic Status: Outdoor Nitrogen Dioxide in
the United States in 2000 and 2010. Environmental Health Perspectives 125:9 CID: 097012
https://doi.org/10.1289/EHP959.
230 Alotaibi, R.; Bechle, M.; Marshall, J.D.; et al. (2019) Traffic related air pollution and the
burden of childhood asthma in the contiguous United States in 2000 and 2010. Environ
International 127: 858-867. https://doi.Org/10.1016/j.envint.2019.03.041
231 Rowangould, G.M. 2013. A census of the US near-roadway population: Public health and
environmental justice considerations. Transportation Research Part D (2013) 59-67.
http://dx.doi.Org/10.1016/j.trd.2013.08.003
232 US EPA (2021). Estimation of Population Size and Demographic Characteristics among
People Living Near Truck Routes in the Conterminous United States. Memorandum to the
Docket.
233 Rowangould, Gregory M., 2013. A census of the US near-roadway population: Public health
and environmental justice considerations. Transportation Research Part D.
http://dx.doi.Org/10.1016/j.trd.2013.08.003
234 US EPA (2021). Estimation of Population Size and Demographic Characteristics among
People Living Near Truck Routes in the Conterminous United States. Memorandum to the
Docket.
235 American Housing Survey, https://www.census.gov/programs-surveys/ahs.html
236 CIA world factbook. https://www.cia.gov/the-world-factbook/
237 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.
452
-------
238 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.
239 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.0b013e3181dbecel.
240 Anderson et al, 2013. Long-term exposure to air pollution and the incidence of asthma:
meta-analysis of cohort studies. Air Qual Atmos Health, doi:10.1007/sll869-011-0144-5.
241 Salam et al, 2008. Recent evidence for adverse effects of residential proximity to traffic
sources on asthma. Current Opinion in Pulmonary Medicine 2008, 14:3-8.
242 Braback, L., Forsberg, B. Does traffic exhaust contribute to the development of asthma and
allergic sensitization in children: findings from recent cohort studies. Environ Health 8, 17
(2009). https://doi.org/10.1186/1476-069X-8-17
243 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.
244 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.
245 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. Eur Respir J 15:
716-724.
246 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.
247 Boothe, VL.; Boehmer, T.K.; Wendel, A.M.; Yip, F.Y. (2014) Residential traffic exposure
and childhood leukemia: a systematic review and meta-analysis. Am J Prev Med 46: 413-422.
248 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.
249 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.
250 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 https://www.nap.edu/catalog/2097/protecting-visibility-in-national-
parks-and-wilderness-areas
453
-------
251 Hand et al. "Spatial and Temporal Variability of Haze and its Constituents in the United
States" - IMPROVE Report 0737-5352 (HERO #3121721).
252 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.
253 U.S. EPA. Integrated Science Assessment (ISA) for Particulate Matter (Final Report, 2019).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
254 Malm, William. Visibility - The Seeing of Near and Distant Landscape Features. Elsevier,
2016.
255 U.S. EPA. Integrated Science Assessment (ISA) for Particulate Matter (Final Report, 2019).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
256 See Section 169(a) of the Clean Air Act.
257 U.S. EPA. Integrated Science Assessment (ISA) for Particulate Matter (Final Report, 2019).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
258 73 FR 16486 (March 27, 2008).
259 U.S. EPA. Integrated Science Assessment (ISA) for Ozone and Related Photochemical
Oxidants (Final Report). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-
20/012, 2020.
260 73 FR 16492 (March 27, 2008).
261 73 FR 16493-16494 (March 27, 2008).
262 73 FR 16490/ 16497 (March 27, 2008).
263 U.S. EPA. Integrated Science Assessment (ISA) for Ozone and Related Photochemical
Oxidants (Final Report). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-
20/012, 2020.
264 U.S. EPA. Integrated Science Assessment (ISA) for Oxides of Nitrogen, Oxides of Sulfur
and Particulate Matter Ecological Criteria (Final Report). U.S. Environmental Protection
Agency, Washington, DC, EPA/600/R-20/278, 2020.
265 U.S. EPA. Integrated Science Assessment (ISA) for Particulate Matter (Final Report, 2019).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
266 U.S. EPA. Integrated Science Assessment (ISA) for Oxides of Nitrogen, Oxides of Sulfur
and Particulate Matter Ecological Criteria (Final Report). U.S. Environmental Protection
Agency, Washington, DC, EPA/600/R-20/278, 2020. See Figure IS-2.
267 Hutchison, R., and C.E. Kraft. 1994. "Hmong Fishing Activity and Fish Consumption."
Journal of Great Lakes Research 20(2):471-487.
454
-------
268 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.
269 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.
270 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.
271 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.
272 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.
273 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 (USDA).
274 Riddell et al, 2008. "The effect of HN03 gas on the lichen Ramalina menziesii." Flora, 203:
47-54.
275 Grantz DA; Garner JHB; Johnson DW. 2003. "Ecological effects of particulate matter."
Environ Int, 29: 213-239.
276 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.
277 SimcikM.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.
278 SimcikM.F., Eisenreich, S.J., GoldenK.A., et al. 1996. "AtmosphericLoading of
Polycyclic Aromatic Hydrocarbons to Lake Michigan as Recorded in the Sediments."
Environmental Science and Technology, 30, 3039-3046.
279 Arzavus K.M., DickhutR.M., and Canuel E.A. 2001. "Fate of Atmospherically Deposited
Polycyclic Aromatic Hydrocarbons (PAHs) in Chesapeake Bay." Environmental Science &
Technology, 35, 2178-2183.
280 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.
455
-------
281 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.
282 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.
283 U.S. EPA. Integrated Science Assessment (ISA) for Particulate Matter (Final Report, 2019).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
284 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.
285 U.S. EPA. (1991). Effects of organic chemicals in the atmosphere on terrestrial plants.
EPA/600/3-91/001.
286 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.
287 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.
288 Viskari E-L. (2000). Epicuticular wax of Norway spruce needles as indicator of traffic
pollutant deposition. Water, Air, and Soil Pollut. 121:327-337.
289 Ugrekhelidze D, F Korte, GKvesitadze. (1997). Uptake and transformation of benzene and
toluene by plant leaves. Ecotox. Environ. Safety 37:24-29.
290 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.
291 U.S. Global Change Research Program (USGCRP). 2018. Impacts, Risks, and Adaptation in
the United States: Fourth National Climate Assessment, Volume II [Reidmiller, D.R., C.W.
Avery, D.R. Easterling, K.E. Kunkel, K.L.M. Lewis, T.K. Maycock, and B.C. Stewart (eds.)].
U.S. Global Change Research Program, Washington, DC, USA, 1515 pp. doi:
10.7930/NCA4.2018.
292 Intergovernmental Panel on Climate Change (IPCC). 2022a. Chapter 2: Terrestrial and
Freshwater Ecosystems and their Services. In: Climate Change 2022: Impacts, Adaptation and
Vulnerability. Contribution of Working Group II to the Sixth Assessment Report of the
Intergovernmental Panel on Climate Change. [Parmesan, Morecroft, Trisurat, et al.] Cambridge
456
-------
University Press. In Press.
https://report.ipcc.ch/ar6wg2/pdf/IPCC AR6 WGII FinalDraftChapter02.pdf.
293 Intergovernmental Panel on Climate Change (IPCC). 2022b. Chapter 14: North America. In:
Climate Change 2022: Impacts, Adaptation and Vulnerability. Contribution of Working Group II
to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. [Hicke,
Lucatello, Mortsch, et al.] Cambridge University Press. In Press.
https://report.ipcc.ch/ar6wg2/pdf/IPCCAR6JWGII FinalDr aft Chapter 14.pdf.
294 U.S. Department of Defense (DOD), Office of the Undersecretary for Policy (Strategy,
Plans, and Capabilities). 2021. Department of Defense Climate Risk Analysis. Report Submitted
to National Security Council. https://media.defense.gOv/2021/Oct/21/20028 77353/-l/-l/0/DOD~
CLIMA TE-RISK-ANALYSIS-FINAL.PDF.
295 U.S. Department of Defense (DOD). 2014. Climate Change Adaptation Roadmap. Available
at: https://www.acq.osd.mil/eie/downloads/CCARprintjvForward e.pdf (accessed February 5,
2021).
296 Center for Climate and Security (CCS), The. 2018. Military Expert Panel Report Sea Level
Rise and the U.S. Military's Mission. Available at:
https://climateandsecurity.org/militaryexpertpanel2018/.
297 U.S. Global Change Research Program (USGCRP). 2016. The Impacts of Climate Change
on Human Health in the United States: A Scientific Assessment. Crimmins, A., J. Balbus, J.L.
Gamble, C.B. Beard, J.E. Bell, D. Dodgen, R.J. Eisen, N. Fann, M.D. Hawkins, S.C. Herring, L.
Jantarasami, D.M. Mills, S. Saha, M.C. Sarofim, J. Trtanj, and L. Ziska, Eds. U.S. Global
Change Research Program, Washington, DC, 312 pp.
298 U.S. Global Change Research Program (USGCRP). 2017. Climate Science Special Report:
Fourth National Climate Assessment, Volume I [Wuebbles, D.J., D.W. Fahey, K.A. Hibbard,
D.J. Dokken, B.C. Stewart, and T.K. May cock (eds.)]. U.S. Global Change Research Program,
Washington, DC, USA, 470 pp, doi: 10.7930/J0J964J6.
299 Intergovernmental Panel on Climate Change (IPCC). 2018. Global Warming of 1.5 °C. An
IPCC Special Report on the impacts of global warming of 1.5 °C above pre-industrial levels and
related global greenhouse gas emission pathways, in the context of strengthening the global
response to the threat of climate change, sustainable development, and efforts to eradicate
poverty [Masson-Delmotte, V., P. Zhai, H.-O. Po'rtner, D. Roberts, J. Skea, P.R. Shukla, A.
Pirani, W. Moufouma-Okia, C. Pe'an, R. Pidcock, S. Connors, J.B.R. Matthews, Y. Chen, X.
Zhou, M.I. Gomis, E. Lonnoy, T. Maycock, M. Tignor, and T. Waterfield (eds.)].
300 Intergovernmental Panel on Climate Change (IPCC). 2019a. Climate Change and Land: an
IPCC special report on climate change, desertification, land degradation, sustainable land
management, food security, and greenhouse gas fluxes in terrestrial ecosystems [P.R. Shukla, J.
Skea, E. Calvo Buendia, V. Masson-Delmotte, H.-O. Po' rtner, D.C. Roberts, P. Zhai, R. Slade,
S. Connors, R. van Diemen, M. Ferrat, E. Haughey, S. Luz, S. Neogi, M. Pathak, J. Petzold, J.
Portugal Pereira, P. Vyas, E. Huntley, K. Kissick, M. Belkacemi, J. Malley, (eds.)]
457
-------
301 Intergovernmental Panel on Climate Change (IPCC). 2019b. IPCC Special Report on the
Ocean and Cryosphere in a Changing Climate [H.-O. Po' rtner, D.C. Roberts, V. Masson-
Delmotte, P. Zhai M. Tignor, E. Poloczanska, K. Mintenbeck, A. Alegn'a, M. Nicolai, A. Okem,
J. Petzold, B. Rama, N.M. Weyer (eds.)].
302 Intergovernmental Panel on Climate Change (IPCC). 2021. Summary for Policymakers. In:
Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth
Assessment Report of the Intergovernmental Panel on Climate Change [Masson-Delmotte, V., P.
Zhai, A. Pirani, S.L. Connors, C. Pe'an, S. Berger, N. Caud, Y. Chen, L. Goldfarb, M.I. Gomis,
M. Huang, K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K. Maycock, T. Waterfield, O. Yelekc^,
R. Yu and B. Zhou (eds.)]. Cambridge University Press. In Press.
303 National Academies of Sciences, Engineering, and Medicine. 2016. Attribution of Extreme
Weather Events in the Context of Climate Change. Washington, DC: The National Academies
Press. Available at: https://www.nap.edu/catalog/21852/attribution-of-extreme-weather-events-
in-the-context-of-climate-change.
304 National Academies of Sciences, Engineering, and Medicine (National Academies). 2017.
Valuing Climate Damages: Updating Estimation of the Social Cost of Carbon Dioxide.
Washington, D.C.: National Academies Press, https://doi.org/10.17226/24651.
305 National Academies of Sciences, Engineering, and Medicine. 2019. Climate Change and
Ecosystems. Washington, DC: The National Academies Press, https://doi.org/10.17226/25504.
306 Blunden, J. and T. Boyer, Eds., 2020: "State of the Climate in 2020". Bull. Amer. Meteor.
Soc., 102 (8), Si-S475, doi:10.1175/2021BAMSStateoftheClimate.l. Available at:
https://www.ametsoc.org/index.cfm/ams/publications/bulletin-of-the-american-meteorological-
society-bams/state-of-the-climate/.
307 U.S. Environmental Protection Agency (U.S. EPA). 2021c. Climate Change and Social
Vulnerability in the United States: A Focus on Six Impacts. U.S. Environmental Protection
Agency, EPA 430-R-21-003.
308 Han, Jaehoon. Memorandum to the Docket EPA-HQ-OAR-2019-0055: "MOVES Modeling-
Related Data Files (MOVES Code, Input Databases and Runspecs) for the Proposed Heavy-Duty
2027 Standards." February 2022.
309 Sonntag, Darrell. Population and Activity of Onroad Vehicles in MOVESCTINPRM.
Attachment to Memorandum to Docket EPA-HQ-OAR-2019-0055: "Updates to MOVES for
Emissions Analysis of the HD 2027 NPRM." May 2021.
310 Sonntag, Darrell. Exhaust Emission Rates for Heavy-Duty Onroad Vehicles in
MOVES CTI NPRM. Attachment to Memorandum to Docket EPA-HQ-OAR-2019-0055:
"Updates to MOVES for Emissions Analysis of the HD 2027 NPRM." May 2021.
311 Sonntag, Darrell. Additional Updates to MOVES CTI NPRM. Attachment to a
Memorandum to Docket EPA-HQ-OAR-2019-0055. Attachment to Memorandum to Docket
EPA-HQ-OAR-2019-0055: "Updates to MOVES for Emissions Analysis of the HD 2027
NPRM." May 2021.
458
-------
312 US Energy Information Administration (EIA), Annual Energy Outlook 2018, Washington,
DC: February 2019, https://www.eia.gov/outlooks/archive/aeol8/pdf/AEO2018.pdf
313 Sonntag, Darrell. Speciation of Total Organic Gas and Particulate Matter Emissions from
Onroad Vehicles in MOVESCTINPRM. Attachment to Memorandum to Docket EPA-HQ-
OAR-2019-005 5: "Updates to MOVES for Emissions Analysis of the HD 2027 NPRM." May
2021.
314 California Air Resources Board. "Heavy-Duty LowNOx: Meetings & Workshops".
Available online: https://ww2.arb.ca.gov/our-work/programs/heavy-duty-low-nox/heavy-duty-
low-nox-meetings-workshops. Note: The initiatives introduced in their 2016 Workshop have
since become components of CARB's Heavy-Duty "Omnibus" Low NOX Rulemaking.
315 Greenhouse Gas Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty
Engines and Vehicles— Phase 2. 81 FR 73941 (October 25, 2016)
316 2007/2010 Heavy-duty rulemaking. 66 FR 5002, January 18, 2001
317 USEPA (2015). Exhaust Emission Rates for Heavy-Duty On-road Vehicles in MOVES2014.
EPA-420-R-15-015a. Assessment and Standards Division. Office of Transportation and Air
Quality. US Environmental Protection Agency. Ann Arbor, MI. November, 2015.
https://www.epa.gov/moves/moves-technical-reports.
318 59 FR 16262, April 6, 1994
319 65 FR 6698, February 10, 2000.
320 79 FR 23414, April 28, 2014 and 80 FR 0978, February 19, 2015.
321 USEPA (2016). Air Toxic Emissions from On-road Vehicles in MOVES2014. EPA-420-R-
16-016. Office of Transportation and Air Quality. US Environmental Protection Agency. Ann
Arbor, MI. November 2016. http://www.epa.gov/moves/moves-technical-reports.
322 U.S. EPA (2021) Technical Support Document: Air Quality Modeling for the HD 2027
Proposal.
323 U.S. EPA (2018) Technical Support Document EPA's 2014 National Air Toxics
Assessment, https://www.epa.gov/national-air-toxics-assessment/2014-nata-assessment-results
324 U.S. Environmental Protection Agency (2007). Control of Hazardous Air Pollutants from
Mobile Sources; Final Rule. 72 FR 8434, February 26, 2007.
325 U.S. EPA. (2018) 2014 NATA: Assessment Results, https://www.epa.gov/national-air-
toxics-assessment/2014-nata-assessment-results.
326 Rich Cook, Sharon Phillips, Madeleine Strum, Alison Eyth & James Thurman
(2020): Contribution of mobile sources to secondary formation of carbonyl compounds, Journal
of the Air & Waste Management Association, DOI: 10.1080/10962247.2020.1813839
327 U.S. EPA. Integrated Science Assessment (ISA) for Particulate Matter (Final Report, 2019).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
459
-------
328 EPA's Report on the Environment: Regional Haze.
https://cfpub.epa.gov/roe/indicator.cfm?i=21
329 Regional Haze Storymap, accessed in 2020 from epa.gov/visibility.
https://epa.maps.arcgis.com/apps/Cascade/index.html?appid=e4dbe2263elf49fb849aflc73a04e2
£2
330 EPA Report on the Environment, technical documentation.
https://cfpub. epa.gov/roe/technical-documentation. cfm?i=l&pvw=.
331 The Sixth External Peer Review of the Community Multiscale Air Quality (CMAQ)
Modeling System. Available online at: https://www.epa.gov/sites/production/files/2019-
08/documents/sixth_cmaq_peer_review_comment_report_6.19.19.pdf
332 National Emissions Inventory Collaborative (2019). 2016vl Emissions Modeling Platform.
Retrieved from http://views.cira.colostate.edu/wiki/wiki/10202.
333 Skamarock, W.C., et al. (2008) A Description of the Advanced Research WRF Version 3.
https://opensky.ucar.edU/islandora/object/technotes:500.
334 USEPA (2019). Meteorological Model Performance for Annual 2016 Simulation WRF v3.8
https://www.epa.gov/sites/default/files/2020-10/documents/met_model_performance-
2016_wrf.pdf. EPA-454/R-19-010.
335 https://www.mmm.ucar.edu/weather-research-and-forecasting-model
336 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: https://www.cmascenter.org/.
337 Henderson, B., et al. (2018) Hemispheric-CMAQ Application and Evaluation for
2016, Presented at 2019 CMAS Conference, available
https://cmascenter.Org/conference//2018/slides/0850_henderson_hemispheriC"
cmaq_application_2018 .pptx.
338 Mathur, R., et al. (2017) Extending the Community Multiscale Air Quality (CMAQ)
modeling system to hemispheric scales: overview of process considerations and initial
applications, Atmos. Chem. Phys., 17, 12449-12474, https://doi.org/10.5194/acp-17-12449-2017.
339 USEPA (2019). Technical Support Document: Preparation of Emissions Inventories for the
Version 7.1 2016 Hemispheric Emissions Modeling Platform. Office of Air Quality Planning and
Standards.
340 US EPA, 2021. Technical Support Document (TSD) Air Quality Modeling for the HD 2027
Proposal.
341 SMAT-CE (Software for the Modeled Attainment Test - Community Edition) vl.6:
https://www.epa.gov/scram/photochemical-modeling-tools; 2014-2018 FRM data:
ftp://newftp.epa.gov/Air/aqmg/SMAT/Ambient_Data/2018/
460
-------
342 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.
343 U.S. EPA, 2011, Final Cross State Air Pollution Rule Air Quality Modeling TSD.
344 U.S. EPA, 2018. Modeling Guidance For Demonstrating Air Quality Goals for Ozone,
PM2.5, and Regional Haze; EPA-454/R-18-009; Research Triangle Park, NC; November 2018.
345 See EPA HD CTI BCA tool, Docket ID No. EPA-HQ-OAR-2019-0055.
346 "Cost Reduction through Learning in Manufacturing Industries and in the Manufacture of
Mobile Sources, Final Report and Peer Review Report," EPA-420-R-16-018, November 2016.
347 See the 2010 light-duty greenhouse gas rule (75 FR 25324, May 7, 2010); the 2012 light-
duty greenhouse gas rule (77 FR 62624, October 15, 2012); the 2011 heavy-duty greenhouse gas
rule (76 FR 57106, September 15, 2011); the 2016 heavy-duty greenhouse gas rule (81 FR
73478, October 25, 2016); the 2014 light-duty Tier 3 rule (79 FR 23414, April 28, 2014).
348 Francisco Posada, Sarah Chambliss, and Kate Blumberg, "Costs of Emission Reduction
Technologies for Heavy-Duty Diesel Vehicles," International Council on Clean Transportation,
February 2016 (ICCT 2016).
349 See ICCT 2016 at page 26.
350 Francisco Posada Sanchez, Anup Bandivadekar, John German, "Estimated Cost of Emission
Reduction Technologies for Light-Duty Vehicles," International Council on Clean
Transportation, March 2012 (ICCT 2012).
351 Mamidanna, S. 2021. Heavy-Duty Engine Valvetrain Technology Cost Assessment. U.S.
EPA Contract with FEV North America, Inc., Contract No. 68HERC19D0008, Task Order No.
68HERH20F0041. Submitted to the Docket.
352 Mamidanna, S. 2021. Heavy-Duty Vehicles Aftertreatment Systems Cost Assessment.
Submitted to the Docket.
353 "Cost Reduction through Learning in Manufacturing Industries and in the Manufacture of
Mobile Sources, Final Report and Peer Review Report," EPA-420-R-16-018, November 2016.
354 See the 2010 light-duty greenhouse gas rule (75 FR 25324, May 7, 2010); the 2012 light-
duty greenhouse gas rule (77 FR 62624, October 15, 2012); the 2011 heavy-duty greenhouse gas
rule (76 FR 57106, September 15, 2011); the 2016 heavy-duty greenhouse gas rule (81 FR
73478, October 25, 2016); the 2014 light-duty Tier 3 rule (79 FR 23414, April 28, 2014).
355 Rogozhin, Alex, Michael Gallaher, Gloria Helfand, and Walter McManus. "Using Indirect
Cost Multipliers to Estimate the Total Cost of Adding New Technology in the Automobile
Industry." International Journal of Production Economics 124 (2010): 360-368; Heavy Duty
Truck Retail Price Equivalent and Indirect Cost Multipliers, Draft Report, RTI International, RTI
Project Number 021 1577.003.002, July 2010.
461
-------
356 Rogozhin, Alex, Michael Gallaher, Gloria Helfand, and Walter McManus. "Using Indirect
Cost Multipliers to Estimate the Total Cost of Adding New Technology in the Automobile
Industry." International Journal of Production Economics 124 (2010): 360-368.
357 See the 2011 heavy-duty greenhouse gas rule (76 FR 57106, September 15, 2011); the 2016
heavy-duty greenhouse gas rule (81 FR 73478, October 25, 2016); the 2014 light-duty Tier 3 rule
(79 FR 23414, April 28, 2014).
358 Heavy Duty Truck Retail Price Equivalent and Indirect Cost Multipliers, Draft Report, RTI
International, RTI Project Number 021 1577.003.002, July 2010.
359 Rogozhin, Alex, Michael Gallaher, Gloria Helfand, and Walter McManus. "Using Indirect
Cost Multipliers to Estimate the Total Cost of Adding New Technology in the Automobile
Industry." International Journal of Production Economics 124 (2010): 360-368.
360 Lynch, Lauren, A. Chad A. Hunter, Bradley T. Zigler, Matthew J. Thornton, and Evan P.
Reznicek. 2020. On-Road Heavy-Duty Low-NOx Technology Cost Study. Golden, CO: National
Renewable Energy Laboratory. NREL/TP-5400-76571.
https://www.nrel.gov/docs/fy20osti/76571.pdf.
361 "Nonconformance Penalties for On-highway Heavy-duty Diesel Engines: Technical Support
Document," EPA-420-R-12-014, Figure 3-1 at page 37.
362 "Tier 3 Certification Fuel Impacts Test Program, Appendix A, Table A-l," EPA-420-R-18-
004.
363 See "TheEngineeringToolBox.pdf," generated via
https://www.engineeringtoolbox.com/gross-net-heating-values-d_420.html, accessed May 29,
2020.
364 "Tier 3 Certification Fuel Impacts Test Program, Appendix A, Table A-l," EPA-420-R-18-
004.
365 Thomas, M., and S. Rao (1999). "Warranty Economic Decision Models: A Summary and
Some Suggested Directions for Future Research." Operations Research 47(6):807-820.
366 Wu, S (2012). Warranty Data Analysis: A Review. Quality and Reliability Engineering
International 28: 795-805.
367 Guajardo, J., M Cohen, and S. Netessine (2016). "Service Competition and Product Quality
in the U.S. Automobile Industry." Management Science 62(7): 1860-1877.
368 Murthy, D., and N. Jack (2009). "Warranty and Maintenance," Chapter 18 in Handbook of
Maintenance Management and Engineering, Mohamed Ben-Daya et al., editors. London:
Springer.
369 Saidi-Mehrabad, M., R. Noorossana, and M. Shafiee (2010). "Modeling and analysis of
effective ways for improving the reliability of second-hand products sold with warranty."
International Journal of Advanced Manufacturing Technology 46: 253-265.
462
-------
370 See "Mitigating Rising Maintenance & Repair Costs for Class-8 Truck Fleets, Effective
Data & Strategies to Leverage Newer Trucks to Reduce M&R Costs," Fleet Advantage
Whitepaper Series, 2018.
371 American Transportation Research Institute, "An Analysis of the Operational Costs of
Trucking: 2019 Update," November 2019, Table
372 See "Mitigating Rising Maintenance & Repair Costs for Class-8 Truck Fleets, Effective
Data & Strategies to Leverage Newer Trucks to Reduce M&R Costs," Fleet Advantage
Whitepaper Series, 2018 at page 8.
373 See docket memo, "Estimated Warranty and Useful Life Ages Used in Estimating Emission
Repair Costs," from Todd Sherwood.
374 See "Mitigating Rising Maintenance & Repair Costs for Class-8 Truck Fleets, Effective
Data & Strategies to Leverage Newer Trucks to Reduce M&R Costs," Fleet Advantage
Whitepaper Series, 2018.
375 See "Mitigating Rising Maintenance & Repair Costs for Class-8 Truck Fleets, Effective
Data & Strategies to Leverage Newer Trucks to Reduce M&R Costs," Fleet Advantage
Whitepaper Series, 2018.
376 See USEPA HD CTI BCA tool contained in the docket for this rule.
377 U.S. Environmental Protection Agency (U.S. EPA). 2021. Regulatory Impact Analysis for
the Final Revised Cross-State Air Pollution Rule (CSAPR) Update for the 2008 Ozone NAAQS.
EPA-452/R-21-002. March.
378 U.S. Environmental Protection Agency (U.S. EPA). 2014. Control of Air Pollution from
Motor Vehicles: Tier 3 Motor Vehicle Emission and Fuel Standards Rule Regulatory Impact
Analysis. EPA-420-R-14-005. March.
379 U.S. Environmental Protection Agency (U.S. EPA). 2019. Integrated Science Assessment
(ISA) for Particulate Matter (Final Report, 2019). U.S. Environmental Protection Agency,
Washington, DC, EPA/600/R-19/188, 2019.
380 U.S. Environmental Protection Agency (U.S. EPA). 2020. Integrated Science Assessment
(ISA) for Ozone and Related Photochemical Oxidants (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-20/012, 2020.
381 U.S. Environmental Protection Agency (U.S. EPA). 2021. Estimating PM2.5- and Ozone-
Attributable Health Benefits. Technical Support Document (TSD) for the Final Revised Cross-
State Air Pollution Rule Update for the 2008 Ozone Season NAAQS. EPA-HQ-OAR-2020-
0272. March.
382 U.S. Environmental Protection Agency (U.S. EPA). 2021. Estimating PM2.5- and Ozone-
Attributable Health Benefits. Technical Support Document (TSD) for the Final Revised Cross-
State Air Pollution Rule Update for the 2008 Ozone Season NAAQS. EPA-HQ-OAR-2020-
0272. March.
383 Woods & Poole (2015). Complete Demographic Database.
463
-------
384 U.S. Energy Information Administration (U.S. EIA). 2020. Annual Energy Outlook 2020
with Projections to 2050. January, www.eia.gov/aeo.
385 U.S. Environmental Protection Agency (U.S. EPA). 2018. Modeling Guidance for
Demonstrating Air Quality Goals for Ozone, PM2.5, and Regional Haze. EPA 454/R-18-009.
https://www.epa.gov/sites/production/files/2020-10/documents/o3-pm-rh-modeling_guidance-
2018.pdf.
386 U.S. Environmental Protection Agency (U.S. EPA). 2015. Regulatory Impact Analysis of
the Final Revisions to the National Ambient Air Quality Standards for Ground-Level Ozone,
EPA-452/R-15-07. EPA-452/R-12-003. Office of Air Quality Planning and Standards, Health
and Environmental Impacts Division, Research Triangle Park, NC.
(https://www.epa.gOv/sites/production/files/2016-02/documents/20151001ria.pdf).
387 U.S. Environmental Protection Agency (U.S. EPA). 2012. Regulatory Impact Analysis for
the Final Revisions to the National Ambient Air Quality Standards for Particulate Matter. EPA-
452/R-12-003. Office of Air Quality Planning and Standards, Health and Environmental Impacts
Division, Research Triangle Park, NC. Available at: <
https://www3.epa.gov/ttnecasl/regdata/RIAs/finalria.pdf>.
388 U.S. Environmental Protection Agency (U.S. EPA). 2018. Environmental Benefits Mapping
and Analysis Program - Community Edition. User's Manual. Office of Air Quality Planning and
Standards, Health and Environmental Impacts Division, Research Triangle Park, NC. Available
at: < https://www.epa.gOv/sites/production/files/2015-04/documents/benmap-
ce_user_manual_march_2015 ,pdf>
389 National Research Council (NRC). 2008. Estimating Mortality Risk Reduction and
Economic Benefits from Controlling Ozone Air Pollution. National Academies Press.
Washington, DC.
390 Bell, M.L., A. McDermott, S.L. Zeger, J.M. Sarnet, and F. Dominici. 2004. "Ozone and
Short-Term Mortality in 95 U.S. Urban Communities, 1987-2000." Journal of the American
Medical Association. 292(19):2372-8.
391 Huang Y, Dominici F, Bell M. 2004. Bayesian Hierarchical Distributed Lag Models for
Summer Ozone Exposure and Cardio-Respiratory Mortality. Johns Hopkins Univ Dept Biostat
Work Pap Ser.
392 Schwartz, J. 2005. "How Sensitive is the Association between Ozone and Daily Deaths to
Control for Temperature?" American Journal of Respiratory and Critical Care Medicine. 171(6):
627-31.
393 Bell, M.L., F. Dominici, and J.M. Samet. 2005. "A Meta-Analysis of Time-Series Studies of
Ozone and Mortality with Comparison to the National Morbidity, Mortality, and Air Pollution
Study." Epidemiology. 16(4):436-45.
394 Ito, K., S.F. De Leon, and M. Lippmann. 2005. "Associations Between Ozone and Daily
Mortality: Analysis and Meta-Analysis." Epidemiology. 16(4):446-57.
464
-------
395 Levy, J.I., S.M. Chemerynski, and J.A. Sarnat. 2005. "Ozone Exposure and Mortality: An
Empiric Bayes Metaregression Analysis." Epidemiology. 16(4):458-68.
396 Smith RL, Xu B, Switzer P. 2009. Reassessing the relationship between ozone and short-
term mortality in U.S. urban communities. Inhal Toxicol 21 Suppl 2:37-61;
doi: 10.1080/08958370903161612.
397 Zanobetti A, Schwartz J. 2008. Mortality displacement in the association of ozone with
mortality: an analysis of 48 cities in the United States. Am J Respir Crit Care Med 177:184-9;
doi: 10.1164/rccm.200706-8230C.
398 Jerrett M, Burnett RT, Pope CA, Ito K, Thurston G, Krewski D, et al. 2009. Long-term
ozone exposure and mortality. N Engl J Med 360:1085-95; doi:10.1056/NEJMoa0803894.
399 U.S. Environmental Protection Agency (U.S. EPA). 2020. Integrated Science Assessment
(ISA) for Ozone and Related Photochemical Oxidants (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-20/012, 2020.
400 Turner, MC, Jerrett, M, Pope, A, III, Krewski, D, Gapstur, SM, Diver, WR, Beckerman, BS,
Marshall, JD, Su, J, Crouse, DL and Burnett, RT (2016). Long-term ozone exposure and
mortality in a large prospective study. Am J Respir Crit Care Med 193(10): 1134-1142.
401 Zanobetti A, Schwartz J. 2008. Mortality displacement in the association of ozone with
mortality: an analysis of 48 cities in the United States. Am J Respir Crit Care Med 177:184-9;
doi: 10.1164/rccm.200706-8230C.
402 Katsouyanni, K, Samet, JM, Anderson, HR, Atkinson, R, Le Tertre, A, Medina, S, Samoli,
E, Touloumi, G, Burnett, RT, Krewski, D, Ramsay, T, Dominici, F, Peng, RD, Schwartz, J,
Zanobetti, A and Committee, HEIHR (2009). Air pollution and health: a European and North
American approach (APHENA). Res Rep Health Eff Inst(142): 5-90.
403 Turner, MC, Jerrett, M, Pope, A, III, Krewski, D, Gapstur, SM, Diver, WR, Beckerman, BS,
Marshall, JD, Su, J, Crouse, DL and Burnett, RT (2016). Long-term ozone exposure and
mortality in a large prospective study. Am J Respir Crit Care Med 193(10): 1134-1142.
404 Di, Q, Wang, Y, Zanobetti, A, Wang, Y, Koutrakis, P, Choirat, C, Dominici, F and
Schwartz, JD (2017). Air pollution and mortality in the Medicare population. New Engl J Med
376(26): 2513-2522.
405 U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2019.
Letter from Louis Anthony Cox, Jr., Chair, Clean Air Scientific Advisory Committee, to
Administrator Andrew R. Wheeler. Re: CASAC Review of the EPA's Integrated Science
Assessment for Particulate Matter (External Review Draft - October 2018). April 11, 2019. EPA-
CASAC-19-002. Office of the Administrator, Science Advisory Board U.S. EPA HQ,
Washington DC. Available at:
https://casac.epa.gov/ords/sab/f?p=105:12:17077188731556::: 12::.
406 U.S. Environmental Protection Agency (U.S. EPA). 2019. Integrated Science Assessment
(ISA) for Particulate Matter (Final Report, 2019). U.S. Environmental Protection Agency,
Washington, DC, EPA/600/R-19/188, 2019.
465
-------
407 U.S. Environmental Protection Agency (U.S. EPA). 2010. Regulatory Impact Analysis
(RIA) for Existing Stationary Compression Ignition Engines NESHAP Final Draft.
408 U.S. Environmental Protection Agency (U.S. EPA). 2010. Regulatory Impact Analysis for
the Proposed Federal Transport Rule.
409 U.S. EPA. Greenhouse gas emissions standards and fuel efficiency standards for medium-
and heavy-duty engines and vehicles. Final rules. Fed Reg. 2011;76(179):57106-57513.
410 U.S. Environmental Protection Agency (U.S. EPA). 2011. Regulatory Impact Analysis for
the Final Mercury and Air Toxics Standards.
411 U.S. Environmental Protection Agency (U.S. EPA). 2012. Regulatory Impact Analysis for
the Final Revisions to the National Ambient Air Quality Standards for Particulate Matter. EPA-
452/R-12-003. Office of Air Quality Planning and Standards, Health and Environmental Impacts
Division, Research Triangle Park, NC. Available at: <
https://www3.epa.gov/ttnecasl/regdata/RIAs/finalria.pdf>.
412 U.S. Environmental Protection Agency (U.S. EPA). 2013. Regulatory Impact Analysis for
the Final Revisions to the National Ambient Air Quality Standards for Particulate Matter.
413 U.S. Environmental Protection Agency (U.S. EPA). 2014. Regulatory Impact Analysis
(RIA) for Proposed Residential Wood Heaters NSPS Revision.
414 U.S. Environmental Protection Agency (U.S. EPA). 2014. Regulatory Impact Analysis for
the Proposed Carbon Pollution Guidelines for Existing Power Plants and Emission Standards for
Modified and Reconstructed Power Plants.
415 U.S. Environmental Protection Agency (U.S. EPA). 2014. Regulatory Impact Analysis of
the Proposed Revisions to the National Ambient Air Quality Standards for Ground-Level Ozone.
416 U.S. Environmental Protection Agency (U.S. EPA). 2015. Regulatory Impact Analysis for
Residential Wood Heaters NSPS Revision: Final Report.
417 U.S. Environmental Protection Agency (U.S. EPA). 2015. Regulatory Impact Analysis for
the Clean Power Plan Final Rule.
418 U.S. Environmental Protection Agency (U.S. EPA). 2015. Regulatory Impact Analysis for
the Proposed Cross-State Air Pollution Rule (CSAPR) Update for the 2008 Ozone National
Ambient Air Quality Standards (NAAQS).
419 U.S. Environmental Protection Agency (U.S. EPA). 2015. Regulatory Impact Analysis for
the Proposed Federal Plan Requirements for Greenhouse Gas Emissions from Electric Utility
Generating Units Constructed on or Before January 8, 2014; Model Trading Rules; Amendments
to Framework Regulations. 4-52
420 U.S. Environmental Protection Agency (U.S. EPA). 2015. Regulatory Impact Analysis of
the Final Revisions to the National Ambient Air Quality Standards for Ground-Level Ozone,
EPA-452/R-15-07. EPA-452/R-12-003. Office of Air Quality Planning and Standards, Health
and Environmental Impacts Division, Research Triangle Park, NC.
(https://www.epa.gOv/sites/production/files/2016-02/documents/20151001ria.pdf).
466
-------
421 U.S. Environmental Protection Agency (U.S. EPA). 2016. Regulatory Impact Analysis of
the Cross-State Air Pollution Rule (CSAPR) Update for the 2008 National Ambient Air Quality
Standards for Ground-Level Ozone.
422 National Research Council (NRC). 2002. Estimating the Public Health Benefits of Proposed
Air Pollution Regulations. National Academies Press. Washington, DC.
423 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, Research Triangle Park, NC. December. Available at:
.
424 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, Research Triangle Park, NC. December. Available at:
. Section 11.2.4
425 Review of the National Ambient Air Quality Standards for Particulate Matter, Federal
Register, 85 FR 82696, December 18, 2020.
426 Woodruff, T.J., J. Grillo, and K.C. Schoendorf. 2008. "Air pollution and postneonatal infant
mortality in the United States, 1999-2002." Environmental Health Perspectives. 116(1): 110-
115.
427 U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2000.
An SAB Report on EPA's White Paper Valuing the Benefits of Fatal Cancer Risk Reduction.
EPA-SAB-EEAC-00-013. July.
428 U.S. Environmental Protection Agency (U.S. EPA). 2016. Guidelines for Preparing
Economic Analyses.
429 U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2017.
Letter from Peter S. Thorne, Chair, Science Advisory Board, and Madhu Khanna, Chair, SAB
Environmental Economics Advisory Committee, to Administrator Scott Pruitt. Re: SAB Review
of EPA's Proposed Methodology for Updating Mortality Risk Valuation Estimates for Policy
Analysis. February 23, 2017. EPA-SAB-17-005. Office of the Administrator, Science Advisory
Board U.S. EPA HQ, Washington DC. Available at:
https://cfpub. epa.gov/si/si_public_file_download. cfm?p_download_id=532807.
430 United States Office of Management and Budget (OMB). 2003. Circular A-4: Regulatory
Analysis. Washington, DC. Available at: <
https://obamawhitehouse.archives.gov/omb/circulars_a004_a-4/ >.
431 U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2004.
Advisory Council on Clean Air Compliance Analysis Response to Agency Request on Cessation
Lag. EPA-COUNCIL-LTR-05-001. December. Available at: <
https://council.epa.gov/ords/sab/f?p=104:12:968651521971>.
467
-------
432 U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2010.
Review of EPA's Draft Health Benefits of the Second Section 812 Prospective Study of the
CAA
433 USGCRP. 2016. The Impacts of Climate Change on Human Health in the United States: A
Scientific Assessment.; http://dx.doi.org/10.7930/J0R49NQX.
434 Fann N, Nolte CG, Dolwick P, Spero TL, Brown AC, Phillips S, et al. 2015. The
geographic distribution and economic value of climate change-related ozone health impacts in
the United States in 2030. J Air Waste Manag Assoc 65; doi: 10.1080/10962247.2014.996270.
435 Jhun I, Fann N, Zanobetti A, Hubbell B. 2014. Effect modification of ozone-related
mortality risks by temperature in 97 US cities. Environment International. 73:128-34.
436 Ren, C., G.M. Williams, K. Mengersen, L. Morawska, and S. Tong. 2008. "Does
Temperature Modify Short-Term Effects of Ozone on Total Mortality in 60 Large Eastern U.S.
Communities? An Assessment Using the NMMAPS Data." Environment International. 34:451-
458.
437 Ren, C., G.M. William, L. Morawska, K. Mengensen, and S. Tong. 2008. "Ozone Modifies
Associations between Temperature and Cardiovascular Mortality: Analysis of the NMMAPS
Data." Occupational and Environmental Medicine. 65:255-260.
438National Research Council (NRC). 2002. Estimating the Public Health Benefits of Proposed
Air Pollution Regulations. National Academies Press. Washington, D
439 U.S. Environmental Protection Agency (U.S. EPA). 2019. Integrated Science Assessment
(ISA) for Particulate Matter (Final Report, 2019). U.S. Environmental Protection Agency,
Washington, DC, EPA/600/R-19/188, 2019.
440 U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2004.
Advisory Council on Clean Air Compliance Analysis Response to Agency Request on Cessation
Lag. EPA-COUNCIL-LTR-05-001. December. Available at: <
https://council.epa.gov/ords/sab/f?p=104:12:968651521971>.
441 U.S. Environmental Protection Agency (U.S. EPA). 2020. Integrated Science Assessment
(ISA) for Ozone and Related Photochemical Oxidants (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-20/012, 2020. Section 6.2.6
442 U.S. Environmental Protection Agency (U.S. EPA). 2018. Technical Support Document:
Estimating the Benefit per Ton of Reducing PM2.5 Precursors from 17 Sectors. February.
443 Turner, MC, Jerrett, M, Pope, A, III, Krewski, D, Gapstur, SM, Diver, WR, Beckerman, BS,
Marshall, JD, Su, J, Crouse, DL and Burnett, RT (2016). Long-term ozone exposure and
mortality in a large prospective study. Am J Respir Crit Care Med 193(10): 1134-1142.
444 Di Q, et al. (2017) Air pollution and mortality in the Medicare population. N Engl J Med
376:2513-2522.
445 Katsouyanni, K, Samet, JM, Anderson, HR, Atkinson, R, Le Tertre, A, Medina, S, Samoli,
E, Touloumi, G, Burnett, RT, Krewski, D, Ramsay, T, Dominici, F, Peng, RD, Schwartz, J,
468
-------
Zanobetti, A and Committee, HEIHR (2009). Air pollution and health: a European and North
American approach (APHENA). Res Rep Health Eff Inst(142): 5-90.
446 Turner, MC, Jerrett, M, Pope, A, III, Krewski, D, Gapstur, SM, Diver, WR, Beckerman,
BS, Marshall, JD, Su, J, Crouse, DL and Burnett, RT (2016). Long-term ozone exposure and
mortality in a large prospective study. Am J Respir Crit Care Med 193(10): 1134-1142.
447 U.S. Environmental Protection Agency (U.S. EPA). 2016. Integrated Science Assessment
for Oxides of Nitrogen - Health Criteria (Final Report). National Center for Environmental
Assessment, Research Triangle Park, NC. July. Available at: <
https://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=310879>.
448 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 - RTP Division, Research Triangle Park, NC. EPA/600/R-
08/139. December. Available at:
.
449 U.S. EPA and National Highway Traffic Safety Administration. "Greenhouse Gas Emissions
and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles - Phase 2
Regulatory Impact Analysis." EPA-420-R-16-900. 2016.
450 National Academies of Sciences, Engineering, and Medicine. "Reducing Fuel Consumption
and Greenhouse Gas Emissions of Medium- and Heavy-Duty Vehicles, Phase Two: Final
Report," p. 329. Washington, DC: The National Academies Press. 2020.
https://doi.org/10.17226/25542.
451 Rubin, J. "A Model of Intertemporal Emission Trading, Banking, and Borrowing." Journal
of Environmental Economics and Management 31: 269-286. 1996.
452 Lam, T., and Bausell, C. "Strategic Behaviors Toward Environmental Regulation: A Case of
Trucking Industry." Contemporary Economic Policy 25(1): 3-13. 2007.
453 Rittenhouse, K., and Zaragoza-Watkins,M. "Anticipation and Environmental Regulation."
Journal of Environmental Economics and Management 89: 255-277. 2018.
454 Harrison, D., Jr., and LeBel, M. "Customer Behavior in Response to the 2007 Heavy-Duty
Engine Emission Standards: Implications for the 2010 NOX Standard." NERA Economic
Consulting. 2008. Available in the docket, Docket ID EPA-HQ-OAR-2019-0055-0576.
455 U.S. Environmental Protection Agency. "Analysis of Heavy-Duty Vehicle Sales Impacts
Due to New Regulation." EPA-420-R-21-013. 2021.
https://cfpub.epa.gov/si/si_public_pra_view.cfm?dirEntryID=349838&Lab=OTAQ.
456 U.S. Environmental Protection Agency. "Analysis of Heavy-Duty Vehicle Sales Impacts
Due to New Regulation: Response to Reviewers Comments." EPA-420-R-21-014. 2021.
https://cfpub.epa.gov/si/si_public_pra_view.cfm?dirEntryID=349838&Lab=OTAQ.
457 New vehicle sales from Wards Intelligence. "U.S. Factory Sales of Medium/Heavy Trucks
and Buses by GVW " UsaFsOl O.xlsx. Registration data from Wards Intelligence. "U.S. Total
469
-------
Truck Registrations by State and Type, 2017," sum of "Truck Tractors and "Other Med./Hvy."
UsaRel0_2017.xlsx. 2019
458 Comer, B.; Corbett, J. J.; Hawker, J.S.; Korfmacher, K.; Lee , E.E.; Prokop, C.; and
Winebrake. J. "Marine Vessels as Substitutes for Heavy-Duty Trucks in Great Lakes Freight
Transportation." Journal of the Air & Waste Management Association 60: 884-890. 2010.
459 U.S. EPA Office of Transportation and Air Quality. "Economic Impacts of the Category 3
Marine Rule on Great Lakes Shipping." EPA-420-R-12-005. 2012.
460 Bushnell, J., and Hughes, J. "Mode Choice, Energy Consumption and Emissions in U.S.
Freight Transportation." Working paper. 2019. Available online:
https://spot.colorado.edu/~jonathug/Jonathan_E._Hughes/Main_files/Freight_Modes.pdf,
accessed 10/21/2019.
461 Morgenstern, R.D.; Pizer, W.A.; and Shih, J.-S. "Jobs Versus the Environment: An Industry-
Level Perspective." Journal of Environmental Economics and Management 43: 412-436. 2002.
462 Berman, E. and Bui, L. T. M. "Environmental Regulation and Labor Demand: Evidence
from the South Coast Air Basin." Journal of Public Economics 79(2): 265-295. 2001.
463 Deschenes, O. "Balancing the Benefits of Environmental Regulations for Everyone and the
Costs to Workers and Firms." IZA World of Labor 22v2. 2018. Available online:
https://wol.iza.org/uploads/articles/458/pdfs/environmental-regulations-and-labor-markets.pdf
464 Ehrenberg, R. G., and Smith, R.S. Modern Labor Economics: Theory and Public Policy
(Addison Wesley Longman, Inc.), p. 108. 2000.
465 Greenstone, M.. "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. 2002.
466 Ferris, A.; Shadbegian, R.J.; and Wolverton, A. "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 1(4): 521-553. 2014.
467 Walker, R.W. "The Transitional Costs of Sectoral Reallocation: Evidence From the Clean
Air Act and the Workforce." The Quarterly Journal of Economics. 1787-1835. 2013.
468 Curtis, M.E. "Who Loses Under Cap-and-Trade Programs? The Labor Market Effects of the
NOx Budget Trading Program." The Review of Economics and Statistics 100(1): 151-166. 2018.
469 Hafstead ,M.A.C. and Williams III, R.C. "Unemployment and Environmental Regulation in
General Equilibrium." Journal of Public Economics 160: 50-65. 2018.
470 "Comments of the American Trucking Associations on the Control of Air Pollution from
New Motor Vehicles: Heavy-Duty Engine Standards Advance Notice of Proposed Rulemaking."
Docket EPA-HQ-OAR-2019-0055-03 57.
471 Comments of the National Association of Small Trucking Companies, "Draining the
Swamp," Docket EPA-HQ-OAR-2019-0055-0456; Comments of the Owner-Operator
Independent Drivers Association, "Re: Docket # EPA-HQ-OAR-2019-0055 'Control of Air
470
-------
Pollution from New Motor Vehicles: Heavy-Duty Engine Standards,'" Docket EPA-HQ-OAR-
2019-0055-0379.
472 Comments of the Owner-Operator Independent Drivers Association, "Re: Docket # EPA-
HQ-OAR-2019-0055 'Control of Air Pollution from New Motor Vehicles: Heavy-Duty Engine
Standards,"' Docket EPA-HQ-OAR-2019-0055-0379.
473 U.S. Bureau of Labor Statistics. National Industry-Specific Occupational Employment and
Wage Estimates: NAICS 336100 - Motor Vehicle Manufacturing. May 2019. Available online:
https://www.bls.gov/oes/2019/may/naics4_336100.htm, accessed 5/21/2020.
474 U.S. Small Business Administration. Table of Small Business Size Standards Matched to
North American Industry Classification System Codes. 2019 Version. Available online:
https ://www. sba.gov/ sites/default/files/2019-
08/SBA%20Table%20of%20Size%20Standards_Effective%20Aug%2019%2C%202019_Rev.p
df
475 NAICS Association. NAICS & SIC Identification Tools. Available online:
https://www.naics.com/search
471
-------
|