Exhaust Emission Rates for Heavy-Duty
Onroad Vehicles in MOVES4
£%	United States
Environmental Protect
Agency

-------
Exhaust Emission Rates for Heavy-Duty
This technical report does not necessarily represent final EPA decisions
or positions. It is intended to present technical analysis of issues using
data that are currently available. The purpose in the release of such
reports is to facilitate the exchange of technical information and to
inform the public of technical developments.
Onroad Vehicles in MOVES4
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
NOTICE
4>EPA
United States
Environmental Protection
Agency
EPA-420-R-23-027
August 2023

-------
Table of Contents
1	Introduction	1
1.1	Pollutant	2
1.2	Emission Process	2
1.2.1	Running Exhaust	2
1.2.2	Start Exhaust	3
1.2.3	Extended Idle and Auxiliary Power Exhaust	3
1.2.4	Crankcase Exhaust	4
1.2.5	Evaporative and Brake and Tire Wear Emissions	4
1.3	Fuel Type	5
1.4	Regulatory Class	5
1.5	Model Year Groups	6
1.6	Operating Modes	6
1.7	Vehicle Age	12
1.8	MO VES4 Updates	13
1.8.1	General Updates	13
1.8.2	Updates to incorporate HD2027 Standards	14
2	Heavy-Duty Diesel Exhaust Emissions	14
2.1	Running Exhaust Emissions	14
2.1.1	Nitrogen Oxides (NOx)	15
2.1.2	Particulate Matter (PM2.5)	60
2.1.3	Total Hydrocarbons (THC) and Carbon Monoxide (CO)	81
2.1.4	Energy	94
2.1.5	Evaluation of Fleet-average Running Rates with Real-World Measurements	104
2.2	Start Exhaust Emissions	105
2.2.1	THC, CO, and NOx	105
2.2.2	Particulate Matter (PM2.5)	115
2.2.3	Adjusting Start Rates for Soak Time	117
2.2.4	Start Energy Rates	128
2.3	Extended Idling Exhaust Emissions	130
2.3.1	1960-2006 Model Years	131
2.3.2	2007-2026 Model Years	134
ii

-------
2.3.3	2027-2060 Model Years	144
2.3.4	Model Year Trends	144
2.3.5	Extended Idle Energy Rates	147
2.4	Auxiliary Power Unit Exhaust	147
2.5	Glider Vehicle Emissions	153
3	Heavy-Duty Gasoline Exhaust Emissions	155
3.1	Running Exhaust Emissions	155
3.1.1	THC, COandNOx	155
3.1.2	Particulate Matter (PM2.5)	174
3.1.3	Energy	182
3.2	Start Emissions	187
3.2.1	THC, CO, andNOx	188
3.2.2	Particulate Matter (PM2.5)	203
3.2.3	Soak Time Adjustments	204
3.2.4	Start Energy Rates	205
4	Heavy-Duty Compressed Natural Gas Exhaust Emissions	207
4.1	Running Exhaust Emission Rates	209
4.1.1	1960-2009 Model Years	209
4.1.2	2010-2060 Model Years	219
4.1.3	Model Year Trends	220
4.2	Start Exhaust Emission Rates	225
4.3	Extended Idle Exhaust Emission Rates	225
5	Heavy-Duty Ammonia Emissions	226
5.1	Heavy-Duty Diesel	226
5.2	Heavy-Duty Gasoline	230
5.2.1	1960-1980 Model Years	231
5.2.2	1981-2060 Model Years	231
5.3	Heavy-Duty Compressed Natural Gas	232
5.4	Summary	233
6	Heavy-Duty Crankcase Exhaust Emissions	235
6.1	Modeling Crankcase Emissions in MOVES	235
6.2	Heavy-Duty Diesel Crankcase Emissions	236
6.2.1 LHD2b3 Crankcase Emissions	237
111

-------
6.2.2	LHD45, MHD and HHD Crankcase Emissions	238
6.2.3	Glider Crankcase Emissions	253
6.3 Heavy-Duty Gasoline and CNG Crankcase Emissions	253
7	Nitrogen Oxide Composition	255
7.1	Heavy-Duty Diesel	256
7.2	Heavy-Duty Gasoline	259
7.3	Heavy-Duty Compressed Natural Gas	259
8	Appendices	260
Appendix A Calculation of Accessory Power Requirements	261
Appendix B Tampering and Mal-maintenance for Diesel Running Exhaust	262
Appendix C Tampering and Mal-maintenance for MY 2007 and Later Diesel Extended Idle. 282
Appendix D Pre-2007 Model Year Extended Idle Data Summary	284
Appendix E Developing Pre-2007 Model Year HD Diesel PM2.5 Emission Rates for Missing
Operating Modes	288
Appendix F Heavy-Duty Gasoline Start Emissions Analysis Figures	289
Appendix G Selection of Fixed Mass Factor (fSCaie) values for MY 2010+ Heavy-Duty Vehicles
294
Appendix H THC and CO Emisssion rates from 2010 and Later Model Year Heavy-duty Vehicles
from the HDIUT	306
Appendix I Analysis of 2010 and Later Model Year Heavy-duty Gasoline Emission Rates	311
Appendix J PM Composition Measurements from Auxiliary Power Units	320
9	References	321
iv

-------
List of Acronyms
ABT
emissions averaging, banking and trading program
A/C
Air Conditioning
ACES
Advanced Collaborative Emission Study (CRC)
APU
auxiliary power units
ARCO
Atlantic Richfield Company
BC
black carbon
bhp
brake horsepower
BTU
British Thermal Unit
CARB
California Air Resources Board
CBD
Central Business District
CFR
Code of Federal Regulations
ch4
methane
CNG
Compressed Natural Gas
CO
carbon monoxide
C02
carbon dioxide
CRC
Coordinating Research Council
CTI
Cleaner Trucks Initiative (later known as "Clean Trucks Rule")
DB
database
DOC
diesel oxidation catalysts
DOE
U.S. Department of Energy
DPF
diesel particulate filter/periodic trap oxidizer
EC
elemental carbon
ECOSTAR
gaseous and exhaust flow measurement system
ECU
Engine Control Unit
EFEE
Engine, Fuel and Emissions Engineering Inc.
EGR
exhaust-gas recirculation
EMFAC
CARB emissions factors model
EPA
U.S. Environmental Protection Agency
ESC
European Stationary Cycle
FEL
family emission limit
FHWA
Federal Highway Administration
FID
Flame Ionization Detection
FTP
Federal Test Procedure
g
Grams
GDI
Gasoline Direct injection engines
GHG
Greenhouse Gases
g/mi
Grams per mile
GPS
Global Positioning System
GVWR
Gross Vehicle Weight Rating
THC
Hydrocarbons
HD
Heavy-Duty
HDIU
Heavy-Duty Diesel In-Use
HDT
Heavy-Duty Truck
HFC
Hydrofluorocarbon
V

-------
H-GAC
Houston-Galveston Area Council
HHD
Heavy-Heavy-Duty Class 8 Trucks (GVWR > 33,000 lbs)
HHDD
Heavy Heavy-Duty Diesel
HN02
nitrous acid (HONO)
HP
horsepower
hr
hour
HV
heating value
H2O
water
I/M
Inspection and Maintenance program
III VP
In-Use Verification Program
kJ
Kilojoules
kW
Kilowatt
LHD
Light-Heavy-Duty
LHD2b3
Light-Heavy-Duty Class 2b and 3 Truck (8,500 < GVWR < 14,000 lbs)
LHD45
Light Heavy-Duty Class 4 or 5 Truck (14,000 < GVWR < 19,500 lbs)
LHDDT
Light Heavy-Duty Diesel Truck
LNT
Lean NOx Trap aftertreatment
MDPV
Medium-Duty Passenger Vehicle
MECA
The Manufacturers of Emission Controls Association
MEMS
Mobile Emissions Measurement System
mg
milligram
MHD
Medium-Heavy-Duty Class 6 and 7 Trucks (19,500 < GVWR < 33,000 lbs)
M0BILE6
EPA Highway Vehicle Emission Factor Model, Version 6
MOVES
Motor Vehicle Emission Simulator Model
MOVES201X
Motor Vehicle Emission Simulator Model development version
MSOD
Mobile Source Observation Database
MY
model year
MYG
model year group
NCHRP
National Cooperative Highway Research Program
NCP
nonconformance penalty
NDIR
non-dispersive infrared
NFRAQS
Northern Front Range Air Quality Study
nh3
ammonia
NMHC
Non-Methane Hydrocarbon
NMOG
non-methane organic gases
NonEC
non-elemental carbon
N onECnon S 04PM
non-elemental carbon non-sulfate particulate matter
NonECPM
non-elemental particulate matter carbon
NO
nitric oxide
NOx
nitrogen oxide
NOy
combined NOx and NOz compound
NOz
nitrous oxide
N02
nitrogen dioxide
NREL
National Renewal Energy Laboratory
NTE
Not-to-Exceed
NYSDEC
New York Department of Environmental Conservation
vi

-------
n2o
nitrous oxide
OBD
On-Board Diagnostics
OC
oxidation catalyst
OEM
Original Equipment Manufacturer
OM
operating mode
OMNMHCE
organic material non-methane hydrocarbon equivalent
PCV
positive crankcase ventilation
PEMS
portable emissions measurement system
PERE
Physical Emission Rate Estimator
PHA
Port of Houston Authority
PM
Particulate Matter
PM2.5
fine particles of particulate matter with aerodynamic diameters <2.5 |im
PM10
particles of particulate matter with aerodynamic diameters <10 |im
ROVER
EPA dataset measurement collection system
RPM
revolutions per minute
SCAQMD
South Coast Air Quality Management District
SCR
selective catalytic reduction
S04
sulfate
STP
scaled tractive power
ST01
258-second driving cycle
T&M
Tampering and Maintenance
TC
total carbon
TEOM
Tapered Element Oscillating Microbalance
THC
Total Hydrocarbon (FID detection)
TOG
Total Organic Gases
TTI
Texas Transportation Institute
TWC
three-way catalysts
UDDS
Urban Dynamometer Driving Schedule
UL
useful life
ULSD
Ultra Low Sulfur Diesel
VIN
Vehicle Identification Number
VIUS
Vehicle Inventory and Use Survey
VMT
Vehicle Miles Traveled
VOC
Volatile Organic Compounds
VSP
vehicle specific power
WMATA
Washington Metropolitan Area Transit Authority
WVU
West Virginia University
ZML
zero-mile emissions level
vii

-------
1 Introduction
The United States Environmental Protection Agency's Motor Vehicle Emission Simulator
(MOVES) is a set of modeling tools for estimating air pollution emissions produced by onroad
(highway) and nonroad mobile sources. MOVES estimates the emissions of greenhouse gases
(GHGs), criteria pollutants and selected air toxics. The MOVES model is currently the official
model for use for state implementation plan (SIP) submissions to EPA and for transportation
conformity analyses outside of California. The model is also the primary modeling tool for
estimating the impact of mobile source regulations on emission inventories.
This report describes the analyses conducted to generate exhaust emission rates and energy rates
representing exhaust emissions and energy consumption for heavy-duty vehicles in MOVES4 as
revised to support EPA rulemakings. Heavy-duty vehicles in MOVES are defined as any vehicle
with a Gross Vehicle Weight Rating (GVWR) above 8,500 lbs.
Emission rates for THC, CO, NOx, PM2.5 and NH3 are stored in the "EmissionRateByAge" table in
the MOVES database according to the following:
•	Pollutant
•	Emission process
•	Fuel type
•	Regulatory class
•	Model year group
•	Operating mode
•	Vehicle age
Energy emission rates are stored in the "EmissionRate" table, which is similar to the
"EmissionRateByAge" table, except emission rates are not differentiated by vehicle age. The
MOVES framework and additional details regarding the "EmissionRateByAge" and
"EmissionRate" table are discussed in the report documenting the rates for light-duty vehicles.9
In Section 1, we provide more background on the factors used to estimate heavy-duty exhaust
emissions in the "EmissionRateByAge" and "EmissionRate" tables. We then discuss the major
updates made to the heavy-duty emissions in MOVES4 in Section 1.8.
Sections 2 through 4 document the tailpipe exhaust emission rates for heavy-duty diesel, heavy-
duty gasoline, and heavy-duty compressed natural gas (CNG) vehicles. Section 5 documents the
crankcase emission rates used for each fuel type of heavy-duty vehicles. Section 7 documents the
methods used to estimate nitric oxide (NO), nitrogen dioxide (NO2), and nitrous acid (HNO2 or
HONO) emissions from NOx emissions using ratios.
1

-------
1.1 Pollutant
This report discusses the development of tailpipe exhaust emission rates for total hydrocarbons
(THC), carbon monoxide (CO), nitrogen oxides (NOx), fine particulate matter, defined as
particulate matter with mean aerodynamic diameter less than 2.5 microns (PM2.5), ammonia (NH3)
and energy consumption (in units of kJ).
Total hydrocarbons (THC) is the measurement of hydrocarbons from a flame ionization detector.1
From THC emissions, MOVES generates other estimates of hydrocarbon and organic gas missions,
including volatile organic compounds (VOCs), methane (CH4) and total organic gases (TOG).
MOVES then uses VOC emission rates to estimate individual toxic compounds such as
formaldehyde and benzene. The derivation of the factors used to compute aggregate measures of
organic gases and individual toxic emissions are available in the Speciation1 and Toxics2 MOVES
Reports.
MOVES reports PM2.5 emissions in terms of elemental carbon (EC) and the remaining non-
elemental carbon PM (nonECPM). This heavy-duty report covers the derivation of EC/PM
fractions used to estimate elemental carbon (EC), and the remaining non-elemental carbon PM
(nonECPM). MOVES also estimates 18 PM subspecies beyond elemental carbon, including
organic carbon, sulfate, nitrate, and other trace elements and ions through the use of speciation
profiles as documented in the Speciation Report.1
In MOVES4, we updated the heavy-duty NH3 emission rates, which are now documented in
Section 5. From NOx, MOVES estimates NO, NO2, and HONO emissions as documented in
Section 7. Nitrous Oxide (N2O) emissions from heavy-duty vehicles are documented in the
MOVES Greenhouse Gas and Energy Report.3 MOVES estimates CO2 emissions from the energy
rates documented in this report, using conversion factors, which are also documented in the
Greenhouse Gas and Energy Report.
In order to incorporate HD2027 final standards, we made updates for NOx, PM, HC, and CO
emission rates for model year 2027 and later. Refer to Section 1.8.2 for an overview of these
changes.
1.2 Emission Process
MOVES models vehicle emissions from fourteen different emission processes as listed in Table
1-1. This report covers the emission rates for the exhaust emission processes (running exhaust, start
exhaust, extended idle exhaust, auxiliary power exhaust, crankcase running exhaust, crankcase start
exhaust, and crankcase extended idle exhaust). We discuss the different processes below:
1.2.1 Running Exhaust
The running exhaust process occurs as the vehicle is operating on the road either under load or in
idle mode, and also includes "off-network idle. The running process is delineated by 23 operating
modes as discussed in Section 1.6.
2

-------
1.2.2 Start Exhaust
The start exhaust process is the incremental emissions that occur when starting a vehicle, including
the incremental emissions that occur after the engine start before the aftertreatment system is fully
functional. Ideally, start emission rates are calculated as the difference in emissions measured
between two otherwise identical drive cycles, where the first cycle includes a start and in the
second drive cycle the vehicle is already running and in a warmed-up condition as shown in
Equation 1-1.
Start emission rate (——)
start
Equation 1-1
= (emissions from drive cycle; with start
— emissions from same drive cycle; without start and with vehicle in warm condition)
Starts also have operating modes to characterize different amount of soak time (time since the
vehicle has last been running before being started again). Cold starts, (or starts after a long soak
period) generally have higher emission rates than warm starts (starts after a short soak period), due
to additional fueling needed due to increased condensation of fuel at colder engine temperatures,
and because the catalytic aftertreatment needs to reach a warm temperature to be fully operational.
MOVES defines eight operating start operating modes based on soak time as discussed in Section
1.6.
Operationally, we typically don't have two identical drive cycles that fit the conditions of Equation
1-1, and cold starts in this report are calculated using Equation 1-2 which is a reasonable
approximation since cold start emissions are typically much higher than hot-start emission.
g
Operationally — defined cold start emission rate (	) =
start
Equation 1-2
= (emissions from drive cycle; with cold start
— emissions from drive cycle; with hot start and vehicle in warm condition)
1.2.3 Extended Idle and Auxiliary Power Exhaust
The extended idle exhaust process in MOVES occurs during periods of hotelling, when long-haul
trucks are used during rest periods, such as when a vehicle is parked for the night and left idling.
Extended idle in generally defined to cover idling periods for longer than one hour. Extended idle
can result in different emissions than incidental idle that occurs during running operation because
the engine may be operated at a higher engine speed and the exhaust aftertreatment system may be
too cool to operate at its full efficiency.
Auxiliary power exhaust are emissions that come from diesel-powered generators that power the
truck's accessory loads, sometimes are used in place of the main engine during periods of hotelling.
In MOVES4, the scope of extended idle and APU emissions was expanded to include CNG and
electric (EV) combination long-haul trucks, as opposed to only diesel. MOVES does not allow
combination long-haul trucks to use the gasoline fuel type. Documentation of the extended idle and
auxiliary power exhaust emissions for heavy-duty diesel trucks are in Sections 2.3 and 2.4 while
documentation of CNG rates are in Section 4.3 and EV energy consumption rates are documented
in the Greenhouse Gas and Energy Consumption Rates Technical Report.3
3

-------
M0VES4 also includes the capability to model energy consumption for trucks that are plugged into
a facility's electricity to run their accessories, known as "shore power." Shore power energy
consumption rates are documented in the Greenhouse Gas and Energy Consumption Technical
Report.
1.2.4	Crankcase Exhaust
Crankcase exhaust emissions (for running, start, and extended idle) include combustion products
and oil that are vented from the engine crankcase to the atmosphere. Crankcase emissions are
estimated for THC, CO, NOx, PM2.5 emissions and their chained pollutants as discussed in Section
6. We do not estimate energy or CO2 emissions because crankcase emissions are a small
contribution to the total CO2 emissions. Crankcase emissions are significant sources of THC and
PM2.5 emissions from heavy-duty diesel engines. Crankcase emission rates for all four pollutants
for all heavy-duty source types and fuels are discussed in Section 6.
1.2.5	Evaporative and Brake and Tire Wear Emissions
Estimation of evaporative hydrocarbon emissions from heavy-duty gasoline vehicles is described in
the evaporative report.4 MOVES does not estimate evaporative emissions for diesel-powered
vehicles, but does estimate fuel spillage emissions which are part of the refueling emissions
documented in the evaporative report.4 Brake and tire wear emission rates from heavy-duty
vehicles are discussed in the Brake and Tire Wear Report.5
Table 1-1 Emission Processes for Onroad Heavy-Duty Vehicles
processID
processName
Covered
in this
report?
1
Running Exhaust
Y
2
Start Exhaust
Y
9
Brakewear
N
10
Tirewear
N
11
Evap Permeation
N
12
Evap Fuel Vapor Venting
N
13
Evap Fuel Leaks
N
15
Crankcase Running Exhaust
Y
16
Crankcase Start Exhaust
Y
17
Crankcase Extended Idle Exhaust
Y
18
Refueling Displacement Vapor
Loss
N
19
Refueling Spillage Loss
N
90
Extended Idle Exhaust
Y
91
Auxiliary Power Exhaust
Y
4

-------
1.3 Fuel Type
This report is primarily organized around the exhaust emission rates by fuel type: heavy-duty diesel
(Section 2), heavy-duty gasoline (Section 3), and heavy-duty CNG (Section 4). These three fuel
types can be modeled in all the heavy-duty sourcetypes, with the exception of long-haul
combination trucks which can only model diesel vehicles. Note that the emissions from the heavy-
duty sector predominantly come from diesel vehicles and the majority of the data analyzed were
from diesel vehicles. MOVES also models E85 for light-duty vehicles, but this fuel type is not
available to be modeled for heavy-duty vehicles.6 And MOVES models energy consumption from
battery electric and fuel cell electric vehicles as explained in the MOVES4 GHG and Energy
report.3
1.4 Regulatory Class
The MOVES regulatory classes group vehicles that have similar emission rates. The MOVES
heavy-duty regulatory classes are largely determined based on gross vehicle weight rating (GVWR)
classifications, because the heavy-duty emission standards are based on GVWR as shown in Table
1-2.
There are additional criteria that define the heavy-duty regulatory classes in MOVES. Urban Bus
vehicles are distinguished from other heavy heavy-duty vehicles (GVWR >33,000 lbs.) because
they have tighter PM emission standards for the 1994 through 2006 model years.7 Urban bus is a
regulatory class that is defined by its intended use as well as the GVWR. EPA regulations define
urban buses as "heavy heavy-duty diesel-powered passenger-carrying vehicles with a load capacity
of fifteen or more passengers and intended primarily for intra-city operation."8.
In MOVES, gliders (regClassID 49) are defined as heavy heavy-duty diesel trucks with an old
powertrain, combined with a new chassis and cab assembly. Currently in MOVES, gliders are
limited to the diesel long-haul and short-haul combination truck source types. As discussed in
Section 2.5, the emissions are equivalent to MY 2000 HHD diesel vehicles.
In MOVES4, we classified diesel light-heavy-duty Class 3 engine-certified vehicles in model year
2017 and later years as LHD45 vehicles. The emission rates for LHD2b3 vehicles are based on the
assumption that all vehicles are chassis-certified, and subject to the Tier 3 standards (Discussed in
Section 2.1.1.5.5), and the Heavy-duty Phase 1 and 2 greenhouse gas emission standards
(Discussed in Section 2.1.4.3.2). Because Class 3 engine-certified vehicles are instead subject to
the same emission standards as Class 4 and 5 engine-certified vehicles, we reclassified these
vehicles as LHD45 vehicles. Model year 2017 is selected because this is the first model year the
emission rates are different between LHD2b3 and LHD45. The reclassification of diesel LHD2b3
as diesel LHD45 vehicles is documented in the Population and Activity report.6
MOVES classifies vehicles of similar activity, usage patterns, and body type into source use types,
often simply referred to as "source types." The MOVES source types are defined in the Population
and Activity Report.6 As shown in Table 1-2, vehicles of a regulatory class may be mapped to
multiple source types. Likewise, each source type used to model heavy-duty vehicles includes
population from several different regulatory classes. For example, single unit short-haul trucks
(sourceTypelD 52), include vehicles from the following regulatory classes: LHD2b3, LHD45,
MHD, and HHD.
5

-------
Regulatory Class
Description
regClassName
regClassID
Gross Vehicle
Weight
Rating
(GVWR) [lb.]
Existing Source Types in
default database
(sourceTypelD)
Light Heavy-Duty
Class 2b and 3
trucks
LHD2b3
41
8,501 -
14,000*
Passenger Trucks (31), and
Light-Commercial Trucks
(32), School Buses (43), and
Single Unit Trucks (51, 52, 53,
54)
Light Heavy-Duty
Class 4 and 5
Trucks
LHD45
42
14,001 -
19,500*
Buses (41, 42, 43) and Single
Unit Trucks (51,52,53,54)
Medium Heavy-
Duty (Class 6 and
7 Trucks)
MHD
46
19,501 -
33,000
Buses (41,42,43), Single Unit
Trucks (51, 52, 53, 54), and
Combination Trucks (61, 62)
Heavy Heavy-
Duty (Class 8
Trucks)
HHD
47
> 33,000
Buses (41, 42, 43), Single Unit
Trucks (51, 52, 53, 54), and
Combination Trucks (61, 62)
Urban Bus
Urban Bus8
48
>33,000
Transit Bus (42)
Gliders (Class 8
Trucks)
Glider Vehicles
49
>33,000
Combination Trucks (61, 62)
*Model year 2017-and-later engine-certified Class 3 (GVWR 10,001-14000 lbs) trucks (only present within source
types 52, 53, and 54) are classified as LHD45 (regclassID 42).
1.5	Model Year Groups
MOVES model year groupings are designed to represent major changes in emission rates due to
changing vehicle and aftertreatment technologies introduced in response to EPA emission
standards. Model year groups in MOVES can represent a single model year (e.g., 2007), or a range
of model years (e.g., 2030-2060). The emission rates discussed in the following sections are
discussed in terms of model year ranges. The model year groups cover all the model years between
1960 through 2060. When data are limited or unavailable for model year groups, we make
assumptions about the impact of emission standards and vehicle aging to estimate those emission
rates.
1.6	Operating Modes
Emission rates in MOVES are stored by regulatory class, fuel type, model year, and operating
mode. To calculate emissions from each process, MOVES sums the product of the emission rate
for each operating mode by the time spent in each operating mode.
For example, the activity basis for running process is source hours operating (SHO). The running
process is divided into 23 operating modes (as shown in Table 1-4). Using Equation 1-3, the total
running emissions is calculated by summing the product of the emission rates with the fraction of
time spent in each operating mode (the operating mode distribution). This is multiplied by the total
hours (SHO) spent in this emission process.
6

-------
Running Emissions
= SHO
23
x Operating k
i=1
x Emission Ratet)
^(Operating Mode Distribution.
Equation 1-3
MOVES performs the calculations shown in Equation 1-3, at a detailed level that accounts for each
factor that impacts the emission rates (e.g., model year, vehicle age, fuel type, regulatory class) and
the operating mode distribution (source type, roadtype, average speed (which varies across hour of
the day)). Then, the emissions can be aggregated to different levels (e.g., by sourcetype). Similar
equations can be constructed for other process, the equation for starts is shown in Equation 1-4,
where the starts are classified into eight operating modes (Table 1-5).
Start Emissions
= Starts
The operating modes for running exhaust are defined in terms of power output (with the exception
of the idle and braking modes). For light-duty vehicles, the parameter used is known as vehicle-
specific power (VSP), which is calculated by normalizing the continuous power output for each
vehicle to its own weight as shown in Equation 1-5 (hence the term "vehicle-specific"). As
discussed in the light-duty emission rate report,9 VSP is a robust predictor of vehicle emissions. In
the laboratory, light-duty vehicles are tested on full chassis dynamometers, and emission standards
are in units of grams per mile. The emission standards are largely independent of the weight (and
other physical characteristics) of the vehicle.
In developing emission rates for MOVES, light-duty emissions data from individual vehicles are
assigned to VSP operating mode bins using Equation 1-5, using the individual vehicle's measured
weight as the source mass, and ideally using vehicle-specific road load coefficients. In contrast,
when MOVES calculates VSP from driving cycles and assigns operating modes for an entire
source type, the average source type mass and average road load coefficients are used instead.
Equation 1-4
i=l
x Emission Ratet)
7

-------
Avt + Bvf + Cv} + m ¦ vt(at + a ¦ sin6t)	„ _
VSPt =	-	-	-	 	_	_	Equation 1-5
m
Where:
VSPt = vehicle specific power at time t [kW/ton]
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],
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 8t = the (fractional) road grade at time t
For heavy-duty vehicles, we classify running exhaust using Scaled Tractive Power (STP) as shown
in Equation 1-6 using road-load coefficients. STP is equivalent to VSP, except the power for all
vehicles within the same regulatory class and model year are scaled using a fixed mass factor,
rather than the individual weight of the vehicle. The fscaie is used to bring the numerical range of
tractive power from heavy-duty vehicles into the same numerical range as the VSP values when
assigning operating modes. When developing emission rates for MOVES, operating modes are
assigned to individual vehicles using both the individual truck mass, m, and the common fscaie
value used for all heavy-duty vehicles from the same regulatory class, source type and model year
group. Because a common fscaievalue is used, individual vehicles assigned to the same STP-
defined operating mode bin are producing the same absolute tractive power, regardless of
differences in their individual source masses.
Avt + Bv} + Cv} + m ¦ vt(at + q ¦ sin6t)
STPt =							—			—	Equation 1-6
fscaie
Where:
STPt = the scaled tractive power at time t [scaled kW or skW]
fscaie = fixed mass factor (see Table 1-3)
Other variables as previously defined in Equation 1-5
When MOVES estimates STP and assigns operating mode distributions for the heavy-duty source
types, Equation 1-6 uses the average source type mass (m) for each regulatory class, source type,
and model year group in the numerator and uses the common fscaie value for the regulatory class
and model year group which was also used in the emission rate analysis. At County and Default
National Scale, MOVES uses the instantaneous speed (vi) from the second-by-second driving
cycles associated with average speed, source types, and road types. For Project Scale, MOVES can
use user-supplied driving cycles and grade. The default average speed driving cycles and the load
road coefficients are discussed in the Population and Activity Report.6
The equation for STP is generalized below in Equation 1-7, with units in scaled kW or skW:
STP = —Equation 1-7
fscale
Where: Paxk is the power demand at the axle for the heavy-duty truck.
8

-------
As presented in Equation 1-6, Paxie can be estimated using the road-load coefficients from chassis-
dynamometer tests (for example, pre-2010 heavy-duty diesel THC, CO, and PM2.5 emission rates
in Sections 2.1.2 and 2.1.3). Road-load coefficients can also be used to estimate power demand
from onroad tests when more accurate power demand measurements are not available (for example,
the 2010+ heavy-duty gasoline emission rates in Section 3.1).
The Paxie can be estimated from an engine dynamometer or from an engine control unit (ECU) for
chassis or onroad testing by measuring the engine power and estimating the accessory loads and
powertrain efficiencies for the vehicle.
For onroad tests, measuring power from the ECU is generally more accurate than estimating power
from road-load coefficients. Unlike a generic road-load equation where vehicle characteristics, such
as aerodynamic drag and rolling resistance are assumed, the ECU measures engine speed and
calculates torque directly during the test, avoiding the need to capture the impact of wind speed and
wind direction, as well as weight and the road grade throughout the entire test cycle and route.
Wind can have a significant impact on power needs, and the payload of heavy-duty vehicles can be
greater than the vehicle weights itself, while also varying significantly over the test. Thus, for
onroad tests, we generally use power calculated from the ECU measurements, because the vehicle
and environmental characteristics determine the axle power (Section 2.1.1.3).
The use of STP instead of VSP is preferable for modeling heavy-duty vehicles emissions because
heavy-duty vehicle emissions are strongly correlated with power output. Heavy-duty vehicles are
regulated using engine dynamometer tests, and emissions standards are in units of grams per brake-
horsepower-hour (g/bhp-hr). Additionally, each heavy-duty regulatory class contains a wide variety
of truck sizes, truck weight, power ratings, and in-use payloads. Using STP, we can scale the
heavy-duty emission rates to different power outputs that were not measured in our emission rate
database. The sample of trucks we used to develop emission rates for each regulatory class has a
limited number of trucks and loads compared to the in-use fleet, which may not be representative
of the average vehicle weight and power output of the in-use fleet. The use of VSP would require
the sample of vehicles to match the average vehicle weights and load to accurately estimate average
in-use emission rates. By using STP, MOVES scales the measured emissions to match the
estimated weight and power output of the modeled in-use fleet.
The fscale values can be considered as a surrogate for the average mass of heavy-duty vehicles
within each regulatory class, as reported in Table 1-3. For the pre-2010 emission rates, we used an
fscale equivalent to the average mass of the light-commercial trucks (2.06 metric tons) for the
LHD2b3 and LHD45 emission rates. For the other heavy-duty source types, a single fscaie (17.1
metric tons) was used, which provided emissions within each operating mode bin for the largest
heavy-duty truck vehicles.
However, in our analysis of in-use data from recent trucks, we found that an fscaie of 17.1 metric
tons limited most of the real-world activity to low and medium power bins within a speed-bin,
especially for the light-heavy and medium-heavy-duty regulatory classes. In MOVES3 (and carried
over to MOVES4), we revised the fscaie values for MY 2010 and later to provide more resolution in
the fcaie by regulatory class. Derivation of the new fcaie values is described in Appendix G.
9

-------
Table 1-3 Fixed Mass Factor, fSCaie (metric tons)
Regulatory Class (regClassID)
MY 1960-2009
MY 2010+
LHD2b3 (41)
2.06
5.0
LHD45 (42)
2.06
5.0
MHD (46)
17.1
7.0
HHD (47)
17.1
10.0
Urban Bus (48)
17.1
10.0
Glider (49)
17.1
17.1
NOTE: OpMode-based emission rates CANNOT be compared directly between regClasses or model years (MYs) with
different fscau values. For example, the OpMode 14 emission rates for MY 2012 MHD (f,caie = 7) cannot be directly
compared to the same OpMode rates from MY 2009 MHD (fscau = 17.1) or MY 2012 HHD (fscau = 10). That is because
data assigned to an OpMode based on different fscau values will have different absolute power (numerator of Equation
1-6 and Equation 1-7). When using vehicle mass in the denominator (Equation 1-5), this is not an issue because the unit
is kW/ton and the power is normalized to the mass of the vehicle. However, when using fscaie in the denominator, as is
the case for all heavy-duty vehicles, the unit is scaled kW and there is no normalization to a physical quantity.
The operating modes bins for running exhaust are shown in Table 1-4.
10

-------
Table 1-4 Operating Mode Definition for Running Exhaust for Heavy-Duty Vehicles
OpModelD
Operating Mode
Description
Scaled Tractive
Power (STP*, skW)
Vehicle Speed
(vt, mph)
Vehicle Acceleration
including grade
(mph/sec)1
0
Deceleration/Braking2


at +g-sin(8t) ^ -2.0 OR
[at +g-sin(8t) < -1.0
AND
flu +g-sin(0t.I) < -1.0
AND
at.2 +g-sin(0t-2) < -1.0)
1
Idle

vt< 1.0

11
Coast
STP,< 0
1 < vt < 25

12
Cruise/Acceleration
0 < STP;< 3
1 < vt < 25

13
Cruise/Acceleration
3 < STP;< 6
1 < vt < 25

14
Cruise/Acceleration
6 < STP;< 9
1 < vt < 25

15
Cruise/Acceleration
9 < STP/< 12
1 < vt < 25

16
Cruise/Acceleration
12 < STPt
1 < vt < 25

21
Coast
STP,< 0
25
-------
Table 1-5 Operating Modes for Start Emissions (as a function of soak time)
Operating Mode
Description
101
Soak Time < 6 minutes
102
6 minutes < Soak Time <30 minutes
103
30 minutes < Soak Time < 60 minutes
104
60 minutes < Soak Time < 90 minutes
105
90 minutes < Soak Time <120 minutes
106
120 minutes < Soak Time < 360 minutes
107
360 minutes < Soak Time < 720 minutes
108
720 minutes < Soak Time
Extended idle exhaust and diesel auxiliary power unit (APU) exhaust are each modeled in MOVES
with a single operating mode (opModelDs 200 and 201, respectively).
1.7 Vehicle Age
In MOVES, the start and running emission rates for THC, CO, NOx, and PM2.5 are stored in the
"emissionRateByAge" table by age group, meaning that different emission rates can be assigned to
different aged vehicles of the same model year, regulatory class, fuel type and operating mode.
MOVES uses six different age classes to model the age effects, as shown in Table 1-6. The effects
of age on the emission rates were developed separately for gasoline and diesel vehicles.
Table 1-6 MOVES Age Group Definitions

Lower
Upper
ageGroupID
bound
bound

(years)
(years)
3
0
3
405
4
5
607
6
7
809
8
9
1014
10
14
1519
15
19
2099
20
None
For diesel running exhaust, we estimated the effects of tampering and mal-maintenance (T&M) on
emission rates as a function of age. Tampering refers to intentional disabling or modifying the
vehicle engine, control systems, and/or exhaust aftertreatment systems that results in increased
emissions. Mal-maintenance refers to lack or improper maintenance of the engine and
aftertreatment, including neglecting to repair broken or mal-functioning engine and aftertreatment
parts, which increase emissions. Based on surveys and studies, we developed estimates of
frequencies and emission impacts of specific emission control component malfunctions, and then
aggregated them to estimate the overall emissions effects for each pollutant (see Appendix B). We
adopted this approach due to the lack of adequate data to directly estimate the deterioration for
heavy-duty vehicles, and because the effects of T&M are believed to be the dominant source of
emission deterioration on fleet-wide heavy-duty diesel emissions. MOVES currently does not allow
12

-------
explicit modeling of tampered vehicles - we hope to address this in future versions of MOVES as
more data on tampering become available.
Start emissions are generally a small contributor to total exhaust emissions from heavy-duty diesel
vehicles. No T&M effects are currently applied to the diesel start emissions.
For gasoline heavy-duty vehicles, the age effects are estimated directly from the a sample of
emissions data, or are adopted from light-duty deterioration as discussed in Section 3.1.1.1. These
effects are applied to both running and start exhaust emissions.
Not all emission rates are distinguished by vehicle age. Rates with no age dependence in MOVES
are stored in the "EmissionRate" table. This table includes energy consumption rates for all
processes as well as THC, CO, NOx, PM2.5, and ammonia (NH3) emission rates for extended idle
and auxiliary power units (APU), nitrous oxide (N2O) rates for start and running emissions, and tire
and brake wear emission rates. We calculate T&M effects for the diesel extended idle emission
rates without regard to vehicle age as discussed in Section 2.3.2.3. This report documents the THC,
CO, NOx, and PM2.5 emissions from extended idle and APU usage, and heavy-duty ammonia
emissions for all emission processes. However, heavy-duty nitrous oxide3 and tire and brake wear5
emission rates are documented in separate MOVES reports.
1.8 MOVES4 Updates
This section provides an overview of the updates to heavy-duty emission rates in MOVES4.
1.8.1 General Updates
MOVES4 updates include:
•	Application of Tier 3 for LHD2b3 diesel vehicles
•	Heavy-duty ammonia (NH3) emission rates
•	Crankcase emissions
•	NOx composition for heavy-duty diesel vehicles
•	Tampering and mal-maintenance effects for LHD vehicles due to a minor correction in
warranty and useful life assumptions.
•	Due to a 2021 appeals court ruling vacating the portions of the HDGHG2 rule that apply to
trailers,10 for combination trucks of model year 2018 and later, we revised MOVES inputs
that describe weight, aerodynamics, and rolling resistance as explained in the vehicle
population and activity report6 and "other efficiency" improvements as described in Section
2.1.4.3 below. This slightly increases the modeled emissions of CO2 and other pollutants
from these trucks.
•	We now expect the Phase 2 requirements for LHD2b3 to be met via electrification starting
in MY2025. Therefore, in MOVES4, we only model reductions in diesel energy rates
through MY2024. See Section 2.1.4.3.2.
13

-------
While most of these updates were routine, we conducted a peer-review of the updated ammonia
emission rates. Materials from peer review, including peer-review comments and EPA responses
are located on the EPA's science inventory webpage. u'12
1.8.2 Updates to incorporate HD2027 Standards
We made additional updates in MOVES4 to incorporate "Control of Air Pollution from New Motor
Vehicles: Heavy-Duty Engine and Vehicle Standards" (also known as the "HD2027 standards")
finalized on December 20, 2022.13 This rule sets more stringent NOx, PM, HC, and CO emissions
standards for heavy-duty vehicles and engines starting in model year 2027.
To account for the impact of the HD2027 standards, we updated the MY2027+ heavy-duty vehicle
exhaust emission rates in MOVES4 for the LHD45, MHD, HHD and Urban Bus regulatory classes,
with focus on:
•	Heavy-duty diesel vehicles:
o Updated NOx emission rates for running, start, and extended idle processes in
response to the HD2027 duty-cycle and off-cycle standards (as described in Sections
2.1.1.6, 2.2.1.3, and 2.3.3)
o Revised running emission rates for NOx, PM2.5, THC, and CO due to the changes to
the regulatory useful life and warranty, by modifying MOVES tampering and mal-
maintenance (T&M) calculations (as described in Sections 2.1.1.7, 2.1.2.3.3, 2.1.3.3,
and Appendix B)
o Revised crankcase emission rates to consider the impacts of the closed crankcase
design option available in the rule (as described in Sections 6.2.2.4)
•	Heavy-duty gasoline vehicles:
o Revised NOx, THC, CO and PM2.5 emission rates for running processes only (as
described in Sections 3.1.1.3.2 and 3.1.2.3)
•	Heavy-duty natural gas vehicles:
o No updates were made since the average NOx FTP emission level for MY 2010-
2017 CNG engine families is already close to the 0.1 g/hp-hr standard and any
further reductions due to the rule are expected to be small.
There were no additional updates affecting heavy-duty gasoline and natural-gas vehicles.
2 Heavy-Duty Diesel Exhaust Emissions
This section details our analysis to develop emission rates for heavy-duty diesel vehicles. Four
emission processes (running, extended idling, starts, and auxiliary power unit exhaust) are
discussed.
2.1 Running Exh aust Emissions
The analysis for running exhaust emissions requires accurate second-by-second measurements of
emission rates and parameters that can be used to estimate the tractive power exerted by a vehicle.
This section describes how we analyzed continuous second-by-second heavy-duty diesel emissions
data to develop emission rates applied within the predefined set of operating modes (Table 1-4).
14

-------
Stratification of the data sample and generation of the final MOVES emission factors was done
according to the combination of regulatory class (shown in Table 1-2) and model year group. As
mentioned in Section 1.6, the emission rates were developed using scaled-tractive power (STP),
using the power scaling factors shown in Table 1-3.
2.1.1 Nitrogen Oxides (NOx)
For NOx rates, we stratified heavy-duty vehicles into the model year groups listed in Table 2-1.
These groups were defined based on changes in NOx emissions standards and the outcome of the
Heavy-Duty Diesel Consent Decree14, which required additional control of NOx emissions during
highway driving for model years 1999-and-later. This measure is referred to as the "Not-to-
Exceed" (NTE) limit.
Table 2-1 Model Year Groups for NOx Analysis Based on Emissions Standards
Model year group
Standard (g/bhp-
hr)
NTE limit (g/bhp-hr)
Pre-1988
None
None
1988-1989
10.7
None
1990
6.0
None
1991-1997
5.0
None
1998
4.0
None
1999-2002
4.0
7.0 HHD; 5.0 other reg. classes
2003-2006
2.41
1.5 times the standard or family emission limit
(PEL)
(or 1.25 standard or FEL, when FEL > 1.50 g/bhp-
hr)
2007-2009
1.21'2
2010-2026
0.2
2027+
0.035 for LHD,
0.05 for
MHD/HHD
The off-cycle NTE standards was replaced by a
moving average window (MAW) approach in the
HD2027 rule.
Notes:
1	NMHC+NOx Standard
2	Assumes Phase-in of NOx standard
2.1.1.1 Data Sources
In modeling NOx emissions from HHD, MHD, LHD, and urban buses, we relied on the following
data sources:
ROVER. This dataset includes measurements collected during onroad operation using the ROVER
system, a portable emissions measurement system (PEMS) developed by the EPA. The
measurements were conducted by the U.S. Army Aberdeen Test Center on behalf of U.S. EPA.15
This program started in October 2000. Due to time constraints and data quality issues, we used only
data collected from October 2003 through September 2007. The data was compiled and reformatted
for MOVES analysis by Sierra Research.16 EPA analyzed the data and developed the emission
rates. The data we used represents approximately 1,400 hours of operation by 124 trucks and buses
of model years 1999 through 2007. The vehicles were driven mainly over two routes:
•	"Marathon" from Aberdeen, Maryland, to Colorado and back along Interstate 70
•	Loop around Aberdeen Proving Grounds in Maryland
15

-------
Consent Decree Testing. These data were conducted by West Virginia University using the
Mobile Emissions Measurement System (MEMS).17'18'19 This program was initiated as a result of
the consent decree between the several heavy-duty engine manufacturers and the US government,
requiring the manufacturers to test in-use trucks over the road. Data was collected from 2001
through 2006. The data we used represented approximately 1,100 hours of operation by 188 trucks
of model years 1994 through 2003. Trucks were heavily loaded and tested over numerous routes
involving urban, suburban, and rural driving. Several trucks were re-acquired and tested a second
time after 2-3 years. Data were collected at 5-Hz frequency, which we averaged around each
second to convert the data to a 1.0-Hz basis.
Heavy-Duty Diesel In-Use Testing (HDIUT). The manufacturer-run in-use testing program for
heavy-duty diesel vehicles was promulgated in June 2005 to monitor the emissions performance of
heavy-duty engines operated under a wide range of real world driving conditions within the
engine's useful life.20 It requires each manufacturer of heavy-duty highway diesel engines to assess
the in-use exhaust emissions from their engines using onboard, portable emissions measurement
systems (PEMS) during typical operation while on the road. The PEMS unit must meet the
requirements of 40 CFR 1065 subpart J. The in-use testing program began with a mandatory two-
year pilot program for gaseous emissions in calendar years 2005 and 2006. The fully enforceable
program began in calendar year 2007 and is ongoing. The vehicles selected for participation in the
program are within the engine's useful life, and generally, five unique vehicles are selected for a
given engine family. This dataset includes results for HHD, MHD, and LHD vehicles. The HDIUT
data are publicly available on EPA's website.21
The data available for use in MOVES2014 were collected during calendar years 2005 through 2010
and represent engines manufactured in model years 2003 to 2009. For MOVES3 and later versions,
we evaluated data from engines selected for testing in calendar years 2010 through 2018. These
engines cover model years 2010 to 2016. The MY 2010+ data set included 40 unique engine
families and 372 vehicles. There are about 10 million seconds of quality-assured second-by-second
data covering about 85,000 miles of instrumented travel. 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.22
For the HHD class, out of a total 159 vehicles selected for testing during 2010-2016, 109 were line-
haul, 46 were delivery, and the remaining were marked as "Other" in the metadata. Since the
HDIUT data is measured and submitted by the manufacturer and the test vehicles are required to be
free of any tampering or mal-maintenance, we can safely assume that they represent zero-mile
vehicles for the purpose of assigning base rates and applying the tampering and mal-maintenance
effects. We have expanded the characterization of the MY 2010+ HDIUT data set by separating the
0.2 NOx FEL group for heavy-duty diesel vehicles into two model year groups (2010-2013 and
2014 and later) as described in 2.1.1.5.
Houston Drayage Data. In coordination with the Texas Commission on Environmental Quality,
the Houston-Galveston Area Council, and the Port of Houston Authority, EPA conducted a study
collecting emissions data from trucks in drayage service using portable emission measurement
systems (PEMS) from December 2009 to March 2010.23 The trucks studied were diesel-fueled,
heavy heavy-duty trucks used to transport containers, bulk and break-bulk goods to and from ports
and intermodal rail yards to other locations. These trucks conducted the majority of their travel on
16

-------
short-haul runs, repeatedly moving containers across fixed urban routes. Note that only small
fractions of trucks involved in drayage service are dedicated solely to this function, with most
trucks spending large fractions of their time performing other types of short-haul service. No
specific drive cycles were used and all PEMS testing was based on actual in-use loads and speeds.
A summary of vehicles by model years for the above-mentioned datasets is provided in Table 2-2.
Table 2-2 Numbers of Diesel Vehicles from the ROVER, WVU MEMS, HDIUT, and Houston Drayage


Regulatory Class
Data Source
MYG
HHD
MHD
LHD1
BUS
ROVER and
Consent Decree
Testing
1991-1997
19
-
-
2
1998
12
-
-
-
1999-2002
78
30
-
25
2003-2006
91
32
-
19
HDIUT
2003-2006
40
25
15
-
2007-2009
68
71
24
-
2010-2016
194
74
94
10
Houston Drayage
1991-1997
8
-
-
-
1998
1
-
-
-
1999-2002
10
-
-
-
2003-2006
8
-
-
-
Note:
1 LHD45 and LHD2b3 vehicles were grouped together and classified as LHD vehicles for analysis of rates. The only
difference in the LHD45 and LHD2b3 emission rates is the application of production volume weighting and T&M
effects summarized in Table 2-16.
For the pre-2010 emission rates, we used a combination of the ROVER, Consent Decree Testing
and the HDIUT testing to estimate the emission rates. The Houston Drayage Study was used as
comparison study only. Additional details are provided in Section 2.1.1.4. The HDIUT data are
used exclusively for the 2010-2026 model year emission rates as discussed in Section 2.1.1.5.
These 2010-2026 rates were also used as the "base rates" to develop the MY2027+ emission rates
based on the duty-cycle standard as explained in Section 2.1.1.6.1.
2.1.1.2 Measurement Accuracy and Quality Assurance
PEMS devices continue to make improvements that affect measurement accuracy, such as sensor
response, sample conditioning, and noise reduction. The data sets represent the accuracy of the
instruments at the time of measurement. In compliance determinations, when determining whether
the tested vehicle meets the in-use emissions standard or not, an "accuracy margin for portable in-
use equipment" (commonly referred to as measurement allowance) is added onto the standard;
increasing the vehicle compliance margin. The accuracy margins vary by model year and type of
measurement method and are described in 40 CFR 86.1912. This is done to prevent measurements
that are biased-high from affecting the compliance decision. However, since the true value for each
second of data is unknown and errors could be biased either high or low, the in-use emission rates
used in MOVES from each of these data sets are not adjusted to reflect the measurement allowance.
17

-------
From each data set, we used only tests we determined to be valid. For the ROVER dataset, we
eliminated all tests with any reported problems, including GPS malfunctions, PEMS malfunctions,
etc., whether or not they affected the actual emissions results. For HDIUT data for MY2009 and
earlier and Houston Drayage, the time-alignment was visually confirmed by comparing relevant
time-series plots, such as exhaust mass-flow rate vs. CO2 concentration, and exhaust-mass flow rate
vs. engine speed, as recorded by the engine control unit (ECU). Data was generally aligned within
one second. When an issue with the time-alignment was found, efforts were made to realign the
data as much as possible. As our own high-level check on the quality of PEMS and ECU output,
we then eliminated any trip from ROVER, HDIUT (MY 2009 and earlier), and Houston Drayage
where the Pearson correlation coefficient between CO2 (from PEMS) and engine power (from
ECU) was less than 0.6. The correlation check removed approximately 7 percent of the ROVER
and HDIUT (MY 2009 and earlier) data. All the data from Houston Drayage met the criteria for
correlation between CO2 and engine power. In addition, data were excluded from the analysis when
the vehicle speed was not available due to GPS and/or ECU malfunctions, when no exhaust flow
was reported, and when a periodic zero correction was being performed on gas analyzers. For the
WVU (West Virginia University) MEMS data, WVU itself reported on test validity under the
consent decree procedure and no additional detailed quality checks were performed by EPA. Table
2-2 shows the total distribution of vehicles by model year group from the emissions test programs
above following evaluation of the validity of the data.
In analyzing the HDIUT data for MY 2010+, we checked the time-alignment and deleted any
second of data that met any of the following conditions: (1) instrument was undergoing zeroing, as
marked by the zero flag field; (2) engine RPM was below 500; (3) vehicle speed was missing or
below zero; (4) acceleration was missing; (5) engine torque was missing; (6) measured exhaust
flow rate was missing, or less than or equal to zero; and (7) as catch-all, if the calculated STP and
OpMode were invalid numbers. We did not verify the accuracy of exhaust flow rate measurement
and CO2 measurement using techniques such as carbon-balance versus ECU reported fuel rate data.
Such verifications are assumed to have been done (by the manufacturer) before data is submitted to
EPA since they are required by 40 CFR 1065 subpart J.
2.1.1.3 Calculation of Operating Modes
As discussed in Section 1.6, we prefer to estimate tractive power from engine data collected during
real-world operation rather than using the road-load equation. To do so, we first identified the
seconds in the data that the truck was either idling or braking based on acceleration and speed
criteria shown in Table 1-4. For all other operation, engine speed (Deng and torque Teng from the ECU
were used to determine engine powering, as shown in Equation 2-1.
Peng =COengTeng	Equation 2-1
We then determined the relationship between the power required at the wheels of the vehicle and
the power required by the engine. We first had to account for the losses due to accessory loads
during operation. Some accessories are engine-based and are required for operation. These include
the engine coolant pump, alternator, fuel pump, engine oil pump, and power steering. Other
accessories are required for vehicle operation, such as cooling fans to keep the powertrain cool and
air compressors to improve braking. The third type of accessories is discretionary, such as air
18

-------
conditioning, lights, and other electrical items used in the cab. None of these power loads are
subtracted in the engine torque values that are output from the engine control unit.
We calculated accessory load requirements and mapped them to STP bins in a number of steps.
First, we grouped the accessories into five categories: cooling fan, air conditioning, engine
accessories, alternator (to run electrical accessories), and air compressor. We identified where the
accessories were predominately used on a vehicle speed versus vehicle load map to properly
allocate the loads. For example, the cooling fan will be on at low vehicle speed where the forced
vehicle cooling is low and at high vehicle loads where the engine requires additional cooling. The
air compressor is used mostly during braking operations; therefore, it will have minimal load
requirements at high vehicle speeds. Table 2-3 identifies the predominant accessory use within
each of the vehicle speed and engine load map areas.
At this point, we translated the vehicle speed and engine load map into engine power levels. The
engine power levels were aggregated into low (green), medium (yellow) and high (red) as
identified in Table 2-3. Low power means the lowest third, medium is the middle third, and high is
the highest third, of the engine's rated power. For example, for an engine rated at 450 hp, the low
power category would include operation between 0 and 150 hp, medium between 150 and 300 hp,
and high between 300 and 450 hp. So, for example, when vehicle operation is in middle of the
engine load map and vehicle speed is low or mid speed, the engine power level is in medium
(yellow) band and the active accessory loads are as listed in the respective cells. However, for the
same engine load map operation (mid) with vehicle speed at high, the engine power level is high
(red) and active accessory loads are as listed in the cell. Some accessory loads, such as cooling fan,
are absent from cells with the same engine power level (identified by color) based on the reasons
given in the previous paragraph.
Table 2-3 Accessory Use as a Function of Speed and Load Ranges, Coded by Power Level


Vehicle Speed (mph)


Low (0-25)
Mid (25-50)
High (50+)


Cooling Fan
Air Cond.
Air Cond.
Air Cond.
Ph
Low
Engine Access.
Alternator
Engine Access.
Alternator
Engine Access.
Alternator
E

Air Compress
Air Compress

ii
o
J

Cooling Fan
Air Cond.
Cooling Fan
Air Cond.
Air Cond.
1)
s
'5b
Mid
Engine Access.
Alternator
Engine Access.
Alternator
Engine Access.
Alternator
W

Air Compress
Air Compress



Cooling Fan
Air Cond.
Cooling Fan
Air Cond.
Cooling Fan
Air Cond.

High
Engine Access.
Alternator
Air Compress
Engine Access.
Alternator
Air Compress
Engine Access.
Alternator
Next, we estimated the power required when the accessory was "on" and percentage of time this
occurred. The majority of the load information and usage rates are based on information from "The
Technology Roadmap for the 21st Century Truck,"24
19

-------
The total accessory load is equal to the power required to operate the accessory multiplied by the
percent of time the accessory is in operation. The total accessory load for each cell in Table 2-3 is
equal to the sum of each accessory load. The calculations are included in Appendix A. The total
accessory loads Pbss.acc listed below in Table 2-4 are subtracted from the engine power determined
from Equation 2-1 to get net engine power available at the engine flywheel.
For pre-2010 model years, LHD losses were set to zero. The losses for MY 2010+ LHD vehicles
were estimated by adjusting the MHD vehicle losses as such: (1) removed the loss for air
compressor; (2) no change to air condition loss; (3) scaled the losses for cooling fan, alternator, and
engine accessories by 5/7 (where 5 and 7 are rough estimates of the average engine displacement in
liters for LHD and MHD engines, respectively). Based on these adjustments, the LHD losses are
estimated to be approximately 60 percent of MHD losses for the low power band and 70 percent
for the mid and high power bands. We acknowledge this calculation relies on a number of
assumptions, but we believe it is a step forward from having LHD losses equal to zero for all model
years (as was the case in MOVES2014).
Table 2-4 Estimates of Accessory Load in kW by Engine Power Level
Engine Power Level
HHD
MHD
LHD1
LHD1
Urban
(of rated power)
(pre-2010)
(2010+)
Bus
Low (0 - l/3rd)
8.1
6.6
0.0
4.1
21.9
Medium (l/3rd to
2/3rd)
8.8
7.0
0.0
4.8
22.4
High (2/3rd to 1)
10.5
7.8
0.0
5.5
24.0
Note:
1 In MOVES2014, the accessory load losses for LHD were assumed to be zero for all model years. In MOVES3 and
later models, MY 2010+ LHD data (Table 2-2) is analyzed with non-zero accessory load losses. However, for pre-
2010 MY LHD, we continue to assume zero accessory load loss.
We then adjusted for the driveline efficiency, accounting for efficiency losses in the wheel
bearings, differential, driveshaft, and transmission. The efficiency values were determined through
literature searches. Driveline efficiency ijdriveime varies with engine speed, vehicle speed, and
vehicle power requirements. Using sources available in the literature, 25-26-27-28-29-30-31-32-33 we
estimated an average value for driveline efficiency. Table 2-5 summarizes our findings.
20

-------
Table 2-5 Driveline Efficiencies Found in the Literature
Vehicle Type
Data Source
Driveline Efficiency
General Truck
Barth (2005)
80-85%
General Truck
Ludic (2001)
75-95%
HDT
Rakha
75-95%
NREL (1998)
91%
Goodyear Tire Company
86%
Ramsay (2003)
91%
21st Century Truck (2000)
94%
HDT Single Drive/direct
SAE J2188 Revised Oct.
2003
94%
HDT Single Drive/indirect
92%
HDT Single Drive/double
indirect
91%
HDT Tandem Drive/direct
93%
HDT Tandem
Drive/indirect
91%
HDT Tandem
Drive/double indirect
89%
Bus
Prtichard (2004):
Transmission Eff.
96%
Hedrick (2004)
96%
MIRA
80%
Based on this research, we used a driveline efficiency of 90 percent for all HD regulatory classes.
Equation 2-2 shows the translation from engine power Peng to axle power Paxie. MOVES uses the
Pioss.acc from Table 2-4 for each regulatory class and engine power-level (high, medium, or low).
Paxle ~ Vdriveline (.Peng ~ Ploss.acc)	Equation 2-2
Finally, we scaled the axle power using Equation 2-3, and the STP-scaling factors fscale presented in
Table 1-3 for every second of data.
STP = —Equation 2-3
fscale
We then constructed operating mode bins defined by STP and vehicle speed according to the
methodology outlined earlier in Section 1.5. It is possible that future test programs might acquire
accessory load information from the ECU and axle efficiency data available through certification
information during the HD GHG Phase 2 compliance program.
2.1.1.4 1960-2009 Model Years
The 1960-2009 model year emission rates were originally developed for MOVES2010 using the
ROVER and Consent Decree Testing data and then evaluated using MY 2003-2009 HDIUT testing
21

-------
data and the Houstan Drayage Program data.34'35 In MOVES2014, the emission rates were
considered for an update using the HDIUT or Houston Drayage Program data if:
1.	MOVES2010 rates were not based on actual data, and
2.	The comparison to the new independent data shows that more than a half of MOVES2010
emission rates are outside the boundary of the 95 percent confidence intervals of the
independent data.
Using this methodology, the emission rates for HHD MY 2003-2006 vehicles and LHD2b3 and
LHD45 MY 2003-2006 vehicles were updated using the MY 2003-2006 HDIUT data.a The
Houston Drayage Program was not used to estimate any of the emission rates in MOVES.
Table 2-16 outlines the data sets used to estimate emission rates for each model year group and
regulatory class. ROVER and Consent Decree testing data was used to estimate the following
regulatory class and model year groups combinations: HHD 1991-1997, 1998, 1999-2002, 2003-
2006, MHD 1999-2002, 2003-2006b, and Urban Bus 1991-1997, 1999-2002, 2003-2006. The
HDIUT was used to estimate emissions for HHD 2007-2009 and LHD 2003-2006.
Emissions in each data set were reported in grams per second. To calculate MOVES heavy-duty
exhust emission rates, we first averaged all the 1-Hz NOx emissions by vehicle and operating mode
because we did not believe the amount of driving done by each truck was necessarily
representative. Then, the emission rates were again averaged by regulatory class and model year
group. For trucks MY 2009 and older, these data sets were assumed to be representative and each
vehicle received the same weighting. Equation 2-4 summarizes how we calculated the mean
emission rate for each stratification group (i.e., model year group, regulatory class, and operating
mode bin).
Where:
tij = the number of 1-Hz data points (for a given operating mode bin) for each vehicle j,
«veh = the total number of vehicles,
rPjj = the emission rate of pollutant p for vehicle j at second
- = the mean emission rate (meanBaseRate) for pollutant p (for a given model year group, regulatory class
P
and operating mode bin).
We calculated a mean emission rate, denoted as the "meanBaseRate" in the MOVES
emissionRateByAge table, for each combination of regulatory class, model year group, and
a This analysis is described in more detail in the MOVES2014 heavy-duty exhaust report However, since we have
updated the fah values used to assign STP bins, the MOVES3 rates are no longer directly comparable (see Presentation
by Choi et al. (201234).
b For 2003-2006 MHD, the emission rates are different than the 2003-2006 HHD emissions for operating modes 0, 1,
11,21, and 33. For the other operating modes, the emission rates are equivalent to the HHD emission rates.
^•veh
Equation 2-4
22

-------
operating mode bin combination. 95% confidence intervals of the mean emission rate were
calculated by accounting for the variability of the averages across different trucks in a regulatory
class and model year group. Examples of mean emission rates are displayed in Figure 2-1. As
expected, the emissions increase with power, with the lowest emissions occurring in the
idling/coasting/braking bins.
The data included in the emissions analysis does not cover all operating modes or regulatory
classes and model year combinations needed for MOVES. In the following sections, we discuss the
methods used to fill missing operating mode bins, and missing regulatory class and model year
combinations. In addition, we also estimate the impact of low- NOx rebuilds which were not
included in the sampled vehicles. To do so, we rely on the heavy-duty diesel emission standards, as
well as engineering knowledge and test data of emission control technologies that were
implemented or forecasted to be implemented to meet the standards.
2.1.1.4.1	Light Heavy-Duty Class 2b3 and Classs 4&5 Trucks
As described in Section 1.5, the LHD regulatory classes were redefined for MOVES3, and the fscaie
value for 2009-and-earlier LHD2b3 and LHD45 regulatory classes is now 2.06 metric tons.c This
differs from MOVES2014 where the value for these vehicles was 17.1 metric tons. Thus, it was
not possible to carry over the emission rates from MOVES2014. Instead, in MOVES3 and later
versions, the 2009-and-earlier LHD2b3 and LHD45 regulatory classes were assigned the same
emission rates as the LHEK10K in MOVES2014.
2.1.1.4.2	High-Pow er Operating Modes
As described in Section 1.5, fscaie values for MHD and HHD trucks were not changed for model
years 2009-and-earlier. Thus, for MHD and HHD trucks, the maximum operating mode
(opModelD = 40) represents a tractive power greater than 513 kW (STP= 30 skW x 17.1). This
value exceeds the capacity of most HHD vehicles; MHD vehicles and buses exert even less power.
Nearly all of the HHD activity occurs in modes 0, 1, 11-16, 21-28, and 33-38, with activity for
buses and MHD vehicles usually occurring over an even smaller range.
To estimate emission rates in the modes beyond the ranges of available data, for each model year
group we linearly extrapolated the rates from the highest operating mode in each speed range where
significant data were collected. In most cases, this mode was mode 16 for the lowest speed range,
mode 27 or 28 for the middle speed range, and mode 37 or 38 for the highest speed range. For each
of these operating modes, work-specific emissions factors (g/kW-hr) were calculated using the
midpoint STP (Table 1-4). Then, these emissions factors were multiplied by the midpoint STP of
the higher operating modes (e.g., modes 39 and 40 for speed>50 mph) to input emission rates for
the modes lacking data. For the highest bins in each speed range, a "midpoint" STP of 33 skW
(564.3 kW) was used. Equation 2-5 displays an example calculation of the emission rate for
opModelD 40, using a mean emission rate from opModelD 37, for a given regulatory class and
model year group.
0 This is consistent with the 2.06 f,caie used to develop LHD2b3 and LHD45 emission rates in MOVES2010.
23

-------
Emission Rate,
opModelD 40
= Emission Rate
STP,
opModelD 37 ^ I ct^d
\ o 1 r,
opModelD 40
Equation 2-5
opModelD 37,
Figure 2-1 shows NOx emission rates by operating mode for HHD trucks in MY 2002. The mean
emission rates for the highest STP operating mode bins (30 and 40) are extrapolated using the
method explained above. In addition, the confidence intervals for the extraplolated bins are copied
from the closest bin with collected data. However, because the data are extrapolated, the
uncertainty of these rates is larger than what is shown by the copied confidence intervals.
Figure 2-2 displays the MY 2002 MHD and Urban Bus regulatory classes emission rates, with the
error bars removed for clarity. For these vehicle, less data was captured at the highest operating
mode bins, and more of the high speed, high power emission rates were extrapolated, which
explains the strictly linear trend in the emission rates between operating mode 27 -30 and 37-40.
4000
3500
_3000 -
-C
tkO
o) 2500 -
+¦»
(0
k.
X
92000
«
01
s 1500
1000
500 -|	*
z
* 5 *
	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
Operating Mode
Figure 2-1 NOx Emissions by Operating Mode from HHD Trucks for Model Year 2002. Error Bars represent
the 95 percent confidence interval of the Mean
24

-------
6000
5000
4000
ttfl
Qi
(V
g 3000

-------
years, using the ratio of emission standards between these two model-year groups (5.0 g/bhp-hr
over 4.0 g/bhp-hr) such that the 1991-1997 rates are 1.25 times the rates for MY 1998.
Table 2-16 provides a summary of the assumptions used to estimate emission rates for regulatory
class-model year groups with missing data.
The default emission rates for HHD and MHD diesel vehicles in MOVES for model years 1991
through 1998 are intended to include the effects of defeat devices as well as the benefits of heavy-
duty low-NOx rebuilds (commonly called reflash) that occurred as the result of the heavy-duty
diesel consent decree.14 Reflashes reduce NOx emissions from these engines by reconfiguring
certain engine calibrations, such as fuel injection timing.
Since defeat devices were in effect mostly during highway or steady cruising operation, we
assumed that NOx emissions were elevated for only the top two speed ranges in the running exhaust
operating modes (>25mph). To modify the relevant emission rates to represent reflash programs,
we first used emission rates from model year 1999 (the first model year with not-to-exceed
emission limits, see Table 1-2) to calculate baseline ratios of the emission rates for operating modes
27 and 37 to the rate for opMode 16 (as discussed in the beginning of Section 2.1.1.4). We then
multiplied the MY 1999 ratios by the emission rates in operating mode 16 for model years 1991
through 1998, to get estimated "reflashed" emission rates for operating modes 27 and 37. This step
is described in Equation 2-7 and Equation 2-8. To estimate "reflashed" rates in the remaining
operating modes, we multiplied the reflashed rates by ratios of the remaining operating modes to
mode 27 for MY 1991-1998, as shown in Equation 2-9 and Equation 2-10. The second step
preserves the relationship among operating mode 21 and operating modes 21-30, and among
operating mode 37 and operating modes 31-40, that existed in the original 1991-1998 data.
2.1.1.4.4 Defeat Device and Low-NOx Rebuilds for 1991-1998 Model Year
HHD and MHD
Where:
operating
modes (OM)
21-30
Equation 2-7
Equation 2-8
Where:
operating
modes (OM)
31-40
Equation 2-10
Equation 2-9
26

-------
Because the reflash occurred over time after the engines were sold, we phase-in the reflash rates
with age. An EPA assessment shows that about 20 percent of all vehicles eligible for reflash had
been reflashed by the end of 2008.37 We assumed that vehicles were reflashed at a steady rate from
the time of the consent decree (1999/2000 calendar year), such that in 2008, about 20 percent had
been reflashed. Figure 2-3 displays the results of the reflash calculations on the HHD fleet-average
emission rate for operating mode 37. For model years 1994-1997 and 1998, we approximated a
linear increase in reflash rate from age zero to the age 20+ age group. For model years 1991-1993,
the adjustments do not start until calendar year 2001, when the 1991 model years exist in the age
10-15 group. When all of the 1991-1998 vehicles have reached the age 20+ group (CY 2018),
MOVES assumes that close to 30% of all the 1991-1998 engines have been reflashed. The
reflashed HHD 1991-1998 MY emission rates were also applied to the 1991-1998 MHD diesel
emission rates. Note that there are no tampering and mal-maintenance aging effects in MOVES for
pre-2010 NOx emission rates. Thus, the only change in the NOx emission rates by vehicle age for
these model years is due to reflashing.
2540
2520
2500
2480
M 2460
a
w 2440
g 2420
H 2400
2380
2360
2340





"0"C
•O-G.. \
-O "G.\






>~O~Q-©"0












—•—J
4odel Year 1991-1993



....... j
4odel Year 1994-1997



—•—I
4odel Year 1998


















0
10	15	20
Age (years)
25
30
Figure 2-3. HHD NOx emission rates for OpModelD 37 for Model Years 1991-1998 Adjusted for Low-NOx
Rebuilds by Vehicle Age
2.1.1.4.5
2007-2009 Model Year HHD, MHD, and Urban Bus
The 2007 Heavy-Duty Rule108 required the use of ultra-low sulfur diesel fuel; this fuel enabled
diesel engines with diesel particulate filters to reach the 0.01 g/bhp-hr PM standard beginning in
2007. In addition, the 2007 Heavy-Duty Rule108 established much tighter NOx emission standards
(0.2 g/bhp-hr). While the NOx standard went into effect for MY 2007 at 0.2 g/bhp-hr, it was phased
in over a three-year period ending in MY 2010. Rather than phasing in the aftertreatment
technology needed to meet the new standard, most manufacturers chose to meet a 1.2 g/bhp-hr
27

-------
standard for MY 2007-2009 (down from 2.4 g/bhp-hr in 2006), which did not require NOx
aftertreatment. For the 2007-2009 HHD, we used the data from the HDIUT program. For the NOx
emission rates within the 2007-2009 model year group for MHD and Urban Bus, we assumed that
the NOx emission rates were 50 percent lower than the corresponding Rover and Consent Decree
derived 2003-2006 emissions (proportional to the reduction in the NOx emission standards
mentioned above). The MHD MY 2007-2009 rates are consistent with NOx emission rates
measured from 2007-2009 MHD trucks measured in HDIUT.35
2.1.1.4.6 2007-2009 Model Year LHD45 andLHD2b3
For LHD2b3 trucks in 2007-2009, we accounted for the penetration of Lean NOx Trap technology.
Cummins began using Lean NOx Trap (LNT) aftertreatment starting in 2007 in engines designed to
meet the 2010 standard in vehicles such as the Dodge Ram. This technology allows for the storage
of NOx during fuel-lean operation and conversion of stored NOx into N2 and H2O during brief
periods of fuel-rich operation. In addition, to meet particulate standards in MY 2007 and later,
heavy-duty vehicles are equipped with diesel particulate filters (DPF). The DPF must be
regenerated at regular intervals to remove and combust accumulated PM to relieve backpressure
and ensure proper engine operation. This step requires high exhaust temperatures. However, these
conditions adversely affect the LNT's NOx storage ability, resulting in elevated NOx emissions.
In order to determine the fraction of time that DPFs spend in PM regeneration mode, in 2007, EPA
acquired a truck equipped with a LNT and a DPF and performed local onroad measurements using
portable instrumentation and chassis dynamometer tests. We distinguished regimes of PM
regeneration from normal operation based on operating characteristics, such as exhaust
temperature, air-fuel ratio, and ECU signals. During the testing conducted onroad with onboard
emission measurement and on the chassis dynamometer, we observed a PM regeneration frequency
of approximately 10 percent of the operating time.
Emissions from this vehicle were not directly used to calculate MOVES emission rates, because
only one vehicle was tested. Rather, to estimate average emissions of LHD2b3 vehicles with LNT,
we calculated a ratio to the MOVES2010d MY 2003-2006 NOx emission rates. During DPF
regeneration, we assumed that the LNT did not reduce emissions from 2003-2006 levels. During all
other times, we assumed that emissions were reduced by the percent reduction in the certification
standards from 2003-2006 levels (i.e., 90 percent). These assumptions result in an estimated NOx
reduction of 81 percent for LNT equipped trucks from 2003-2006 to 2007-2009, as shown in
Equation 2-11.
d In MOVES2014, we updated the diesel NOx emission rates for 2003-2006 based on the HDIUT program.
28

-------
LNT LHD2b3 NOx emissions
Baseline Emissions
(LNT normal emissions \
= (normal op. frequency) x 			
\ Baseline Emissions 1
(Baseline Emissions\
+ (DPF reg. frequency) x 			
\Baseline Emissions1
= (0.90) x (0.10) + (0.10) x (1) = 0.19
Where Baseline Emissions = MOVES2010 MY 2003-2006 NOx emission rates for LHD2b3
We weighted the average rates for the LHD2b3 regulatory class (regClassID 41) for model years
2007-2009 assuming that LNT-equipped trucks account for 25 percent of the LHD2b3 diesel
vehicles. We assume that the remaining 75 percent of MY 2007-2009 LHD2b3 diesel trucks will
not have LNT aftertreatment and will exhibit a 50% NOx reduction from the 2003-2006 model year
emission rates as was assumed for the HHD and MHD described in the previous section. Overall,
these assumptions result in a 58 percent reduction in NOx emission rates in the model year 2007-
2009 emission rates from the MOVES2010 MY 2003-2006 NOx emission rates as shown in
Equation 2-12.
2007 — 2009 LHD2b3 NOx emissions
Baseline emissions
(LNT NOx emissions\
= (LNT market share) —				—;—;	
\Baseline Emissions J
+ (non
(2007 — 2009 emission standards
— LNT market share) 			
\ Baseline Emissions
= (0.25) x (0.19) + (0.75) x (0.5) = 0.4225
Where Baseline Emissions = MOVES2010 MY 2003-2006 NOx emission rates for LHD2b3
In the absence of other data, we applied the LHD2b3 emission rates to LHD45 vehicles. Newer
data shows that LNT is not being used in LHD45 vehicles, however, we have not updated this
assumption because the LHD results compare well to the HDIUT data.34
2.1.1.5 2010-2026Model Years
In MOVES3 and later versions, the MY 2010-2026 emission rates for HHD, MHD, Urban Bus, and
LHD45, and LHD2b3 are based on analysis of the HDIUT data and model-year specific production
volume weighting for each model year from 2010 through 2018. The rates for HHD, MHD, and the
two LHD classes use data from vehicles with HHD, MHD, and LHD engines, respectively. The
NOx emission rates are projected to remain constant for MY 2018 and later vehicles for regulatory
classes HHD, MHD, Urban Buses, and LHD45. The LHD2b3 trucks are projected to have a further
decrease in NOx emissions through the implementation of the Tier 3 program as discussed in
Section 2.1.1.5.5.
Equation
2-11
Equation
2-12
29

-------
In calculating the mean emission rates for MY 2010+ vehicles, we made several additions to the
method presented for the pre-2010 model years described in Section 2.1.1.4:
1.	For a given regulatory class (HHD, MHD, LHD), we grouped the vehicles in the HDIUT
data set into three NOx family emission limit (FEL)e groups as detailed in Section 2.1.1.5.1
below.
2.	Within the NOx FEL group and regulatory classes, we grouped vehicles into model year
groups as detailed in Section 2.1.1.5.2.
3.	We calculated the operating mode-based average emission rate for each vehicle by
regulatory class, NOx FEL group, and model year group (Equation 2-13). Then, we
calculated the operating mode-based average emission rate for all vehicles in the NOx FEL
group and model year group (Equation 2-14).
4.	We weighted the operating mode-based average emission rates for each of the NOx FEL
groups, and model year group within each regulatory class by the model year specific
production volumes of the engines in the NOx FEL group to the total production volume of
the regulatory class (Equation 2-15). Thus, we created operating mode-based average
emission rates for each model year and regulatory class.
ER
Usee ERpol.OM.C,FEL,
MYG.veh.sec
pol,OM,C,FEL,MYG,veh
sec,
count
Equation 2-13
ER
Hveh ERpol.OM.C.FEL,
MYG.veh
pol,OM,C,FEL,MYG
veh,
count
Equation 2-14
ER
pol,OM,C,MYG,MY
ERpol,OM,C,FEL,MYG *
PV,
C, MY, FEL
FEL
jFEL
PV,
Equation 2-15
C,MY,FEL,
Where:
C
ERx,y,z
FEL
MYG
MY
OM
pol
P V C,MY,FEL
sec, seCcount
veh, vehcount
= Regulatory class (LHD, MHD, HHD, and Urban Bus)
= Emission rate in mass/time. The subscripts show the categorization.
= NOx FEL group of engine (0.20 g/bhp-hr, 0.35 g/bhp-hr, and 0.50 g/bhp-hr)
= Model year group (2010-2013, 2014-2016)
= Model year
= Running exhaust emissions operating mode
= Pollutant (NOx, THC, CO)
= Production volume by class, model year, and FEL group
= a second of data (for a given veh and OM); number of seconds in that category
= a vehicle (in the class and FEL grouping); number of vehicles in that category
Figure 2-4 displays the average NOx emission rates using the described method for HHD model
year 2013 vehicles. To calculate the 95% confidence intervals for the model years, the confidence
intervals were first calculated for each FEL group, treating the mean emission rate by operating
e A Family Emission Limit is the maximum emission level established by a manufacturer for the certification of an
engine family.
30

-------
mode from each individual truck as an independent random variable. Then, the confidence interals
from the different FEL groups were weighted together using the production volumes, similar to
how the means were calculated.
o.io -
0.08 -
< 0.06 -
3B
x
O
Z 0.04 -
0.02 -
0.00 -
Figure 2-4
More details about the selection of the NOx FEL groupings, updated /scale values, and the methods
for the production volume weighting are provided in Section 2.1.1.5.1 through 2.1.1.5.4 below.
2.1.1.5.1 NOx Family Emission Level Groups
We grouped engines, within a regulatory class, by their NOx FEL. These groups are shown in Table
2-6.
Table 2-6 NOx Family Emission Limit (FEL) based Groups for LHD, MHD, HHD, and Urban Bus Classes in the
HDIUT Data
Group Name
Range of NOx
FEL (g/bhp-hr)
Lower Limit
(Excluded)
Upper Limit
(Included)
0.20
0.00
0.20
0.35
0.20
0.35
0.50
0.35
0.50
Each test vehicle, within a regulatory class, was assigned to one of these three groups. These
groupings were applied not only to NOx, but to all pollutants for emission rate calculations. We
chose to use NOx as the basis for creation of these groups because data for NOx FEL is most
abundant in the heavy-duty engine certification database and, for MY 2010-2015 engines, the
biggest technology changes and tailpipe exhaust emissions impacts are due to emissions control
measures for NOx. We arrived at the specific group bins based on the spread of NOx FELs for MY
2010-2015 engine families reported in the certification database available at the time of the analysis
(not just the engine families tested under the HDIUT program). The NOx FELs for MY 2010-2015
HD diesel engine families in the certification database are shown in Figure 2-5. As highlighted by

¦ HDIUT, MY 2013
—


—











i


















i
il
I




T



i i
¦





11







il



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
NOx Emissions by Operating Mode from HHD Trucks for Model Year 2013. Error Bars represent
the 95 percent confidence interval of the Mean
31

-------
the shaded rectangles, most of the engine families are concentrated in the 0.05-0.20, 0.30-0.35,
and 0.45-0.50 bands and this trend guided our bin choices, represented by the curly braces, for the
three NOx FEL groups.
100
O-^
10
0)
E
05
LL.
V
c
lo
O
rsi
<2P S
£ s
05 >
»4-
O
so
80
60
40
20

	 . — w
-0 0-^ ,	
lO-TP'	" '



\
1
P'



/ f




[V





/
1 J





0.1 0.2 0.3 0.4
NOx FEL(g/bhp-hr)
0.5
0.6
Figure 2-5 Distribution of NOx Family Emission Limit (FEL) for Model Year 2010 -2015 Heavy-Duty Diesel
Engine Families, as Reported in the Certification Database
Table 2-7 shows the number of vehicles by regulatory class and NOx FEL group for MY2010+
engines in the HDIUT program. The number of vehicles by regulatory class in this table match the
number of vehicles in Table 2-2. The 10 Urban Bus vehicles in the HDIUT data set were not used
in MOVES because they only represented one engine family. The Urban Bus emission rates are set
equal to the HHD emission rates because: Urban Buses are in the same GVWR class as HHD;
some engines certified as HHD are used in the Urban Bus application; and there is no separate NOx
standard (for MY 2010+) for the Urban Bus regulatory class.
32

-------
Table 2-7 Number of 2010 and Later Model Year HDIUT Vehicles by NOx FEL Group and Model Year
Groups1 Used for Emission Rate Analysis

\() l i:
. (¦ ionp

0.2
0.35
0.5

MY
MY
MY
MY


2010-
2014-
2010-
2010-


2013
2016
2016
2016
Total
LHD2
52
27
0
15
94
MHD
19
23
23
9
74
HHD3
78
50
31
35
194
Total
149
100
54
59
362
NOTE:
1	THC, CO, and CO2 emission rates were analyzed using the same model year groups. Sample size is generally the same, with a few
exceptions for vehicles with invalid pollutant measurements for one or more pollutant. For example, one of the MHD 0.35 NOx FEL
vehicles did not have NOx measurements.
2	LF1D data for the 0.2 FEL group includes MY up to MY 2016. MHD and HHD data for the 0.2 FEL group only include up to MY
2015
3	The HHD 0.35 group contains five MY 2009 vehicles, which have similar NOx emission rates as the other vehicles in this group.
The average NOx emission rates for the 0.2 NOx FEL Group for model year 2010-2013, compared
to the 0.35 and 0.5 NOx FEL Groups are shown below for LHD (Figure 2-6), MHD (Figure 2-7)
and HHD (Figure 2-8). Because the 0.35 and 0.5 NOx FEL groups do not have a model year
distinction like the 0.2 NOx FEL group, the figures display all the available MY 2010 and later
vehicles for the 0.35 and 0.5 NOx FEL groups.
As shown, the NOx emission rates in the 0.2 NOx FEL Group and model year 2010-2013 are
consistently lower than those in the 0.35 and 0.50 NOx FEL groups, with a few minor exceptions.
In addition, for many operating mode and regulatory class combinations, the NOx emission rates in
the 0.2 NOx FEL Group are statistically significantly lower than the NOx emission rates in the 0.35
and 0.5 NOx FEL Groups. However, no such trend exists in the NOx emission rates for the 0.35 and
the 0.5 NOx FEL groups. In general, the NOx emission rates in the 0.35 and the 0.5 NOx FEL
groups are not statistically different than one another by operating mode for the MHD and HHD
regulatory classes.
33

-------
0.10
0.08 -
<. 0.06 -
•SB
x
O
Z 0.04 H
0.02 -
0.00
I LHD FEL 0.20, MY 2010-2013- N=52
! LHD FEL 0.35, N=0
l LHD FEL 0.50, MY 2010-2015; N=15
±1
iiiili
iii
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 Average LHD NO* Emission Rates by Operating Mode for the 0.2 NO* FEL for MY 2010-2013 and
the 0.5 NOx FEL for MY 2010-2015. Error Bars are 95% Confidence Intervals of the Mean
0.10
0.08
< 0.06 -
.25
x
O
Z 0.04
0.02
0.00
ta MHD FEL 0.20, MY 2010-2013!, N=23
¦ MHD FEL 0.35, MY 2010-2013v N=23
B MHD FEL 0.50, MY 2010-2013- N=9
.1.1

•II
Ji
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 2-7 Average MHD NOx Emission Rates by Operating Mode for the 0.2,0.35 and 0.50 NOx FEL Groups
for MY 2010-2013 Vehicles. Error Bars are 95% Confidence Intervals of the Mean
34

-------
0.20
0.15
in
Quo
~ 0.10
o
z
0.05
0.00
Figure 2-8 Average HHD NOx Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013 and
the 0.35 and 0.5 NOx FEL for MY 2010-2015. Error Bars are 95% Confidence Intervals of the Mean
2.1.1.5.2 Model Year Groups within the 0.2 NOx FEL Group
We grouped the vehicles within the 0.2 NOx FEL Group further into 2010-2013 and 2014 and later
model year groups to account for differences in emissions performance of more recent engines and
aftertreatment systems. In a subsequent update for MOVES3 based on studies49 50 51 that suggested
SCR control had improved, we were able to incorporate data from MY 2014-2016 engines to
evaluate this recommendation.
Table 2-7 displays the model year groupings used in developing MOVES emission rates by NOx
FEL groups and pollutant. For NOx, THC, CO, and CO2, we grouped 0.2 NOx FEL group into two
model year groups given the sufficient sample size. Within the 0.35 and 0.5 NOx FEL groups, we
only have a single model year group (2010 and later) due to the smaller sample size of HDIUT test
vehicles in these groups. For example, for the MHD regulatory class, the HDIUT data set only
includes MY 2010-2013 vehicles in the 0.35 and 0.5 NOx FEL Group (see Table 2-29). Aside from
the small sample size, we believe it is defensible to have single model year groups for the 0.35 and
0.5 NOx FEL groups for two additional reasons. First, with the exception of the 0.35 NOx FEL
Group for MHD, heavy-duty vehicle manufacturers are no longer producing engines certified in the
0.35 or 0.5 NOx FEL groups beyond MY 2017. Second, we anticipate the MY 2014-2017 MHD
engines certified to the 0.35 NOx FEL group would have similar emission control technology as
earlier MY 2010-2013 vehicles certified to the same emission levels. For these reasons, we believe
it is reasonable to use only one model year group for the 0.35 and 0.5 NOx FEL groups, within each
regulatory class.
A comparison of the emission rates for the 0.2 NOx FEL Group by model year group and
regulatory class are shown in Figure 2-9 through Figure 2-11. For the LHD vehicles, the NOx
emission rates for MY 2014 and later are lower than the MY 2010-2013 vehicles, with some of the
differences being statistically significant. For the MHD and HHD regulatory classes, no significant
I HHD FEL 0.20, MY 2010-2013
1 HHD FEL 0.35, MY 2010-2015
I HHD FEL 0.50, MY 2010-2015
	
L4
N=78
N=31
N=35
A
11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES OpMode
35

-------
differences in the NOx emission rates between the two model year groups are observed across
operating modes. In contrast, significant differences are observed for THC and CO emissions for
the MHD and HHD regulatory classes between the MY 2010-2013 and MY 2014-2015 model year
groups as discussed in Section 2.1.3.2. Even though the differences in NOx emission rates across
regulatory classes and the model year groups were not statistically significant, we separated the 0.2
NOx FEL group for all three regulatory classes into two model year groups (2010-2013 and 2014
and later) to be consistent in our analysis for all pollutants, and to account for the potential
differences in real-world NOx emissions performance due to the updated engine and aftertreatment
systems
0.030
0.025
	0.020
-59
x 0.015
O
¦z.
0.010
0.005
0.000
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 2-9 NOx emission rates for the MY 2010-2013 and MY 2014-2016 vehicles in the LHD 0.20 NOx FEL
Group
MY 2010-2013
MY 2014-2016

				:	
ii
ji
0.030
0.025
l/>
M 0.020
CD
rtJ
^ 0.015
c
o
j/i
0.010
E
LU
0.005
0.000
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 2-10 NOx emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the MHD 0.20 NOx FEL
Group
MY 2010-2013
MY 2014-2015
»il ii
ii
I
36

-------
0.08
0.07
~ 0.06
0.05
x
O
2 0.04
0.03
0.02
0.01
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
MOVES OpMode
Figure 2-11 NO NOx emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the HHD 0.20 NOx
FEL Group
2.1.1.5.3 Weighting by Production Volume
We collected production volume data by the same regulatory classes (LHD, MHD, HHD, Urban
Bus) and NOx FEL groups (0.20, 0.35, 0.50) that we used for the emission rate analysis. We then
combined the NOx FEL group-based rates (averaged within the model year groups described
above) with the production volume data (distinct for each model year) to create emission rates
unique to each model year. We did this for each model year from 2010 through 2018 (the last
model year for which we have production volume data). For MY 2019 and later, we used the same
production volume weighting as MY 2018. The per-model-year production volume weighting, by
regulatory class and NOx FEL groups, is shown in Figure 2-12. This method allows us to better
represent the technology adoption landscape in the heavy-duty domain. For example, for HHD,
model years 2010 through 2013 had engines with NOx FEL in the 0.50 group (0.35 g/bhp-hr < NOx
FEL <0.50 g/bhp-hr), but starting with model year 2014, there are no engines in the 0.50 group.
Compared to engines in the 0.20 group and 0.35 group, engines in the 0.50 group predominantly
use a different emissions control strategy to reduce tailpipe NOx emissions. Thus, our approach
using the NOx FEL groups and per-model-year production volume correctly captures the
prevalence and influence (on emissions) of different technologies in the fleet. Production volume
percent contributions of the three NOx FEL groups sum to 100 percent of the production volume
for each regulatory class and model year.

¦ MY 2010-2013






¦ MY 2014-2015






























tt
T














I









it




ii





i. Ii ii



ii i1





ii








37

-------
Figure 2-12 Production Volume Contribution of Heavy-Duty Diesel Engine Families by NOx Family Emission
Limit Group for Each Regulatory Class
2.1.1.5.4 Operating Modes andfscale Values
For the updates to THC, CO, NOx, PM2.5, and energy rates for MY 2010-2026 HD vehicles, we
used new fscale values (see Table 1-3 and Appendix G) that allowed all OpModes to be populated
with rates based on real-world data. Thus, there was no need to perform the hole-filling approach
used to populate the high-power operating modes for the pre-2010 MY vehicles (see 2.1.1.4.2).
Figure 2-13 to Figure 2-15 show the effect of the new fscale values on OpMode coverage using the
example of NOx emission rates for the vehicles in the NOx FEL=0.20 group for LHD, MHD, and
HHD regulatory classes, respectively/Note that the absolute mass/time OpMode-based emissions
rates between the two series based on different fscale cannot be compared. The main benefit of the
new fscale values is that all the operating modes are populated even if the trends may not be
perfectly monotonically increasing for each pollutant in a regulatory class. The comparison is
similar for the 0.35 and 0.50 NOx FEL groups. Note the final rates input into MOVES are estimated
as production volume weighted rates from each of the NOx FEL groups.
f These plots were generated using model year 2010-2015 HDIUT data that was available at the time of the analysis.
The final emission rates were developed using the complete HDIUT data that includes both MY 2010-2015 and MY
2016-2018.
38

-------
0.025
0.020 H
^ 0.015
cm
X
o
Z 0.010 -
0.005 -
0.000
¦ L H D DO. 20_f s=5.00
HLHDD 0.20 fs=17.1
i=
IlJ
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 2-13 Effect of MOVES2014 (17.1) and MOVES3 (5.00)/^ Values on OpMode Coverage for NOx
Emission Rates for Light Heavy-Duty Vehicles in the NOx FEL = 0.20 Group
0.060
0.050
0.040
"w"
M
0.030 -|
o
z
0.020 -
0.010 -
0.000
¦ MHDD_0.20_fs=7.00
BMHDD 0.20 fs=17.1
E IB I
1 11 12 13 14 15 16

ill
21 22 23 24 25 27 28 29 30
MOVES OpMode


11
1
1
. 1
33 35 37 38 39 40
Figure 2-14 Effect of MOVES2014 (17.1) and MOVES3 (7.00) fscaie Value on OpMode Coverage for NOx
Emission Rates for Medium Heavy-Duty Vehicles in the NOx FEL = 0.20 Group
39

-------
M
x
O
0.080
0.070
0.060
0.050
0.040
0.030
0.020
0.010
0.000
¦ H H D DO. 20_f s=10.00
BHHDD 0.20 fs=17.1
MS.
II
0|1 11 12 13 14 15 16 121 22 23 24 25 27 28 29 30 ¦ 33 35 37 38 39 40
MOVES OpMode
Figure 2-15 Effect of MOVES2014 (17.1) and MOVES3 (10.00)/^ Value on OpMode Coverage for NOx
Emission Rates for Heavy Heavy-Duty Vehicles in the NOx FEL = 0.20 Group
2.1.1.5.5
HD Incorporation of Tier 3 Standards for LHD2b3
In addition to regulating light-duty vehicles, the Tier 3 vehicle emission standard38 also applies to
chassis-certified light heavy-duty diesel vehicles, which with the new MOVES4 definition,
includes all of regclass 41 (LHD2b3). In addition, we updated the T&M assumptions in MOVES4
for regclass 41 by reflecting the 50,000 miles warranty requirement for LHD vehicles, and the
150,000 useful life mileage for Tier 3 LHD2b3 vehicles (see Section B.l). For these LHD2b3
diesel vehicles, we estimate reductions in the NOx zero-mile emission rates attributable to the Tier
3 standards beginning in 2018. We did not estimate Tier 3 reductions in the THC, CO, and PM2.5
emission rates as discussed in subsequent sections.
For diesel vehicles in the LHD2b3 regulatory class, we estimate that the Tier 3 NOx standard
results in a different percent reduction of start and running emissions. We applied a greater portion
of the reduction to running emissions and a smaller reduction to start emissions. These reductions
were phased-in over the same schedule as for gasoline vehicles, as detailed in Table 2-8. The
derivation of the phase-in assumptions is discussed in the MOVES2014 heavy-duty exhaust
report.35
40

-------
Table 2-8 Phase-in Assumptions for Tier 3
NOx Standards for Light Heavy-Duty 2b3 Diesel Vehicles
Model Year
Phase-in
fraction (%)
Reduction in Running
Emission Rate (%)'
Reduction in Start
Emission Rate (%):
2017
0
0.0
0.0
2018
49
30.1
11.2
2019
62
38.1
14.2
2020
75
46.1
17.2
2021
87
53.5
19.9
2022
100
61.5
22.9
Note:
1 These reductions are based on comparison of Tier 3 standards against Tier 2 standards.
In generating the reduced rates for running operation, the starting point (or pre-Tier 3 baseline) are
the LHD2b3 rates for MY2017. The ending point, representing full Tier 3 control, was model year
2022. The MY 2022 rates were calculated by multiplying the rates for MY2017 by 0.3855. This
fraction reflects the percent reduction in running emission rates for MY2022 as shown in Table 2-8.
In addition to tightening emission standards, the Tier 3 regulations require an increase in the
regulatory useful life from 120,000 miles to 150,000 miles. We used the Tampering and Mal-
maintenance (T&M) methodology presented in Appendix B to estimate the emissions impact of the
lengthened useful life, and to estimate different age effects for Tier 3 LHD2b3 vehicles as shown in
Table B-3. For the phase-in model years, we calculated a weighted average of the Model Year 2017
and Model Year 2022 emission rates using the Tier 3 phase-in, as shown in Equation 2-16.
LHD2b3 ERmy = (LHD2b3 ER20i7 ) x (1 — Tier 3 PhaseinMy)
+ (LHD2b3 ER2022) x (Tier 3 PhaseinMy)	Equation 2-16
Where:
LHD2b3 ERmy = the LHD2b3 diesel emission rate for each process (start/running), age, operating mode, and
model year between 2018 and 2021.
Equation 2-16 is also used to estimate the impact of the lengthened Tier 3 useful life standard on
NOx, PM2.5, THC, and CO emissions.
Tier 3 also allows fleet-wide averaging with electric vehicles as explained in the MOVES
adjustments report.63 While this averaging is allowed for LHD2b3s, it is not modelled in MOVES.
The reduction in NOx grams per mile from the Tier 3 rulemaking across model years is displayed in
Figure 2-25.
2.1.1.6 202 7-2060 Model Years
In MOVES4, we updated the MY 2027 and later emission rates for the LHD45, MHD, HHD and
Urban Bus regulatory classes to account for the impact from the HD2027 standards. The HD2027
standards include duty-cycle standards, off-cycle standards and changes to warranty and useful life
41

-------
requirements. This section describes the methodology to derive the new zero-mile NOx emission
rates for running emissions. Temperature adjustments for running and extended idle NOx
emissions are described in the MOVES temperature report.63
We first estimated the effects of the duty-cycle standards and the off-cycle standards separately, as
discussed in Sections 2.1.1.6.1 and 2.1.1.6.2, respectively. Then, we estimated the combined effect
of both the duty-cycle standards and off-cycle standards on the zero-mile emission rates used in
MOVES as discussed in 2.1.1.6.3. The effects of aging are described in Section 2.1.1.7.
2.1.1.6.1
Emission Rates Based on Duty-Cycle Standards
The HD2027 NOx heavy-duty compression ignition duty-cycle exhaust emission standards for
MY2027+ are shown in Table 2-9. The duty-cycle standards include three separate tests: Federal
Test Procedure (FTP), Supplemental Emission Test - Ramped Modal Cycle (SET-RMC) and low
load cycle (LLC). We used the 2027+ 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 LLC standard in MOVES, we used the FTP standard to model the impact of the standards on
low-power operation.
Table 2-9 Heavy-duty Compression Ignition Duty-Cycle NOx Standards for the HD2027 Rule
Regulatory
Classes
FTP
(g/hp-hr)
SET-RMC
(g/hp-hr)
LLC (g/hp-
hr)
LHD
0.035
0.035
0.05
MHD
0.05
0.05
0.065
HHD
0.05
0.05
0.065
HD2027 FTP and SET RMC standard	r .. . 1>7
d 			Equation 2-17
duty	MY2010-2026 standard
Equation 2-17 through 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 standards. The term Rduty is the ratio between the new emission standards and
the MY 2010-2026 FTP and SET-RMC duty-cycle standards (0.2 g/hp-hr).
HD2027 FTP and SET RMC standard	r .. . 1>7
d —		Equation 2-17
duty	MY2010 - 2026 standard
Rduty ranges between 17 and 25 percent as shown in Table 2-10.
42

-------
Regulatory Classes
FTP and SET
standard
(g/hp-hr)
Rduty
LHD
0.035
17.5%
MHD, HHD
0.05
25%
To estimate the effect of the new engine dynamometer duty-cycle standards on in-use emissions,
we looked at this relationship for the previous change in NOx standards. The 2010 0.2 g/hp-hr NOx
emission standard39 is the most recent previous 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 pre-2010 heavy-duty NOx standard with the 2010 standard.
Equation 2-18 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 pre-2010
standard, and the percent change in the 2010 standard FTP standards relative to the pre-2010
standard.8 In other words, Rin uSe is the ratio between the relative change in in-use emissions
compared to the relative change in the FTP duty-cycle emission standard. This ratio is operating
mode specific and the ratio computed for HHD is applied to MHD and LHD45.
Rir
% Change in the in — use emission rates from 2010 compliant vehicles
% Change in the 2010 FTP standard
Equation 2-18
The percent change in in-use emission rates from vehicles certified to the 2010 standard (the
numerator in Equation 2-18) was estimated using Equation 2-19. For each operating mode, the
MOVES emission rates for HHD vehicles certified to the 2010 0.2 g/hp-hr standard (the
numerator) were calculated from 93 MY 2010-2015 HHD vehicles with a certified engine family
emission limit (FEL) below the 0.2 g/hp-hr NOx emissions level and tested as part of the Heavy -
Duty In-Use Testing program.21 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. Because data was sparse for LHD and MHD, the HHD data
ratios were also used for LHD and MHD.
g 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.
43

-------
% Change in the in_use emission rates from 2010 compliant vehicles
Emission rate from HHD 2010 compliant vehicles	Equation 2-19
HHD MY 2006 MOVES emission rate
The percent change from Equation 2-19 was calculated separately for each MOVES operating
mode to evaluate the effectiveness of the 2010 standard across different ranges of in-use operating
conditions, as shown in Figure 2-16 and Table 2-11. 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 2-20 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 2-18. The 2004-2006 NOx
emission standard was estimated as 1.68 g/hp-hr, assuming that NOx emissions consist of 70
percent of the combined NMHC (Non-Methane Hydrocarbons) plus NOx standard, consistent with
the assumption used in MOVES340. The percent change associated with the standard change is
plotted as a horizontal line in Figure 2-16 to compare to the in-use emission rate reductions.
0.2
Percent Change in the 2010 FTP standard = (7QO/o x 2 4) ~ 1 = ~QQ-1% Equation 2-20
QJ
DO
X
O
50
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 2-16 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
Table 2-11 displays the Rin uSe as calculated by Equation 2-18. For operating modes with Rin use
values greater than one, the observed in-use emissions are less than would be expected due to the
44

-------
change in FTP emission standard. Operating modes with Rin use values less than one see less
impact than implied by the change in the FTP emissions standards.
Table 2-11 Calculation of Rm use by MOVES Operation Mode

HHD MOVES
NOx Emission rate
from 2010 compliant
HHD vehicles (g/hr)a
Percent change in in-use

MOVES
OpMode
MY 2006 NOx
emission rates
(g/hr)
NOx emission rates from
2010 compliant vehicles
(%)
Rin use C
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 fscaie of 17.1 metric tons to be consistent with the fscaie
of the MY 2006 HHD emission rates in MOVES3 and later versions. Note that the fscaie for
model year 2010 and later is 10 for HHD, 7 for MHD, and 5 for LHD45 and LHD2b3.
b For operating modes lacking data, we used the same Rm use for the closest operating mode.
0 Rm use is calculated based HHD data, but also applied to LHD45 and MHD.
Equation 2-21 is used to estimate the percentage reduction to NOx running emissions from the
change in the duty-cycle standard for each operating mode.
Rduty_in_use — (l ~~ Rduty) x Rin_use	Equation 2-21
Where:
Rduty_inuse= the percent emission reductions in the in-use running NOx emissions estimated from changing the FTP
duty-cycle standard.
45

-------
Equation 2-22 is used to estimate the age 0-3 heavy-duty diesel NOx running emission rates from
the changes in the duty-cycle standards. The same calculations were applied to estimate rates for all
the heavy-duty diesel regulatory classes.11
ER
duty_in_use
= (1-R
duty_in_use
) x ER
MOVES_baseline
Equation 2-22
Where:
ER
duty jnjxse
= the MOVES running NOx emission rates for the control scenarios based on reduction in the duty-cycle
standard
R
= the percent emission reductions in running NOx emissions estimated from changing to FTP duty-cycle
duty Jnjxse
standard
ERM0VES_baseiine= the MOVES baseline running NOx emission rates for each regulatory class
The estimated HHD MOVES running NOx emission rates for the HD2027 standards, estimated
based solely on the duty-cycle standards in Table 2-9, are shown in Figure 2-17.
300
,250
1200
w 150
8
100
50
i Without HD2027
HD2027 Duty-Cycle Standard Compliant Rates for MY2027+
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 2-17 Duty-cycle-based running NOx emissions, ERduty_in_use, for HHD diesel for the HD2027 standards
h We assumed that the Rm lse values 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 1-4. Changing the fscaie values does
change the definition of vehicle operation in the other operating modes. However, the R,„ ,.s,. values are relatively
constant for the positive power operating modes within each speed range as observed in Table 2-11. In calculating the
results shown in Figure 2-16, we deemed it was not necessary to attempt to account for the fscaie differences when
applying the Rm use values.
46

-------
2.1.1.6.2 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 HD2027 off-cycle standards for heavy-duty diesel vehicles. Table 2-12 presents the
calculated off-cycle standards used to develop MOVES inputs. The off-cycle standards include
requirements for operating conditions in two bins: idle bin (less than 6 percent of maximum power)
and non-idle load bin. In developing inputs to MOVES, we assumed manufacturers will comply
with the optional idle standard in all off-cycle idle operation. We then developed the off-cycle
standards using the procedures as described below. These calculations are detailed here for HHD,
but we used the same approach for MHD and LHD45.
In modeling the HD2027 off-cycle standards in MOVES4, we incorporated the impact of the
compliance margin EPA used in setting the off-cycle standards. The in-use compliance margin of
approximately 40% of the standards, accounted for a number of in-use factors, including test
procedure variability, production and engine variability, fuel and diesel exhaust fluid (DEF)
quality, aftertreatment aging due to severe-service operation, and aftertreatment packaging to name
a few. When manufactures design engines, they must include similar margin to ensure each of their
engines meet the standards, which results in the average fleet emissions being below the standards.
Table 2-12 Calculated Off-Cycle NOx Standards based on the HD2027 Standards
Regulatory Class
Off-cycle
Bin
Off-Cycle NOx Standards (g/hr for idling,
g/hp-hr for non-idle operation)
LHD
Idle
9 for MY2027-2028; 13.5 forMY2029+A
Non-idle
0.052 for MY2027-2028, 0.051 forMY2029+B
MHD
Idle
9 for MY2027-2028; 13.5 forMY2029+A
Non-idle
0.065 for MY2027-2028; 0.064 forMY2029+B
HHD
Idle
9 for MY2027-2028; 13.5 forMY2029+A
Non-idle
0.065 for MY2027-2028; 0.064 forMY2029+B
A The optional idle standards were 10 g/hr in the HD2027 final standards. However, MOVES uses the values in the table
to specifically account for the compliance margin to the off-cycle standards.
B The non-idle standards in the table are different from the HD2027 final standards values (0.058 g/hp-hr and 0.073
g/hp-hr for LHD and MHD/HHD/Urban bus, respectively). They are now based on the modeling of off-cycle standards
compliance margin in MOVES4.
In order to apply these off-cycle standards to MOVES operating modes, we first converted the
standards to a fuel consumption (i.e., CO2 emissions) basis. We calculated the optional idle NOx
g/hr standard in units of NOx g/C02 kg using Equation 2-23 and the resulting values are displayed
in Table 2-13.
47

-------
N0X
Optional Idle standard
C02 " " " Vkg/
(£
= [Optional Idle NOx standard
Idle CO-
(kg\
Vhr.
Equation 2-23
Where Idle C02 (j^)= the MOVES average C02 (kg/hr) emission rate for HHD diesel vehicles for MOVES idle
(operating mode 1). We assume the CO2 rates are unaffected by the HD2027 standard.
Table 2-13 Calculation of Optional Idle NOx /CO2 Standard (g/kg)
Optional Idle NOx standard (g/hr)
Average HHD Idle
(Operating Mode=l) CO2
emission rate (kg/hr)
Optional Idle NOx /CO2
standard (g/kg)
9 for MY2027-2028; 13.5 for MY2029+
7.68
1.17 for MY2027-2028;
1.13 for MY2029+
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 2-14 in Columns (F) for HHD vehicles for the final standards.
In Table 2-14, Column (B) lists the MOVES CO2 emission rate for Model Year 2027 HHD diesel
vehicles for each MOVES operating mode. Column (C) lists the mean power for each operating
mode bin as 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 (see Section 2.1.4.3).21 The percent load,
Column (D), is calculated for each operating mode bin using Equation 2-24.
Mean Power opMode=i
Percent Load0pMode=i - Mean PowerOpMode=40	Equation 2-24
Mean Power 0pMode=i
Percent Load0pMode=i - Mean PowerOpMode=40	Equation 2-24
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)) as defined in the rule, where percent load less than 6 percent of
maximum power is defined as idle and above 6 percent of maximum power is non-idle load.
48

-------
Table 2-14 Calculation of the Off-cycle NOx Standard Compliant Emission Rate for HHD Diesel Vehicles for the
HD2027 Standards
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-
2028 off-
cycle
compliant
emission
rate (g/hr)
MY 2029+
off-cycle
compliant
emission
rate (g/hr)
0
4.92
6.04
1.3%
Idle
5.76
5.57
1
7.68
8.10
1.7%
Idle
9.00
8.70
11
13.42
1.04
0.2%
Idle
15.72
15.19
12
21.69
28.90
6.2%
Non-idle
1.88
1.85
13
37.31
75.16
16.1%
Non-idle
4.89
4.81
14
52.20
121.18
26.0%
Non-idle
7.88
7.76
15
68.68
166.98
35.8%
Non-idle
10.85
10.69
16
110.42
282.24
60.5%
Non-idle
18.35
18.06
21
13.92
-1.61
-0.3%
Idle
16.31
15.76
22
32.99
34.43
7.4%
Non-idle
2.24
2.20
23
44.71
77.71
16.7%
Non-idle
5.05
4.97
24
59.82
121.62
26.1%
Non-idle
7.91
7.78
25
77.03
167.82
36.0%
Non-idle
10.91
10.74
27
102.53
230.56
49.4%
Non-idle
14.99
14.76
28
142.09
327.41
70.2%
Non-idle
21.28
20.95
29
181.90
403.76
86.5%
Non-idle
26.24
25.84
30
212.63
470.01
100.7%
Non-idle
30.55
30.08
33
28.36
34.76
7.4%
Non-idle
2.26
2.22
35
71.87
145.03
31.1%
Non-idle
9.43
9.28
37
106.93
227.23
48.7%
Non-idle
14.77
14.54
38
148.35
323.23
69.3%
Non-idle
21.01
20.69
39
183.17
396.00
84.9%
Non-idle
25.74
25.34
40
196.42
466.62
100.0%
Non-idle
30.33
29.86
Finally, the off-cycle NOx standard compliant emission rate in Column (F) was 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) by the N0x/C02 off-cycle idle
standard calculated in Table 2-13 using Equation 2-25.
Idle Emission Rate
= MOVES MY 2027 HHD C02 Emission Rate ®
Vnr/	Equation 2-25
NOx	( g \
x Voluntary Idle inusestandard J
49

-------
For the operating modes classified as non-idle load, we multiplied the off-cycle (g/hp-hr) standard
in Table 2-12 for the corresponding power classification by the mean power (Column C), as shown
in Equation 2-26.
Non — idle Emission Rate
= Mean Power(hp) x In_use standard (-—]	Equation 2-26
\hp ¦ hr/
The estimated off-cycle NOx standard compliant emission rates for heavy heavy-duty diesel
vehicles are shown in Figure 2-18. Similarly, we applied Equation 2-24 through Equation 2-26 to
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.
I Without HD2027
HD2027 Off-Cycle Standard Compliant Rates for MY2027-2028
HD2027 Off-Cycle Standard Compliant Rates for MY2029+
3 300

E
LU
X
O
Z
w
LU
>
O
250
200
150
100
50


























































































ix.
1.
L





1





1 ¦
1.
I,

0 1 11 12 1314151621 2223242527282930333537383940
MOVES Operating Mode
Figure 2-18 Base NOx rates and off-cycle NOx standard compliant emission rates for HHD diesel
2.1.1.6.3 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 2-19 shows the HHD duty-cycle and off-cycle standards for MY 2027+ across
MOVES operating modes. The duty-cycle standard is estimated to constrain emissions in four
operating modes (operating modes 0, 21, 35 and 37), while the off-cycle standard constrains
emissions in the remaining operating modes.
50

-------
	Duty Cycle Compliant Rates for MY2027+
	Off-cycle Compliant Rates for MY2027-2028
	Off-cycle Compliant Rates for MY2029+
=5 120
3
% 100
a:
.! 80
CO
<2
E 60
LU
| 40
w
£ 20
O
5 0
Figure 2-19 Comparison of Running NOx emission rates for diesel-fueled HHD compliant with the duty-cycle
and off-cycle standards
Because manufacturers will need to simultaneously 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 off-cycle standards is selected for operating mode 12, but the emission rate based on
the duty-cycle standards is selected for operating mode 35). Figure 2-20, Figure 2-21 and Figure
2-22 present the estimated emission rates for HHD, MHD, and LHD45 diesel vehicles that meet
both the final duty-cycle and off-cycle standards, respectively.


1-25 mph


25-50 mph

>50 mph






































































/














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
51

-------
i Without HD2027	
HD2027 Standards Rates for MY2027-2028
HD2027 Standards Rates for MY2029+
u
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 2-20 Estimated zero-mile NOx emission rates for HHD diesel vehicles due to the HD2027 duty-cycle and
off-cycle standards
i Without HD2027
HD2027 Standards Rates for MY2027-2028
HD2027 Standards Rates for MY2029+
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 2-21 Estimated zero-mile NOx emission rates for MHD diesel vehicles due to the HD2027 duty-cycle and
off-cycle standards
52

-------

¦ Without HD2027
i HD2027 Standards Rates for MY2027-2028
HD2027 Standards Rates for MY2029+





































































































1







¦









¦



M	
1.
ll
1.
L
.



Ii
1.
L
L
.

_

_
_
1
i.
1.
L
.
_

_
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 2-22 Estimated zero-mile NOx emission rates for LHD45 diesel vehicles due to the HD2027 duty-cycle
and off-cycle standards
2.1.1.7 Tampering and Mai-maintenance
MOVES accounts for the fleet-average increase in emissions with vehicle age. For heavy-duty
diesel vehicles, the increase with age is modeled in the development of the emission rates by
applying tampering & mal-maintenance (T&M) adjustment factors to age zero rates. We assume
that the T&M effects are the dominant cause of emission increases with age in heavy-duty diesel
vehicles. No effort has been made to explicitly model emission increases due to normal wear and
tear (e.g. catalyst degradation, sensor deterioration from properly maintained vehicles) from heavy-
duty diesel vehicles.
Table 2-15 shows the percent increases due to T&M adjustment factors for aggregate NOx
emissions by regulatory class and model year group. As described in Appendix B, the T&M
adjustment factors in Table 2-15 are calculated by combining information regarding the assumed
frequency rate of an equipment failure at the useful life of the engine with the estimated emission
impact of the equipment failure.
53

-------
Table 2-15 NOx T&M adjustment factor (percent) by regulatory class
Model years
LHD2b3 trucks (%)
LHD45 trucks (%)
MHD, HHD, Urban Bus (%)
1994-1997
0
0
0
1998-2002
0
0
0
2003-2006
0
0
0
2007-2009
3.8
3.84
0
2010-2012
55.9
77.9
77.9
2013-2026
58.6
58.6
58.6
2027-2028
58.6
329.7
324.0 (MHD),
410.5 (HHD),
433.7 (Urban Bus)
2029-2060
58.6
331.6
325.6 (MHD),
416.3 (HHD),
440.6 (Urban Bus)
We also apply an age effect; that is, we assume that emissions begin to deteriorate at the age
vehicles pass the warranty requirements and increase to the full T&M adjustment factors at the age
the vehicles reach the useful life mileage requirement (Figure B-l). Because we expect the
warranty period for the HHD, MHD, and Urban Buses will be reached within the first three years
(Table B-l), the T&M adjustments factors are applied starting with the age 0-3 group for these
vehicles. Similarly, the T&M adjustments factors are applied starting in the age 4-5 group for the
LHD45 and LHD2b3 vehicles.
The emission rates described in Sections 2.1.1.4 through Section 2.1.1.6 are assumed to represent
new or "zero-mile" emission rates. The 2007-2009 LHD emission rates are derived from the 2003-
2006 HDIUT data as discussed in Section 2.1.1.4.6. The MY 2010 and later heavy-duty emission
rates are also based on the HDIUT data which tested newer and well-maintained vehicles. We then
account for the emission increases with age by multiplying the MY 2007-2009 LHD emission
rates, and MY-2010-and-later emission rates by a function of the corresponding T&M adjustment
factors and age effects (Equation 8-3). The MOVES emission rates for regulatory classes with the
same zero-mile emission rates (Table 2-16) are different due to the T&M NOx effects (Table 2-15)
and phase-in of T&M effects by age (Table B-4).
The T&M adjustments for NOx are zero for some of the model year groups because these vehicles
lack the heavy-duty advanced aftertreatment systems that we assume are most affected by
tampering and mal-maintenance, as discussed more in Appendix B.
The LHD vehicles have different T&M NOx increases than HHD, MHD, and Urban Bus vehicles
due to the penetration of lean NOx trap (LNT) aftertreatment. For MY 2007-2009 we assumed that
there was a 25 percent penetration of LNT-equipped vehicles within both LHD2b3 and LHD45
regulatory classes, with the remaining 75% having no NOx aftertreatment equipment. Subsequent
certification data shows that LNT was not actually used in LHD45 vehicles, however, we have not
updated this assumption, or the T&M adjustment factors, for MY 2007-2009 because the resulting
LHD45 2007-2009 emission rates compare well to the HDIUT data for 2007-2009 vehicles.34 For
MY 2010-2012, we assumed a 25 percent penetration of LNT and 75 percent of selective catalytic
54

-------
reduction (SCR) NOx aftertreatment system within the LHD2b3 regulatory class. We assume that
all heavier heavy-duty vehicles (LHD45, MHD, HHD, Urban bus) have 100 percent penetration of
SCR systems starting in model year 2010. For model years 2027+, the NOx standard is much more
stringent and we assume that aftertreatment systems are much more effective in reducing
emissions. As detailed in Section B.7, we also assume the converse—that aftertreatment failure in a
MY2027 vehicle would bring NOx emissions to the same level as estimated for a MY 2010
vehicle's engine-out emissions without any reductions from aftertreatment system.
The T&M values for model year 2010 and later vehicles also account for implementation of OBD.
For LHD2b3 trucks, OBD systems were assumed to be fully implemented in MY 2010. For the
other HD regulatory classes (LHD45, MHD, HHD), we assumed there would be a phase-in period
from MY 2010 to 2012 where one-third of those trucks were equipped with OBD systems. In MY
2013 and later, all trucks have OBD systems. These OBD adoption rates have been incorporated
into the tampering and mal-maintenance emission increases in Table 2-15 based on the
assumptions and calculations detailed in Appendix B.
Figure 2-23 displays the NOx rates in gram per mile by age and regulatory class for model year
2015 vehicles. Due to faster mileage accumulation, the HHD trucks tend to reach their useful life
quicker than other heavy-duty vehicles, with the maximum emission rates by the 4-5 age group. In
contrast Urban Buses do not reach the useful life until the age 10-14 group. The gram per mile
(g/mile) emission rates were calculated outside the model by using operating mode distributions
and average speeds for each regulatory class estimated from a national MOVES run. The emission
rates showed here do not include any adjustments to the rates due to fuel effects or humidity that
are applied during MOVES run, however they do incorporate differences due to activity.1
Despite the significant T&M adjustment factors for MY 2010+ heavy-duty diesel vehicles, the
fully-aged NOx emissions of MY 2010+ vehicles are significantly lower than the emission rates of
MY 2009 and earlier vehicles. As shown in Figure 2-25, the age 0-3 MY 2009 NOx g/mile
emission rates are over 100% higher than the MY 2010 vehicle emission rates, while the T&M
adjustment factors are less than 100% (Table 2-15). Thus, MOVES estimates that, on average, a
fully-aged MY 2010 heavy-duty truck has lower NOx emissions than a new MY 2009 heavy-duty
truck.
1 For example, the MY 2015 NOx emission rates by operating mode for LHD2b3 and LHD45 diesel vehicles are the
same, but Figure 2-17 shows different gram per mile emission rates due to vehicle and activity differences, including
heavier weights of LHD45 vehicles (source mass), and because the two regulatory classes are distributed differently
among the source types in MOVES, which have different operating mode distributions. See the MOVES Population
and Activity ReportError! Bookmark not defined, for more information. The zero-mile HHD and Urban Bus NOx
emission rates by operating mode are equivalent. However, the emission rates in the age groups: 0-3, 4-5, 6-7, and 8-9
are different because they have a different phase-in of the T&M effects. The difference between HHD and Urban Bus
in grams per mile for ages 10+ are due to differences in the operating mode distributions and average speeds.
55

-------
Reg Class
-+¦ 41-LHD2b3
-+¦ 42-LHD45
46-MHD67
-+¦ 47-HHD8
48-Urban Bus
0-3	4-5	6-7	8-9	10-14	15-19	20+
Age (years)
Figure 2-23 Base Heavy-duty Diesel NOx Running Emission Rates (g/mile) by Age for Model Year 2015 by
Regulatory Class Estimated using Nationally Representative Operating Mode Distribution
2.1.1.8 Summary and Model Year Trends
Table 2-16 summarizes the methods used to estimate emission rates for each regulatory class and
model year group combination. The emission rates in MOVES are based on the analysis of
ROVER, Consent Decree testing data, and HDIUT data. Using the HDIUT data, we updated the
HHD, MHD, LHD, and Urban Bus rates for MY 2010+ vehicles in MOVES3. Rates for MY 2027+
were updated in MOVES4.
56

-------
Table 2-16 Summary of Methods for Heavy-Duty Diesel NOx Emission Rate Development for Each Regulatory
			Class and Model Year Group		
Model
HHD
MHD
Urban Bus
LHD45
LHD2b3
year group
(regClass 47)
(regClass 46)
(regClass 48)
(regClass 42)
(regClass 41)

HHD 1991-

Urban Bus 1991-


1960-
1989,
1990
1997 rates

1997 rates


proportioned to
ratio of
Same rates as
HHD
proportioned
using ratio of
LHD 1991-1993 rates proportioned to
LHD certification levels
certification
levels

HHD certification
levels


1991-1997
Data analysis30,
with
adjustments for
Low-NOx
Same rates as
HHD
Data analysis3
LHD 1999-2002 rates proportioned to
1991-1997 FTP standards per Table 2-1

rebuilds







Urban Bus 1999-



Data analysis3 0,

2002 rates


1998
with
adjustments for
Low-NOx
Same rates as
HHD
proportioned
using ratio of
HHD 1998 rates
Same rates as 1999-2002

rebuilds

to HHD 1999-
2002 rates


1999-2002
Data analysis3'0
Data analysis3
Data analysis3
MHD data analyzed with 2.06 fscaie
2003-2006
Data analysis3 0
Data analysis3 0'
Data analysis3
Data analysis b


MHD 2003-
Urban Bus 2003-




2006 rates
2006 rates
Percent reductions from the
2007-2009
Data analysis'3
proportioned to
proportioned to
MOVES2010 LHD 2003-2006 rates


FTP standards
FTP standards per
(Section 2.1.1.4.6)°


per Table 2-1°
Table 2-1


2010-
2018
HHD data
analysis'3 with
MY specific
production
volume
weighting
MHD data
analysis'3 with
MY specific
production
volume
weighting
Same as HHD
except T&M
adjustment
factors specific to
Urban Bus
LHD data
analysis'3 with MY
specific
production
volume weighting;
T&M specific to
LHD45
LHD data
analysis'3 with MY
specific production
volume weighting;
T&M specific to
LHD2b3 & Tier 3
reductions starting




in MY 2018





Same baseline as





LHD2b3 MY 2018
2019-2026
Same as HHD
Same as MHD
Same as Urban
Same as LHD45
with Tier 3
MY 2018
MY 2018
Bus MY 2018
MY 2018
reduction phase-in
from MY 2019-
2022

Revised rates
based on
HD2027
standards
Revised rates
based on
HD2027
standards
Revised rates
Revised rates
Same as 2019-
2026 rates
2027-2060
based on HD2027
based on HD2027

standards
standards
Notes:
a Analysis based on ROVER and Consent Decree testing data
b Analysis based on HDIUT data
0 Confirmed by HDIUT and/or Houston Drayage Program data
57

-------
The role of the model year groups, representing a rough surrogate for technology or standards, can
be seen in Figure 2-24, which shows NOx emission rates for 0-3 age group by model year and
regulatory class estimated in grams per mile (g/mile) using nationally representative operating
mode distributions and average speeds. The MOVES model estimates drastic reductions in NOx
emissions from pre-1990 technologies to modern 2027 and later model year engines. Some of the
differences in the emission rates for the 0-3 age group are caused by different phase-in of T&M
adjustment factors, which began to be applied in the first age group for HHD, MHD, and Urban
Bus regulatory classes (Section 2.1.1.5.5). Because the nationally representative operating modes
are different across model years and regulatory classes, some of the differences between model
years and the regulatory classes are due to activity differences and not the emission rates by
operating mode.1
The figure also shows NOx emissions for "gliders", which are trucks with a new chassis and cab
assembly and equipped with a rebuilt engine typically without an exhaust aftertreatment system.
These emissions are of similar magnitude to emission rates from the model year 2000 HHD
vehicles as discussed in Section 2.5. In MOVES4, zero population of gliders is estimated starting
MY2020+.
Figure 2-25 shows the model year trend for NOx in gram per mile for the 2007 and later model
years. The reduction in LHD2b3 emission rates due to the phase-in of the Tier 3 standards between
2018 and 2022 is evident. All other heavy-duty emission rates by operating mode are unchanged
from model year 2018-2026 until the HD2027 rule introduces additional reductions for MY2027+
as documented in Table 2-16. The slight reductions observed in the gram per mile emission rates
for all heavy-duty vehicles (including gliders) in model years 2014 through 2027 are due to
changes in the operating mode distribution from lower weights, lower rolling resistance, and
improved aerodynamics of trucks implemented in the Phase 1 and Phase 2 Heavy-duty Greenhouse
Gas Regulations as documented in the MOVES4 Population and Activity Report.6
58

-------
1980
2000
2020
2040
Reg Class
41-LHD2b3
¦	42-LHD45
46-MHD67
¦	47-HHD8
48-Urban	Bus
49-Gliders
Model Year
Figure 2-24 Base running emission rates for NOx from age 0-3 diesel heavy-duty vehicles averaged over a
nationally representative operating mode distribution.
Reg Class
41-LHD2b3
¦	42-LHD45
46-MHD67
¦	47-HHD8
48-Urban Bus
2010
2020
2030
2040
Model Year
Figure 2-25 Base running emission rates for NOx from age 0-3 gasoline heavy-duty vehicles averaged over a
nationally representative operating mode distribution for model years 2007 through 2040.
59

-------
2.1.2 Particulate Matter (PM2.5 )
PM2.5 refers to particulate matter with a mean aerodynamic diameter less than 2.5 microns.
Particulate matter is a complex mixture of particles that are composed of one or many of the
following substances: organic material, metals, elements, and ions, including sulfate.
Measurements from which the MOVES PM2.5 emission rates are based on are typically filter-based,
which include the mass of all the chemical components in the particle-phase. MOVES reports PM
emissions in terms of elemental carbon (EC) and the remaining non-elemental carbon PM
(nonECPM).
As described above for NOx, the heavy-duty diesel PM2.5 emission rates in MOVES are a function
of: (1) fueltype, (2) regulatory class, (3) model year group, (4) operating mode, and (5) age group.
We classified heavy-duty diesel exhaust PM emission data into the following model year groups
for purposes of emission rate development. These groups are generally based on the introduction of
emissions standards for heavy-duty diesel engines. They also serve as a surrogate for continually
advancing emission control technology on heavy-duty engines. For example, MY 2010 and beyond
is defined as a separate group even though the PM standard is unchanged from the previous group
MY 2007-2009. This is because, starting with MY 2010, the wide adoption of SCR systems and
improvements in DPFs likely resulted in a shift in tailpipe PM2.5emissions. Other, secondary
reasons include wider availability of PM2.5 data in the HDIUT data set (section 2.1.1.1) for MY
2010-2011 and the fscale updates to all HD regClasses for MY 2010 and later (see section 2.1.1.4.2
and Appendix G). Table 2-17 shows the model year group ranges and the applicable brake-specific
emissions standards.
Table 2-17 Model Year Groups Used for Analysis Based on the PM Emissions Standard
Model Year Group
PM Standard (g/bhp-hr)
Range

1960-1987
No transient cycle
standard
1988-1990
0.60
1991-1993
0.25
1994-1997
0.10
1998-2006
0.10
2007-2009
0.01
2010+
0.01
Section 1.2.1 and Section 2.1.2.2 discuss the derivation of the MY 1960-2009 and 2010-2060
heavy-duty diesel PM2.5 emission rates, respectively. The discussion of the tampering and mal-
maintenance factors applied to emissions in both model year groups are discussed in Section
2.1.2.3.
60

-------
2.1.2.1 1960-2009 Model Years
The PM2.5 data from was analyzed in several steps to obtain MY 1960-2009 PM2.5 emission rates.
First, STP operating mode bins were calculated from the chassis dynamometer data from the CRC
E55/59. Second, continuous PM2.5 data measured by the TEOM was normalized to gravimetric PM
filters. Third, MOVES PM2.5 emission rates were calculated for the STP operating mode bins for
the available regulatory class and model year combinations. Then, steps were taken to estimate
missing operating modes, regulatory classes, and model years from the E55/59 program. In
addition, we estimate the EC/PM fraction and adjust the emission rates to account for tampering
and mal-maintenance. The E55/59 data and analysis steps are explained in detail in the following
subsections.
2.1.2.1.1 Data Sources
All of the data used to develop the MOVES PM2.5 emission rates for MY 1960-2009 was generated
based on the CRC E-55/59 research program.41 The following description in the "Compilation of
Diesel Emissions Speciation Data - Final Report'42 provides a good summary of the program:
The objective of the CRC E55/59 test program was to improve the understanding of the
California heavy-duty vehicle emissions inventory by obtaining emissions from a
representative vehicle fleet, and to include unregulated emissions measured for a subset of
the tested fleet. The sponsors of this project include CARB, EPA, Engine Manufacturers
Association, DOE/NREL, and SCAQMD. The project consisted of four segments,
designated as Phases 1, 1.5, 2, and 3. Seventy-five vehicles were recruited in total for the
program, and recruitment covered the model year range of 1974 through 2004. The number
and types of vehicles tested in each phase are as follows:
Phase 1: 25 heavy heavy-duty (HHD) diesel trucks
Phase 1.5: 13 HHD diesel trucks
Phase 2: 10 HHD diesel trucks, 7 medium heavy-duty (MHD) diesel trucks,
2 MHD gasoline trucks
Phase 3: 9 MHD diesel, 8 HHD diesel, and 2 MHD gasoline
The vehicles tested in this study were procured in the Los Angeles area, based on model
years specified by the sponsors and by engine types determined from a survey. WVU
measured regulated emissions data from these vehicles and gathered emissions samples.
Emission samples from a subset of the vehicles were analyzed by Desert Research Institute
for chemical species detail. The California Trucking Association assisted in the selection of
vehicles to be included in this study. Speciation data were obtained from a total of nine
different vehicles. Emissions were measured using WVU's Transportable Heavy-Duty
Vehicle Emissions Testing Laboratory. The laboratory employed a chassis dynamometer,
with flywheels and eddy-current power absorbers, a full-scale dilution tunnel, heated probes
and sample lines and research grade gas analyzers. PM was measured gravimetrically.
Additional sampling ports on the dilution tunnel supplied dilute exhaust for capturing
unregulated species and PM size fractions. Background data for gaseous emissions were
gathered for each vehicle test and separate tests were performed to capture background
samples of PM and unregulated species. In addition, a sample of the vehicles received
61

-------
Tapered Element Oscillating Microbalance (TEOM) measurement of real time particulate
emissions.
The HHDDTs were tested under unladen, 56,000 lb., and 30,000 lb. truck load weights. The
driving cycles used for the HHDDT testing included:
•	AC50/80;
•	UDDS;
•	Five modes of an HHDDT test schedule proposed by CARB: Idle, Creep, Transient,
Cruise, and HHDDT S (a high-speed cruise mode of shortened duration);
•	The U.S. EPA Transient test.
The CARB HHDDT test cycle is based on California truck activity data and was developed
to improve the accuracy of emissions inventories. It should be noted that the transient
portion of this CARB test schedule is similar but not the same as the EPA certification
transient test.
The tables below provide a greater detail on the CRC E-55/59 data used in the analysis. Both the
number of tests and number of vehicles by model year group and regulatory class (MHD, HHD)
are provided in Table 2-18.
Table 2-18 Vehicle and Test Counts by Regulatory Class and Model Year Group
Regulatory
Class
Model Year
Group
Number of
tests
Number of
vehicles

1960- 1987
82
7

1988 - 1990
39
5
MHD
1991 - 1993
22
2

1994 - 1997
39
4

1998 - 2006
43
5

2007 - 2009
0
0

1960- 1987
31
6

1988 - 1990
7
2
HHD
1991 - 1993
14
2
1994 - 1997
22
5

1998 - 2006
171
18

2007 - 2009
0
0
Counts of tests are provided by test cycle in Table 2-19.
62

-------
Table 2-19 Vehicle Test Counts by Test Cycle
Test Cycle
Number of tests
CARB-T
71
CARB-R
66
CARB-I
42
UDDSW
65
AC5080
42
CARB-C
24
CARBCL
34
MHDTCS
63
MHDTLO
23
MHDTHI
24
MHDTCR
29
2.1.2.1.2	Calculate STP from Second-By-Second Data
For each second of operation on the chassis-dynamometer the instantaneous scaled tractive power
(STPt) was calculated using Equation 1-6, and then subsequently classified to one of the 23
operating modes defined above in Table 1-4.
The values of coefficients A, B, and C are the road-load coefficients pertaining to the heavy-duty
vehicles as determined through previous analyses for EPA's Physical Emission Rate Estimator
(PERE).43 The chassis dynamometer cycles used in E55/59 include the impact of speed,
acceleration, and loaded weight on the vehicle load, but grade effects are not included and the
grade value is set equal to zero in Equation 1-6.
Note that this approach differs from the NOx emission rates analysis described in Section 2.1.1.3,
since the particulate data was collected on a chassis dynamometer from vehicles lacking electronic
control units (ECU). We have not formally compared the results of the two methods of calculating
STP. However, on average, we did find the operating-mode distributions to be similar between the
two calculation methods for a given vehicle type. For example, we found that the maximum STP in
each speed range was approximately the same.
2.1.2.1.3	Compute Normalized TEOM Readings
The TEOM readings were obtained for a subset of tests in the E-55/59 test program. Only 29
vehicles had a full complement of 1-Hz TEOM measurements. However, the continuous particulate
values were modeled for the remaining vehicles by West Virginia University, and results were
provided to EPA. In the end, a total of 56 vehicles and 470 tests were used in the analysis out of a
possible 75 vehicles. Vehicles and tests were excluded if the total TEOM PM2.5 reading was
negative or zero, or if corresponding full-cycle filter masses were not available. Table 2-20
provides vehicle and test counts by vehicle class and model year. The MHD (Class 6 and Class 7)
trucks tested in the study included seven Class 6 and eleven Class 7 vehicles, representing only a
limited model years.
63

-------
Table 2-20 Vehicle and Test Counts by Heavy-Duty Class and Model Year
Model Year
HDD Class 6/7(MHD)
HDD Class 8 (HHD)
No. Vehicles
No. Tests
No. Vehicles
No. Tests
1969
-
-
1
6
1974
1
10
-
-
1975
-
-
2
10
1978
-
-
1
5
1982
1
5
-
-
1983
1
10
1
6
1985
1
28
1
10
1986
1
3
1
4
1989

11
1
4
1990
1
12
1
3
1992
1
11
1
11
1993
1
11
1
3
1994
1
9
3
15
1995
2
24
3
13
1998
2
20
3
28
1999
-
-
3
43
2000
2
18
5
44
2001
1
5
2
21
2004
-
-
4
29
2005
-
-
1
6
Since the development of MOVES emission rates is cycle independent, all available cycles/tests
which met the above requirements were utilized. As a result, 488,881 seconds of TEOM data were
used. The process required that each individual second-by-second TEOM rate be normalized to its
corresponding full-cycle filter mass, available for each combination of vehicle and test. This step
was necessary because individual TEOM measurements are highly uncertain and vary widely in
terms of magnitude (extreme positive and negative absolute readings can occur). Kinsey et al.
(2006)44 demonstrated that time-integrated TEOM measurements compare well with gravimetric
filter measurements of diesel-generated particulate matter. Equation 2-27 shows the normalization
process for a particular one second TEOM measurement.
P terj
^^normalized,i,j vT~DA/f	^^TEOM,j,i	Equation 2-27
2/ r™TEOM,i
Where:
/' = an individual 1-Hz measurement (g/sec),
j = an individual test on an individual vehicle,
PMteomjj = an individual TEOM measurement on vehicle j at second
PMfaterj= the total PM2 5 filter mass on vehicle j,
Phi normalized, ij = an estimated continuous emission result (PM25) emission result on vehicle j at second
64

-------
2.1.2.1.4	Compute Average Rates by MOVES Operating Mode
After normalization, the data were classified into the 23 operating modes defined in Section 1.6 by
regulatory class and model-year group. Mean average results, and standard deviation for PM2.5
emission values were computed in terms of g/hour for each operating mode. In cases where the
vehicle and TEOM samples were sufficient for a given mode (based on the number of points within
each operating mode bin), these mean values were adopted as the MOVES emission rates for total
PM2.5.
2.1.2.1.5	Populating Missing and High-Power Operating Modes
As detailed in Appendix E, a log-linear regression was performed on the existing PM2.5 data against
STP to fill in emission rates for missing operating mode bins. Similar to the NOx rates for MY
2009 and older vehicles, emission rates were extrapolated for the highest STP operating modes.
2.1.2.1.6	LHD and Urban Bus Emission Rates
The PM2.5 emission rates for LHD and Urban Buses are based on the available TEOM data
collected on MHD and HHD vehicles. We believe this is reasonable because the certification
standards in terms of brake horsepower-hour (bhp-hr) are the same for LHD, MHD, and HHD
regulatory classes.
The following steps were conducted to adjust the emissions estimated from the MHD and HHD
regulatory classes because the data were not analyzed for the fscale.used for LHD. The emission rates
of pre-2010 LHD (LHD2b3 and LHD45 (regClassID 41 and 42) are based on an fscale of 2.06 as
discussed in Section 1.5, whereas MHD and HHD are based on an fscaie of 17.1. The PM2.5 emission
rates for the pre-2010 LHD regulatory classes are derived from the VSP-based MHD PM2.5
emission factors derived from the E55/59 TEOM data as analysed for MOVES2009.45
With VSP-based emission rates, the power of the vehicle is scaled to the mass of the individual
tested vehicle. Because LHD have lower vehicle weights and power outputs than the MHD and
HHD vehicles, we scaled the VSP-emission rates down to the power requirements of the LHD
vehicles. To estimate the LHD2b3 and LHD45 PM2.5 emission rates, we multiplied the VSP-based
MHD PM2.5 emission rates by a factor of 0.46 obtained from the MOBILE6.2 heavy-duty
conversion factors46, which accounts for the lower power requirements per mile (bhp-hr/mile) of
light heavy-duty trucks versus MHD trucks. This scaling factor estimates VSP-based emissions
rates for LHD vehicles. We approximated the STP-based MOVES emission rates for LHD vehicles
using these VSP-based rates.-" Equation 2-28 used to derive the PM2.5 emission rates for LHD
regulatory class is shown below:
LHDPM? r emission rate	^
Fnii/itinii / —/7*
= 0.46 x MHD (ySPbased)PM2.s emission rate
¦> When this approximation was conducted in MOVES2010, the fSCau of 2.06 matched the average mass of LHD vehicles
for the source types. This approximation need to be revisited now that we have updated the mass of LHD vehicles in
MOVES to range from 3.5 to 7.8 metric tons.
65

-------
Despite the uncertainty used in this approach, the representative PM2.5 gram per mile emission rates
show reasonable trends (see Figure 2-33). LHD2b3 and LHD45 emission rates for pre-2010 model
years are lower than the MHD and HHD emission rates, but higher than the Urban Bus emission
rates.
Urban Bus emission rates are generally assumed to be the same as the HHD emission rates.
However, they have different emission deterioration effects as discussed in Appendix B.l. Also,
for some model year groups, the Urban Buses are subject to a different emission standard, so we
adjust the emission rates before deterioration by applying a ratio of the EPA certification standards.
Table 2-21 displays the model years for which the Urban Bus regulatory class has different PM
emission standards from other heavy-duty compression-ignition engines. For these model years
(1991-2006), the Urban Bus PM emission rates were set equal to the HHD emission rates
multiplied by the ratio in emission standards before applying deterioration. The gram per mile
emission rates for Urban Bus presented in Figure 2-33 show expected trends, with lower emission
rates than the other regulatory classes starting in MY 1991 through 2006.
Table 2-21 Urban Bus PM Standards in Comparison to Heavy-Duty Highway Compression Engine Standards
Engine
Model
Year
Standard for Heavy-
Duty Highway
Compression-Ignition
Engines (g/bhp-hr)
Standard for
Urban Buses
(g/bhp-hr)
Ratio in
standards
1991-19933
0.25
0.1
0.4
1994-1995
0.1
0.07
0.7
1996-2006
0.1
0.05
0.5
Note:
a The 0.1 g/bhp-hr US EPA Urban Bus standard began with model year 1993. In California, the 0.1 g/bhp-hr Urban Bus
standard began in 1991. MOVES assumes all Urban Buses met the stricter CA standard beginning in 1991.
2.1.2.1.7 Model Year 2007-2009 Vehicles (with Diesel Particulate Filters)
The heavy-duty diesel emission regulations were made considerably more stringent for PM2.5
emissions starting in model year 2007 - even considering the phase-ins and average banking and
trading program, the emission standard fell by a factor of ten from 0.1 g/bhp-hr to 0.01 g/bhp-hr.
This increase in regulatory stringency required the use of diesel particulate filter (DPF) systems on
heavy-duty diesels. As a result, the PM2.5 emission performance of diesel vehicles has improved
dramatically.
At the time of analysis (originally done for MOVES2014, but carried over to MOVES3 and
MOVES4 for the 2007-2009 vehicles), no continuous PM2.5 emissions data were available on the
2007-2009 model-year vehicles. However, heavy and medium heavy-duty diesel PM2.5 data were
available from the EPA engine certification program on model years 2003 through 2007. These
data provided a snapshot of new engine emission performance before and after the introduction of
particulate trap technology in 2007 and made it possible to determine the relative improvement in
PM2.5 emissions from model years 2003 through 2006 to model year 2007. This same relative
improvement was applied to the existing, operating mode-based, 1998-2006 model year PM2.5
running emission rates to estimate in-use rates for MY 2007-2009 vehicles.
66

-------
An analysis of the certification data is shown in Table 2-22 below. It suggests that the particulate
trap reduced PM2.5 emissions by a factor of 27.7. This factor is considerably higher than the relative
change in the certification standards (i.e., a factor of 10). The reason for the difference is that the
new trap-equipped vehicles certify at emission levels which are much lower than the standard, and
thus, create a much larger compliance margin than previous technologies could achieve.
As an additional check on the effectiveness of the trap technology, EPA conducted some limited in-
house testing of a Dodge Ram truck, and carefully reviewed the test results from the CRC
Advanced Collaborative Emission Study (ACES) phase-one program, designed to characterize
emissions from diesel engines meeting 2007 standards. The results from these studies demonstrated
that the effectiveness of working particulate traps is very high.47
Table 2-22 Average Certification Results for Model Years 2003-2007
Certification Model
Year
Mean
St. Dev.
n
(g/bhp-hr)a
2003
0.084
0.014
91
2004
0.088
0.013
59
2005
0.085
0.014
60
2006
0.085
0.014
60
2007
0.003
0.002
21
Note:a Average ratio from MYs 2003-2006 to MY 2007 is 27.7.
2.1.2.1.8 Elemental Carbon and Non-Elemental Carbon Emission Factors
Particulate matter from conventional (pre-2007) diesel engines is largely composed of elemental
carbon. Elemental carbon is often used synonymously with soot and black carbon. Black carbon is
important because of its negative health effects and its environmental impacts as a climate forcer.48
Elemental carbon from vehicle exhaust is measured with filter-based measurements using thermal
optical methods. Continuous surrogate measures of elemental carbon can also be made with
photoacoustic instruments.
MOVES models Total PM2.5 emissions by vehicle operating mode using elemental carbon (EC)
and non-elemental particulate matter carbon (NonECPM), as shown in Equation 2-29.
PM2.5 = EC + NonECPM	Equation 2-29
By having emission rates for EC and nonECPM for each operating mode, the MOVES design
permits the EC fraction of PM (EC/PM) to vary for each operating mode. In practice, the data used
to develop EC and nonECPM emission rates does not support such fine resolution, and the EC/PM
is the same across all the running exhaust operating modesexcept for the idle operating mode.
For pre-2007 diesel trucks, we developed EC and nonECPM emission rates by applying EC/PM
fractions to the modal-based emission rates. For the idle operating mode (opModelD 1), we applied
an EC/PM fraction of 46.4 percent from the PM2.5 speciation profile developed from the idle mode
from the UDDS tests from the E55/59 program. For all the other operating modes within the
67

-------
running emission process, we used an EC/PM fraction of 79.0 percent from the PM2.5 speciation
profile developed from the transient mode of the UDDS tests from the E55/59 program. The
development of the pre-2007 PM2.5 speciation profiles from the E55/59 program are documented in
the Onroad Speciation Report.1
For 2007-2009 DPF-equipped diesel engines, we used the EC/PM fraction of 9.98 percent
measured in Phase 1 of the Advanced Collaborative Emissions Study (ACES) Report.47 Diesel
particulate filters preferentially reduce elemental carbon emissions, resulting in the low percentage
of elemental carbon emissions. The average EC/PM fraction is based on four engines run on the
16-hour cycle which composes several different operating cycles. Because the fraction is based
upon a range of driving conditions, we applied the constant 9.98 percent EC/PM fraction across all
operating modes for the 2007+ diesel emissions rates, including the idle operating mode
(opModelD 1).
Figure 2-26 and Figure 2-27 show the operating mode trend for PM2.5 with the EC and nonECPM
fractions. As with NOx, the highest operating modes in each speed range will rarely be attained due
to the power limitations of heavy-duty vehicles and the high fscaie used for these model years, but
are included in the figures for completeness. At high speeds (greater than 50 mph; operating modes
> 30), the overall PM2.5 rates are lower than the other speed ranges. For pre-2007 model years
(Figure 2-26), the PM2.5 rates are dominated by EC (except for the idle operating mode, opModelD
1). With the introduction of DPFs in model year 2007 (Figure 2-27), we model the large reductions
in overall PM2.5 rates and the smaller relative EC contribution to PM emissions. Figure 2-35 shows
the PM2.5 gram per mile emission rates separated into the elemental carbon and non-elemental
carbon fractions for HHD vehicles across model years.
35
30
" 25
v 20
TO
i 15
c
TO
CU
^ 10
5
0
Figure 2-26 MHD Diesel PM2.5 Emission Rates for MY 2006 (age 0-3) by Operating Mode
¦	EC
¦	nonEC
llxl
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
Operating mode
68

-------

1.4

1.2

1
JZ



CD
0.8
TO

s_


0.6
Q.
C

TO

0J


0.4

0.2

0
I EC
InonEC
Mil
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
Operating mode
Figure 2-27 MHD Diesel PM2.5 Emission Rates for MY 2007 (age 0-3) by Operating Mode
2.1.2.2 2010-2060Model Years
The MY 2010+ HDIUT data set described in Section 2.1.1.1 and Table 2-2 was used to update
PM2.5 emissions rates for MY 2010+ vehicles. Operating modes (Table 1-4) were assigned to the 1
hz data using the method to calculate STP described in Section 2.1.1.3, and the updated fscale values
developed for the 2010+ MYNOx analysis described in Section 2.1.1.4.2 and Appendix G.
2.1.2.2.1 Estimating Base Rates
As compared to the NOx, THC, and CO data, the PM2.5 emissions data reported from the HDIUT
program had many missing, negative and zero values. When the PM2.5 emissions data are
distributed over the NOx FEL groups, the number of vehicles with valid measurements (non-zero or
non-negative) by regulatory class and model year group was very limited as shown in Table 2-23.
The number of vehicles with valid measurements for each model year group and regulatory class
are significantly smaller than the total number of vehicles tested, with the exception of the HHD
2014-2015 model year group.
69

-------
Table 2-23. HDIU Vehicles with Valid PM2.5 Measurements By Regulatory Class, Model Year Group and NOx
		FEL Group			


Valid Measurements by NOx FEL





Group

Total
Total
Reg
Model Year
0.2
0.35
0.5
Valid
Vehicles
Class
Group

Tested
LHD
2010-2013
52
0
4
56
64
2014-2016
19
0
2
21
32
MHD
2010-2013
20
6
6
32
55
2014-2015
17
61
61
29
51
HHD
2010-2013
57
9
26
92
139
2014-2015
48
4
0
52
55
Note:
1 Due to an absence of MHD 2014-2015 vehicles certified to 0.35 and 0.5 NOx in the HDIUT, we replicated the
HDIUT MHD 2010-2013 vehicles certified to 0.35 and 0.5 NOx FEL to the 2014-2015 model year group.
When evaluating the data by operating mode, the data are even more sparse. Table 2-23 shows the
total number of vehicles tested by model year group and regulatory class that had valid (positive)
PM2.5 emissions data in at least one operating mode. However, the number of vehicles with valid
measurements is less when the data is evaluated by operating mode. Table 2-24 shows the
minimum and maximum range of valid vehicle measurements by individual operating modes. The
operating modes with the smallest number of valid measurements tended to be for the high-power
operating modes (opModelD 29, 30, 39 and 40). The highest number of valid measurements tend to
be for the high speed (>50 mph) at moderate power (opModelD 33, 35, 37, and 38).
We addressed the issue of data sparsity in the PM2.5 emissions data by not using NOx FEL based
grouping and production volume weighting. By aggregating the data across NOx FEL groups, there
was sufficient data to divide the data into the two model year groups used for the NOx analysis:
MY 2010-2013 and 2014-2015/2016. Note that for the current HDIUT dataset, LHD includes MY
2016 vehicles, but MHD and HHD only include up to MY 2015.
Table 2-24. Range of Valid Vehicle Measurements by Operating Mode by Regulatory Class and Model Year
			 Group		



Minimum Valid



Total Valid Vehicle
Vehicle



Measurements
Measurements for
Maximum Valid Vehicle

Model Year
(in at least one
any operating
Measurements for any

Group
operating mode)
mode
operating mode
LHD
2010-2013
56
11
52
2014-2016
21
3
18
MHD
2010-2013
32
4
27
2014-2015
29
3
26
HHD
2010-2013
92
17
83
2014-2015
52
4
45
The operating mode-based PM2.5 emission rates were estimated using Equation 2-30 and Equation
2-31, which are similar to Equation 2-13 and Equation 2-14 used for the development of NOx
emission rates except the FEL grouping factor is removed. Zero or negative emission rates are as
70

-------
treated as missing values. Despite the sparseness of the data by operating mode, due to the revised
fscale factors for 2010 and later vehicles (2.1.1.5.4), in general, there was no need for estimating
rates for missing high power operating modes.k
„ „	^sec ERpol,OM,C}MYG,veh}sec	, ,,,
ERpol,OM,c,MYG,veh = 		Equation 2-30
S^Ccount
„ n	^ivehERpol.OM.C.veh	, ,,
ERpol,OM,c,MYG —	7	Equation 2-31
vencount
Where:
C	= Regulatory class (LHD, MHD, HHD, and Urban Bus)
ERx,y,z = Emission rate in mass/time. The subscripts show the categorization
MYG	= Model year group (2010-2013 or 2014-2016)
MY	= Model year
OM	= running exhaust emissions operating mode
pol	= Pollutant (PM2 5)
sec; seCcomt = a second of data (for a given veh and OM); number of seconds in that category
veh; vehcomt = a vehicle (in the class); number of vehicles in that category
Similar to NOx, the PM2.5 rates for LHD2b3 and LHD45 are identical and based on the combined
LHD class vehicles in HDIUT, while the MHD rates are based on MHD class vehicles, and HHD
and Urban Bus rates are based on HHD class vehicles.
Because we did not use production volume weighting by NOx FEL, the PM2.5 emission rates are the
same for each model year within the two model year groups. The 2014-2015 model year group
sample contains a higher penetration of HHD vehicles certified to the 0.2 NOx FEL group, than the
2010-2013 model year HHD group (Table 2-23) which is consistent with the trend in production
volumes (Figure 2-12). The LHD vehicles had a similar proportion of vehicles certified to the 0.2
and 0.35 NOx FELs within the 2010-2013 and 2014-2016 model year groups. The PM2.5 emission
rates decrease for each of the regulatory classes between the MY 2010-2013 and 2014 and later
model years as observed in Figure 2-34.
There was no valid HDIUT PM data for 2014-2015 MY MHD vehicles within the 0.35 and 0.50
FEL groups. Because there are 2014-2015 MY MHD vehicles certified to these higher levels in the
production volume data (Figure 2-12), we supplemented the 2014-2016 MY MHD data with data
from the MY 2010-2013 MHD vehicles in the 0.35 and 0.5 FEL groups. As such, the MHD
vehicles have a similar proportion of vehicles certified to the 0.35 and 0.5 NOx FEL Groups within
the 2010-2013 and 2014-2015 model year groups.
Figure 2-28 through Figure 2-30 display the average PM2.5 emission rates by regulatory class,
model year group, and operating mode. The error bars are the 95% confidence intervals of the
mean, calculated by treating the mean emission rates from each vehicle within each operating mode
k The MHD 2010-2013 vehicles were an exception, missing data in operating modes 29, 30, and 40. Using the methods
outlined in Appendix 1.3 the emission rates in operating mode 29 and 30 were set equal to 28, and 40 was set equal to
39.
71

-------
as an independent random variable. While the average PM2.5 emission rates by individual operating
mode bin and regulatory class generally are not significantly different between the 2010-2013 and
2014-2016 model year groups, when you look across the operating modes, there is a fairly
consistent decrease in PM2.5 emission rates in the 2014-2016 compared to the 2010-2013 model
year groups. The observed decrease in PM2.5 emission rates in the 2014+ model year emission rates
is consistent with the decrease in extended idle PM2.5 emissions observed with the full-phase in of
SCR engines1 (Section 2.3.2.2).
0.60
0.50
0.40 -
0.30
q: 0.20 -
0.10 :

¦ MY 2010-2013
1
:

¦ MY 2014-2016
1



1












j I 11

I
; t 11
Tt II
i X
1 i. It II II 1
Hi- 1- 1, 1 1.


11 !L 1 III
I 11
M
E
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
MOVES OpMode
Figure 2-28 PM2.5 Emissions by Operating Mode for LHP Model Year Groups 2010-2013 and 2014-2016
1 As discussed in Section 2.3.2.2, we believe the reduction in THC and PMiswith the SCR systems is because the SCR
aftertreatment classification is a surrogate for the combined engine control and aftertreatment system used with SCR
equipped trucks that have a large impact on THC emissions. With the use of SCR, engines can be calibrated to run
leaner, producing lower engine-out PM25. Additionally, SCR systems rely on oxidation catalysts and/or catalyzed
DPFs to convert NO to NO2, which also reduces PM2 5 tailpipe emissions.
72

-------
Li
I MY 2010-2013
l MY 2014-2015
1,1. LI

0 1
11 12 13 14 15 16
21 22 23 24 25 27 28 29 30
MOVES OpMode
33 35 37 38 39 40
Figure 2-29 I'M:.? Emissions by Operating Mode for MHT) Model Year Groups 2010-2013 and 2014-2015

0.20 q

0.18 :

0.16 :

0.14 :
"UT
0.12 i


ClO
£
0.10 :

0.08 :
CL
0.06 j

0.04 :

0.02 i

0.00 ^
\, ii
0 1 11
I MY 2010-2013
l MY 2014-2015
12 13 14 15 16
iii
33 35 37 38 39 40
21 22 23 24 25 27 28 29 30
MOVES OpMode
Figure 2-30 PM2.5 Emissions by Operating Mode for HHD Model Year Groups 2010-2013 and 2014-2015
As observed in Figure 2-28 through Figure 2-30, the PM2.5 emission rates are highest for the MHD
vehicles, followed by the HHD, and then the LF1D vehicles. The corresponding uncertainty of the
MHD rates is also largest compared to the other rates. We do not have a reason to suspect the
accuracy of the HDIUT PM2.5 measurements from the MFID vehicles, and we can only speculate on
the reasons the MFID emission rates would be higher than the other regulatory classes. For
example, the fraction of vehicles produced in the less stringent 0.35 and 0.5 NOx FEL groups is
higher for the MFID than most of the other regulatory class and model year combinations, but not
all (Table 2-23).
Because much of the FIDIUT PM data are missing or reported as zero, and given the additional
uncertainty regarding the MFID rates, we compared our HDIUT-based PM2.5 rates against values
73

-------
reported in the literature. As shown in Figure 2-34, the MOVES age 0-3 PM2.5 rate ranges from 3 to
7 mg/mile for MY 2010-2014 HHD and LHD vehicles with the MHD significantly higher at 26
mg/mile and from 2 to 7 mg/mile for all HD MY 2014+ vehicles. Other studies have reported PM2.5
rates in the range of 1-7 mg/mi for MY 2010+ vehicles equipped with DPF and SCR and certified
to NOx standard of 0.20 g/bhp-hr.49'50'51 The rates from the MOVES run and other studies are
dependent on driving cycle, however, since the MOVES rates are generally within the range of
reported values, we believe it is reasonable to use the HDIUT-based PM2.5 data for the update for
all regulatory classes.
The MY 2014-2016 emission rates are applied to all future model yearsfor each regulatory class.
The Tier 3 rulemaking sets PM FTP emission standards for Class 2b and Class 3 of 8 mg/mile and
10 mg/mile respectively.52 Because the age 0-3 LHD2b3 MOVES emissions rates are well below
the Tier 3 standard, we do not estimate a reduction in zero-mile PM2.5 rates with the phase-in of
Tier 3 in MY 2018 and later vehicles as we do for NOx emissions. For PM2.5, the MOVES LHD2b3
rates are in compliance with the 3 mg/mile standard as shown in Figure 2-34 However, we do
incorporate different aging effects due to the extended useful life in the Tier 3 program as discussed
below in Section 2.1.2.3.2.
2.1.2.2.2	Elemental Carbon and Non-Elemental Carbon Emission Rates
The EC (9.98 percent) and non-EC (90.02 percent) fractions for MY 2010 and later are unchanged
from MY 2007-2009 analysis described in section 2.1.2.1.8. As discussed in the Speciation
Report1, the ACES Phase 1 is deemed to be an appriorate source of the EC/PM fractions for 2010+
engines which continue to use DPF technology along with other selective catalyctic converters for
controlling NOx. One of the reasons why we deemed applying the ACES Phase 1 PM speciation
preferable to using the ACES Phase 2 profile (which is tested on MY 2011 diesel engines and
aftertreatment systems) is because the ACES Phase 2 program did not include DPF regeneration
events that can have a large impact on PM composition, as discussed in the MOVES speciation
report.1
2.1.2.2.3	DPF Regeneration Events
The MOVES 3 emission rates include active DPF regeneration effects because the HDIUT data set
includes active regeneration activity, but MOVES does not model active regeneration explicitly. To
do this, we would like to have detailed information on the frequency and emission effect of real
world regeneration events by operating mode and regulatory class. Until we have that kind of data
and see a need for that detail in MOVES, we assume that the emission rates in MOVES for MY
2010+ HD vehicles reasonably capture the average effect of active DPF regeneration events.
To assess the amount of active regeneration activity in the HDIUT data, we examined the ECU
codes. Modern DPFs have catalyzed substrate that allow them to undergo passive regeneration
when the vehicle is operating at high-speeds and/or high-loads such that the exhaust temperature is
sufficient to induce the regeneration. The passive regeneration events are "silent" and happen in the
background without any regeneration code in the ECU data. On the other hand, active regeneration
events happen when the ECU actively raises the temperature in the exhaust so that the soot
captured in the DPF can be combusted. One way to increase the temperature is to inject additional
fuel which gets burned off and raises the temperature. These events can raise the PM2.5
74

-------
concentrations considerably, but may only occur infrequently. We analyzed the "Regen Signal"
column in the quality-assured 1 hz emissions data files for 77 vehicles in the HHD 0.20 NOx FEL
group to estimate the frequency and count of regeneration events. It is our understanding that the
"Regen Signal" flag only accounts for active regen events. There were 11 vehicles with the
Regen Signal set to "Y" and the regen events totaled 60,576 seconds, which is about 18% of the
data from just those 11 vehicles and 3% of the data from all 77 vehicles. Future work could
evaluate whether the active regeneration observed in the HDIUT data is consistent with other
studies and whether the PM2.5 second-by-second measurements accurately capture the elevated
PM2.5 concentrations that occur during active regeneration events.™
2.1.2.3 Tampering and Mal-maintenance
Tampering refers to intentional disabling or modifying the vehicle engine, control systems, and/or
exhaust aftertreatment systems that results in increased emissions. Mal-maintenance refers to lack
or improper maintenance of the engine and aftertreatment, including neglecting to repair broken or
mal-functioning engine and aftertreatment parts, which also increases emissions. MOVES uses the
same methodology to apply tampering and mal-maintenance (T&M) adjustment factors to both the
pre-2010 and 2010 and later PM2.5 emission rates
We followed the same tampering and mal-maintenance methodology and analysis for PM2.5 as we
did for NOx, as described in Appendix B. We account for the emission increases with age by
multiplying the zero-mile emission rates by T&M adjustment factors and scaled aged effects
(Equation 8-3). The MOVES T&M adjustment factors on PM2.5 emissions over the fleet's useful
life are shown in Table 2-25 and derived in Appendix B.8. The value of 89 percent for 2010-2012
model years reflects the projected effect of heavy-duty on-board diagnostic deterrence/early repair
of T&M effects. It is an eleven percent improvement from model years which do not have OBD
(i.e., 2007-2009). The 67 percent value for 2013+ is driven by the assumed full-implementation of
the OBD in 2013 and later trucks, which assumes a 33 percent decrease in T&M emission effects.
m As discussed in the MOVES speciation report1, the PM2 5 composition can change significantly during regeneration
events. Second-by-second PM2 5 measurements made with a photoacoustic or optical method are dependent on the
properties of the PM2 5 composition
75

-------
Table 2-25 Tampering and Mal-maintenance Adjustment Factors for PM2.5 over the Useful Life2
Model Year Group
Increase in PM2 5 Emissions (%)
Pre-1998
85
1998 -2002
74
2003 - 2006
48
2007 - 2009
100
2010-20121
89 (HHD, MHD, LHD45, and Bus)
67 (LHD2b3)
2013+
67
Note:
1 LHD2b3 achieve full OBD adoption in MY 2010. HHD, MHD, LHD45, and Bus are at partial (33%) and full OBD
adoption in MY 2010-2012 and MY 2013, respectively.
2Useful life varies by regulatory class (Table B-4)
2.1.2.3.1 1960-2009 Model Years
The CRC E-55/59 emissions data set used for the pre-2010 emission rates was collected during a
limited calendar year period, yet MOVES requires data from a complete range of model year/age
combinations. For example, for the 1981 through 1983 model year group, the primary dataset
contained data which was in either the 15-to-19 or the 20+ age groups. However, for completeness,
MOVES must have emission rates for these model years for age groups 0-3, 4-5, 6-7, etc. In
populating the emission rates in MOVES, we used the age group that had the most data in each
regulatory class and model year group combination. Then, we used the T&M methodology
discussed in the previous section to model age and model year group combinations.
One criticism of the T&M approach is that it may double count the effect of T&M on the fleet
because the primary emission measurements, and base emission rates, were made on in-use
vehicles that may have had maintenance issues during the testing period. This issue would be most
acute for the 2007 and later model year vehicles where all of the deterioration is subject to
projection. However, for 2007 and later model year vehicles, the base emission rates start at low
levels, and represent vehicles that are considered to be free from the effects of T&M.
Figure 2-31 shows the estimated PM2.5 emission rates from MY 2006 heavy-duty diesel vehicles,
accounting for the effect of tampering and mal-maintenance. The different ages at which each
regulatory class reaches full useful life can be observed as the age when the emissions reach their
maximum value. The age 0-3 emission rates for the 2006 and earlier model year were extrapolated
using the T&M adjustment factors since the majority of these engines were older than three years
when tested in the E-55/59 program.
76

-------
0.012-
0.009-
0.006-
0.003-
Reg Class
41-LHD2b3
¦ 42-LHD45
46-MHD67
- 47-HHD8
48-Urban Bus
0.000
0-3	4-5	6-7	8-9	10-14	15-19	20+
Age (years)
Figure 2-31 Heavy-duty Diesel PM2.5 Emission Rates (g/mile) by Age Group and Regulatory Class for Model
Year 2006 using Nationally Representative Operating Mode Distribution
2.1.2.3.2 2010-2026Model Years
As mentioned in Section 2.1.1.1, the vehicles in HDIUT program are generally well-maintained
and therefore, are used to represent zero-mile emission rates in MOVES. As such, we apply the
T&M adjustment factors shown in Table 2-25 and scaled age effects (Table B-4) to estimate
emission rates for the different ages. As shown in Table B-4, there are different age effects for Tier
2 and Tier 3 LHD2b3 vehicles. We used Equation 2-27 to estimate a weighted average of the PM2 5
emission rates during the Tier 3 phase-in (Model year 2017-2022) as discussed in 2.1.1.5.5.
Note that for regulatory class LHD45, MHD, HHD and Urban Bus, the PM2.5 emission rates in the
MOVES database for all age groups for MY 2013+ have lower T&M age adjustments than their
counterpart rates for MY 2010-2012 due to the HD OBD phase-in (see Section 2.1.2.3 and
Appendix B).
Figure 2-32 shows the impact of tampering and mal-maintenance on PM2.5 emission rates by
vehicle age for MY 2015 heavy-duty vehicles. Again, the rate at which emissions increase toward
their maximum varies by regulatory class.
77

-------
0.012
0.009
E
"S>
re
en
c
o
E
HI
in
c\i
0.006
£ 0.003
0.000
Reg Class
41-LHD2b3
¦ 42-LHD45
46-MHD67
- 47-HHD8
48-Urban Bus
10-14
15-19
8-9
Age (years)
Figure 2-32 Heavy-duty Diesel PM2.5 Emission Rates (g/mile) by Age Group and Regulatory Class for Model
Year 2015 using Nationally Representative Operating Mode Distribution
2.1.2.3.3
2027-2060 Model Years
The T&M approach for model years 2027 through 2060 is similar to the approach for 2010 through
2026, but since the HD2027 standards set longer warranty period and useful life requirements (as
shown in Table B-2) for diesel heavy-duty vehicles starting with the 2027 model year, we adjusted
the T&M effects and the resulting PM2.5 emission rates to account for those changes.
2.1.2.4 Model Year Trends
Figure 2-33 displays the PM2.5 rates by model year and regulatory class for 0-3 age group estimated
in grams per mile (g/mile) using nationally representative operating mode distributions and average
speeds. MOVES models a very large decrease in PM2.5 emission rates starting in model year 2007
(decrease on order of-10 to 40 times), when all regulatory classes are assumed to have
implemented diesel particulate filters, with the exception of gliders (Section 2.5). As discussed in
Section 2.1.1.8, some of the variation between regulatory classes is also due to differences in the
application of T&M adjustment factors and differences in the operating mode distributions and
average speeds.
Figure 2-34 provides resolution to the model year changes in PM2.5 emission rates for the 2007 and
later model years. Further reductions in PM2.5 emissions are observed for each regulatory class
between the 2007-2009 and the 2014 and later emission rates. The higher rates for MY 2010 and
later MHD vehicles stem directly from the HDIUT data as shown in Figure 2-29 and discussed in
Section 2.1.2.2.1. The minor variation in the gram per mile emission rates within the 2010-2013
78

-------
and 2014+ model year groups by model year and regulatory class are due to differences in
operating mode distributions. Finally, there is a small change in PM2.5 rates beginning in MY2027
due to warranty period provisions of the HD2027 rule.
1980	2000	2020	2040
Model Year
Figure 2-33. Base running emission rates for PIVhsfrom age 0-3 diesel heavy-duty vehicles averaged over a
nationally representative operating mode distribution.
Reg Class
41-LHD2b3
42-LHD45
46-MHD67
47-HHD8
48-Urban	Bus
¦*- 49-Gliders
79

-------
2010	2020	2030	2040
Model Year
0.01
Reg Class
¦*- 41-LHD2b3
42-LHD45
46-MHD67
47-HHD8
48-Urban	Bus
Figure 2-34 Base running emission rates for PIVI2.5from age 0-3 diesel heavy-duty vehicles for MY 2007-2040
averaged over a nationally representative operating mode distribution.
Figure 2-35 shows the PM2.5 emission rates separated into elemental carbon (EC) and non-
elemental carbon (nonEC) fractions for age 0-3 HHD diesel vehicles using nationally
representative operating mode distributions and average speeds. The EC fraction stays constant
until model year 2007, when it is reduced to less than -10 percent due the implementation of diesel
particulate filters.
80

-------
EC
NonEC
1980
1990
2000
2010
2020
2030
Model Year
Figure 2-35 Heavy Heavy-duty (HHD) Diesel PM2.5 Emission Rates by Elemental Carbon (EC) and Non-
Elemental Carbon (nonEC) Fraction for the 0-3 Age Group by Model Year using Nationally Representative
Operating Mode Distributions
2.1.3 Total Hydrocarbons (THC) and Carbon Monoxide (CO)
While diesel engine emissions of THC and CO are important, they are not the largest contributors
to mobile source THC and CO emission inventories. Regulations of non-methane hydrocarbons
(NMHC), combined with the common use of diesel oxidation catalysts have yielded reductions in
both THC and CO emissions from later model year heavy-duty diesel engines. As a result, data
collection efforts typically do not focus on THC or CO from heavy-duty engines, and less data is
available. As discussed in Section 1.1, this report discusses the derivation of total hydrocarbons
(THC), from which MOVES estimates other hydrocarbons and organic gaseous pollutants.
2.1.3.1 1960-2009 Model Years
We used emissions data combined with emissions standards to develop appropriate model year
groups. Since standards did not change frequently in the past for either NMHC or CO, we created
fewer model year groups than we did for NOx and PM. The MOVES THC and CO model year
groups are:
•	1960-1989
•	1990-2006
•	2007-2009
2.1.3.1.1 Data Sources
The heavy-duty diesel THC and CO emission rate development followed a methodology that
resembles the light-duty methodology9, where emission rates were calculated from 1-hz data
81

-------
produced from chassis dynamometer testing. Data sources were all heavy-duty chassis test
programs:
1.	CRC E-55/5941: As mentioned earlier, this program represents the largest volume of heavy-
duty emissions data collected from chassis dynamometer tests. All tests were used, not just
those using the TEOM. Overall, 75 trucks were tested on a variety of drive cycles. Model
years ranged from 1969 to 2005, with testing conducted by West Virginia University from
2001 to 2005.
2.	Northern Front Range Air Quality Study (NFRAQS)53: This study was performed by
the Colorado Institute for Fuels and High-Altitude Engine Research in 1997. Twenty-one
HD diesel vehicles from model years 1981 to 1995 selected to be representative of the in-
use fleet in the Northern Front Range of Colorado were tested over three different transient
drive cycles.
3.	New York Department of Environmental Conservation (NYSDEC)54: NYSDEC
sponsored this study to investigate the nature and extent of heavy-duty diesel vehicle
emissions in the New York Metropolitan Area. West Virginia University tested 25 heavy
heavy-duty and 12 medium heavy-duty diesel trucks under transient and steady-state drive
cycles.
4.	West Virginia University: Additional historical data collected on chassis dynamometers
by WVU is available in the EPA Mobile Source Observation Database.
The pre-2010 onroad data used for the NOx analysis was not used since THC and CO were not
collected in the MEMS program, and the ROVER program used the less accurate non-dispersive
infrared (NDIR) technology instead of flame-ionization detection (FID) to measure HC. To keep
THC and CO definitions and data sources consistent, we only used chassis test programs which
measured THC using a FID exclusively for the analysis. Time-series alignment was performed
using a method similar to that used for light-duty chassis test data.
Table 2-26 Numbers of Vehicles by Model Year Group, Regulatory Class, and Age Group
Model year
group
Regulatory class
Age group
0-3
4-5
6-7
8-9
10-14
15-19
20+

HHD
58
19
16
9
16
6
7

MHD
9
6
5
4
12
15
6
1960-2002
LHD45
2


1




LHD2b3
6







Bus
26


1
3


2003-2006
HHD
6






2007-2009
HHD, MHD, LHD45,
LH2b3, Bus
No vehicles for this model year group. Rates for
this model year group are based on MY 2003-
2006 with 80 percent reduction.
2.1.3.1.2 1960-2006Model Years
Similar to the analysis done for PM2.5, for each second of operation on the chassis dynamometer,
the instantaneous scaled tractive power (STPt) was calculated using Equation 1-6 and the second
82

-------
was subsequently classified to one of the 23 operating modes defined in Table 1-4. We used the
same track-load coefficients, A, B, and C pertaining to heavy-duty vehicles that were used in the
PM2.5 analysis.
Using the methods introduced in the NOx analysis, we averaged emissions by vehicle and operating
mode. We then averaged across all vehicles by model year group, age group, and operating mode.
In populating the emission rates in MOVES, we used the age group that had the most data in each
regulatory class and model year group combination. These age groups are shown in Table 2-27. We
then used the T&M effects discussed in Section 2.1.3.3 to extrapolate the emission rates for each
age group. For missing operating modes, we extrapolated using STP as was discussed for NOx in
Section 2.1.1.4.2. For the 1960-2002 group, data for the HHD and Urban Bus regulatory classes
were combined because they have the same CO and NMHC emission standards, although they
have separate age effects as discussed in Section 2.1.3.3.
Regulatory class
Model year group
Age group
HHD/Urban Bus
1960-2002
0-3
MHD
1960-2002
15-19
LHD2b3
1960-2002
0-3
HHD
2003-2006
0-3
Based on the Data
With limited data on LHD45 vehicles, we applied the LHD2b3 emissions data to all LHD vehicles.
We also applied the LHD emission rates from 1960-2002 to the LHD 2003-2006 model year group.
For 2003-2006 MHD and Urban Bus regulatory classes emission rates, we applied the HHD 2003-
2006 emission rates.
Figure 2-36 and Figure 2-37 show the rates for MHD and HHD for MY 2002 for THC and CO,
respectively, based on the methods described above. The THC and CO mean emission rates
increase with STP, though there is much higher uncertainty than for the NOx rates (Figure 2-1).
This pattern could be due to the smaller data set or may reflect a less direct correlation of THC and
CO to STP as is observed for the 2010 and later model year rates.
83

-------
100
01
+¦»
(S
t—
U
90
80
70
60
50
a 40
ot
30
0
~ MHD
¦ HHD/Bus
20


¦

<
<
~
~
i
10
'
i i
i f
~i	1	1	1	1	1	r
~i	1	1	1	1	1	1	r
~i	1	1	1	1	1
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
Operating mode
Figure 2-36 THC Emission Rates [g/hr] by Operating Mode for Model Year 2002 and Age Group 0-3. Error
Bars Represent the 95 Percent Confidence Interval of the Mean
600 n
500
T 400 -
.c
Bfl
01
a
O
u
c
a
o»
300 ^
2 200
100
~ MHD
¦ HHD/Bus
i i
« • 5 1
Hi
n	1	1	1	1	1	r
fhi
ii
	1	1	1	1	1	1	1	r
ijli
-1—i—i—i—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
Operating mode
Figure 2-37 CO Emission Rates [g/hr] by Operating Mode for Model Year 2002 and Age Group 0-3. Error Bars
Represent the 95 Percent Confidence Interval of the Mean
84

-------
2.1.3.1.3
2007-2009Model Years
With the increased use of diesel oxidation catalysts (DOCs) in conjunction with DPFs, we assumed
an 80 percent reduction in zero-mile emission rates for both THC and CO for 2007-2009 model
years. The derivation of the T&M effects for 2007-2009 model years are presented in Table 2-28
and discussed in Appendix Section B.9. As shown in Figure 2-48, the CO emission rates developed
using this assumption are significantly lower than the model year 2010 and later emission rates that
were developed based on the HDIUT data, and should be re-evaluated in future versions of
MOVES.
2.1.3.2 2010-2060 Model Years
We used the MY 2010+ HDIUT data set, using the same vehicles as used for NOx and described in
Section 2.1.1.1 and Table 2-2. The HDIUT dataset includes vehicles in the HHD, MHD, LHD45,
LHD2b3 and Urban Bus regulatory classes. The HDIUT emission measurements are made using
instruments that conform to the requirements described in 40 CFR Part 1065, which require the use
of a flame ionization detector (FID) for measuring total hydrocarbons (THC)55 and a non-
dispersive infrared (NDIR) analyzer for carbon monoxide (CO)56
The THC and CO emission rates have more uncertainty than the NOx emission rates, which
suggests a less direct correlation of THC and CO to STP. Nevertheless, we followed the analysis
methodology used for MY 2010+ NOx rates as described in Sections 1.6 (calculation of STP and
assignment of operating modes), 2.1.1.5 (calculation of mean emission rates), 2.1.1.5.1 (NOx FEL
groups) and Appendix G (selection offscale). Figure 2-38 and Figure 2-39 display the HHD THC
and CO emission rates estimated from the HDIUT data by the NOx FEL Groups used to develop
the MY 2010-2013 emission rates. Comparisons of the THC and CO emission rates by NOx FEL
Groups for the LHD and MHD regulatory classes are provided in Appendix H. These comparisons
show that there are significant differences among the emission rates in different NOx FEL groups
for THC emissions. The THC emission rates in the 0.2 and 0.35 NOx FEL group are lower than the
THC emission rates from the vehicles in the 0.5 NOx FEL groups for each regulatory class, with
the differences being the most significant for MHD (Appendix H.1.2) and especially HHD (Figure
2-38). For CO, there is not a consistent trend among the different NOx FEL groups and regulatory
classes. Regardless, we have analyzed the CO emission rates using the NOx FEL groups for
consistency.
85

-------
0.020
0.015 -
no
™ 0.010
u
X
0.005 -
0.000
ta HHD FEL 0.20, MY 2010-2013,!N=78
I
Ic HHD FEL 0.35, MY2010-2015,!N=31
I	I
¦ HHD FEL 0.50, MY2010-2015,! N=35
i .i
0 1, 11 12 13 14 15 16
i
i
iii
21 22 23 24 25 27 28 29 30
MOVES OpMode
i I
J
33 35 37 38 39 40
Figure 2-38 Average HHD THC Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013 and
the 0.35 and 0.5 NOx FEL for MY 2010-2015. Error Bars are 95% Confidence Intervals of the Mean
lU
Jl

I HHD FEL 0.20, MY 2010-2013, N=78
i HHD FEL 0.35, MY 2010-2015, N=31
I HHD FEL 0.50, MY 2010-2015, N=35
I
1
i
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 2-39 Average HHD CO Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013 and
the 0.35 and 0.5 NOx FEL for MY 2010-2015. Error Bars are 95% Confidence Intervals of the Mean
Figure 2-40 and Figure 2-41 display the production-weighted average emission rates for THC and
CO emissions for model year 2013 HHD trucks. Production-weighted averages are calculated for
each model year between model year 2010 and 2018 using the production volumes displayed in
Figure 2-12.
86

-------
0.0050
0.0040
«T 0.0030
.25
u
IE 0.0020
0.0010
0.0000

¦ HDIUT, MY 2013













. , ,11
1

i I 1
1 II
i i

1
1
u i i 1
II
0 1
11 12 13 14 15 16
21 22 23 24 25 27 28 29 30
MOVES OpMode
33 35 37 38 39 40
Figure 2-40 THC Emissions by Operating Mode from HHD Trucks for Model Year 2013. Error Bars represent
the 95 percent confidence interval of the Mean
0.12
0.09 -
hjQ
™ 0.06
O
U
0.03 -
0.00
J	i.
¦ HDIUT, MY 2013
til
i
0 1 , 11 12 13 14 15 16 I 21 22 23 24 25 27 28 29 30 i 33 35 37 38 39 40
MOVES OpMode
Figure 2-41 CO Emissions by Operating Mode from HHD Trucks for Model Year 2013. Error Bars represent
the 95 percent confidence interval of the Mean
Figure 2-42 and Figure 2-43 display the comparison of the MY 2010-2013 and MY 2014-2015
groups within the HHD 0.2 NOx FEE Groups. In general, the newer vehicles (MY 2014-2015) have
lower TFIC and CO emission rates than the corresponding MY 2010-2013 emission rates. Similar
model year trends are observed for LFID TFIC emissions, and MHF) THC and CO emissions in
Appendix H. As discussed in Section 2.1.1.5.2 regarding NOx emissions, we attribute the model
year differences within the 0.2 NOxFEL Group to improved emission control hardware and engine
and aftertreatment operation.
87

-------
0.0040
0.0035
—	0.0030
i/i
no
—	0.0025
U
I- 0.0020
0.0015
0.0010
0.0005
0.0000
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 2-42 THC emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the HHD 0.20 NOx FEL
Group
0.14
0.12
0.10
1/T
"SB 0-08
O
0.06
0.04
0.02
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
MOVES OpMode
Figure 2-43 CO emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the HHD 0.20 NOx FEL
Group
LHD2b3 and LHD45 emission rates are based on the vehicles with "LHD" service class in the
HDIUT data set, which only contains engine-certified LHD45 vehicles. Urban Bus emission rates
are based on HHD vehicles in the HDIUT data set. MHD and HHD emission rates are based on the
emission rates from those vehicle classes. The
HDIUT data set predominantly contains vehicles in the 0-3 age group with only a handful vehicles
in the 4-5 age group. Since the HDIUT data is measured and submitted by the manufacturer and the
test vehicles are required to be free of any tampering or mal-maintenance, we can safely assume
that they represent zero-mile vehicles for the purpose of assigning base rates and applying the
tampering and mal-maintenance effects.
A comparison of HDIUT-based THC and CO emission rates for MY 2010+ heavy-duty vehicles by
regulatory class are shown in Figure 2-47 and Figure 2-48, respectively. The THC rates, generally
¦ MY 2010-2013

¦ MY 2014-2015












|




ii ii ii li ii ii
1 ii ii ii ii ii Ii
I

I 1
T I


i. .. ii Ii 1

¦	MY 2010-2013
¦	MY 2014-2015
88

-------
low for diesel vehicles, are comparable to MY 2007-2009 rates for both MHD and HHD. However,
for CO, the HHD rates for MY 2010+ are significantly higher compared to MY 2007-2009, but are
comparable to the pre-2007 data which are based on emission measurements. The variation in the
2010-2018 rates reflects the model year variation in the production volume by NOx FEL group, and
use of the different 0.2 FEL NOx model year group between 2010-2013 and 2014 and later.
In the 2017 review of a draft version of this report, we received a comment that single-cell-NDIR-
based CO measurements suffer from severe drift that is not corrected by zero and span checks
because the calibration gases are dry, while vehicle tailpipe exhaust gases are not dry. Based on the
HDIUT data, it is not possible for us to determine if MY 2010+ CO emission rates are affected by
the alleged drift in the CO measurements. We looked at the CO emissions for each of 93 vehicles
in the HHD 0.20 FEL group (from the 2010-2016 selection years) and confirmed the high average
CO rate is not due to a few outliers. Further, the CO emission rate for the MHD and LHD vehicles
is significantly lower (see Figure 2-48). Based on the available data and trends, we are unable to
confirm whether or not the high CO emissions for the HHD vehicles is real or an artifact of CO
sensor drift. In Section 2.1.5, we demonstrated that the fleet-average heavy-duty CO emission rate
estimates from MOVES compare well with measurements from heavy-duty exhaust plume capture
and tunnel measurement campaigns conducted in 2015 and 2017, which increased our confidence
that the CO emission rates measured from HDIUT are reasonable. Thus, we decided to accept the
reported HDIU CO emission rates as valid.
As discussed in 2.1.1.5.5, we did not reduce the LHD2b3 zero-mile THC and CO emission rates
due to the implementation of the Tier 3 standard. For LHD2b3 2010 and later vehicles, the
MOVES emissions rates are based on LHD45 vehicles measured in the HDIU program, as
described above. The surrogate LHD45 emission rates, for THC, CO, and PM2.5emissions, imply
that current levels on the FTP cycle are substantially below the Tier 3 standards. For example,
when MOVES rates are used to simulate FTP cycle for NMHC, the result is a rate of approximately
0.05 grams per mile, while the simulated FTP estimate for CO is less than 1.0 gram/mile. However,
we did account for the lengthened useful life standard required by the Tier 3 standard in the
Tampering & Mal-maintenance standards as discussed in the next section.
2.1.3.3 Tampering and Mal-maintenance
For all model years, we applied tampering and mal-maintenance effects to adjust emissions from
the measured age to all age groups, lowering emissions for younger ages and raising them for older
ages, using the methodology described in Appendix B. We applied the tampering and mal-
maintenance effects shown below in Table 2-28 to CO and THC.
For MY2027+ LHD45, MHD, HHD vehicles, we adjusted the T&M effects in estimating the
emission rates for CO and THC to account for the longer warranty period and useful life
requirements in HD2027 standards (as shown in Table B-2).
89

-------
Table 2-28 Tampering and Mal-maintenance Effects for THC and CO over the Useful Life
Model years
Increase in THC and CO Emissions (%)
1994-2003
300
2003 - 2006
150
2007 - 2009
150
2010-20121
29 (HHD, MHD, LHD45, and Bus)
22 (LHD2b3)
2013-2026
22
2027+
22
Note:
1 LHD2b3 achieve full OBD adoption in MY 2010. HHD, MHD, LHD45, and Bus are at partial (33%) and full OBD
adoption in MY 2010-2012 and MY 2013, respectively.
While LHD2b3, LHD45 and MHD vehicles share the same pre-2010 MY fully deteriorated
emission rates for THC and CO, they deteriorate differently as they age. Table B-4 estimates the
degree of T&M that occurs by age by using the warranty and full useful life requirements for each
heavy-duty regulatory class with the average mileage accumulation rates. We multiplied these
increases by the T&M age-based adjustment factors shown in Table B-4 and applied the result to
the zero-mile (or age 0) emissions rate to estimate the emissions rate by age group using Equation
8-3. As shown in Table B-4, there are different age effects for Tier 2 and Tier 3 LHD2b3 vehicles.
We used Equation 2-27 to estimate a weighted average of the THC and CO emission rates during
the Tier 3 phase-in (Model year 2017-2022) as discussed in Section 2.1.1.5.5.
Figure 2-44 and Figure 2-45 show THC and CO emission rates by age group for MY 2015. Due to
our projections of T&M effects, there are large increases as a function of age. Additional data
collection would be valuable to determine if real-world deterioration effects are consistent with
those in MOVES.
90

-------
0-3	4-5	6-7	8-9	10-14	15-19	20+
Age (years)
Figure 2-44 Heavy-duty Diesel THC Running Emission Rates (g/mile) by Age for Model Year 2015 by
Regulatory Class Estimated using Nationally Representative Operating Mode Distribution
Reg Class
41-LHD2b3
-+¦ 42-LHD45
46-MHD67
47-HHD8
48-Urban	Bus
0-3	4-5	6-7	8-9	10-14	15-19	20+
Age (years)
Figure 2-45 Heavy-duty Diesel CO Running Emission Rates (g/mile) by Age for Model Year 2015 by Regulatory
Class Estimated using Nationally Representative Operating Mode Distribution
91

-------
2.1.3.4 Model Year Trends
Figure 2-46 through Figure 2-48 display the THC and CO emission rates by model year and
regulatory class for age group 0-3 estimated in grams per mile (g/mile) using nationally
representative operating mode distributions and average speeds. As discussed in Section 2.1.1.8,
some of the minor variation in the gram per mile emission rates within the model year groups and
between regulatory classes are due to differences in operating mode distributions. Differences in
the emission rates for age group 0-3 between regulatory classes are also due to different application
of the T&M adjustment factors (Section 2.1.3.3). For example, as discussed in Section 2.1.3.1.3,
the zero-mile MY 2003-2006 emission rates by operating mode are equivalent for HHD, MHD, and
Urban Bus, but the T&M adjustment factors are applied differently for each regulatory class, and
the operating mode distributions are difference, resulting in the differing gram per mile emission
rates observed for 2003-2006 in Figure 2-46 and Figure 2-48.
The MY 2007-2009 emission rates reflect the use of diesel oxidation catalysts and are derived by
reducing the CO and THC emissions in MY 2003-2006 by 80 percent and applying the model-year
and regulatory class specific T&M adjustment factors. For MY 2010-2018, the significant
variation in the emission rates by model year are due to the model year specific production volumes
of the NOx FEL Group and the model year split of the 2010-2013 and 2014 and later 0.2 NOx FEL
group (Section 2.1.1.5.2 and Section 2.1.3.2). For example, the spike in THC emissions observed in
the 2010 model year HHD vehicles (Figure 2-47) is explained by the high THC emissions of the
0.5 NOx FEL group (Figure 2-38), and the high production volumes of the 0.5 NOx FEL engines in
MY 2010 (-50% of the total HHD, see Figure 2-12).
Finally, there are small reductions in emission rates attributable to the longer warranty period and
useful life requirements in the HD2027 rule.
92

-------
£
B>
re
ct
o
X
Reg Class
-+¦ 41-LHD2b3
42-LHD45
46-MHD67
47-HHD8
48-Urban	Bus
49-Gliders
1980	2000	2020	2040
Model Year
Figure 2-46 Base running emission rates for THC from age 0-3 diesel heavy-duty vehicles averaged over a
nationally representative operating mode distribution.
2010	2020	2030	2040
Model Year
Figure 2-47 Base running emission rates for THC from age 0-3 diesel heavy-duty vehicles for MY 2007-2040
averaged over a nationally representative operating mode distribution.
— 0.10
v
|
Reg Class
41-LHD2b3
42-LHD45
46-MHD67
47-HHD8
48-Urban	Bus
93

-------
1980
2000
2020
2040
Model Year
Figure 2-48 Base running emission rates for CO from age 0-3 diesel heavy-duty vehicles averaged over a
nationally representative operating mode distribution.
2.1.4 Energy
2.1.4.1 1960-2009 Model Years
2.1.4.1.1 LHD
Reg Class
-+¦ 41-LHD2b3
42-LHD45
46-MHD67
47-HHD8
48-Urban	Bus
49-Gliders
In MOVES4, the energy rates for LHD (LHD2b3 and LHD45) for pre-2010 MY diesel vehicles are
unchanged from MOVES2010a. In MOVES2010, the energy rates for LHD2b3 regulatory class,
along with the light-duty regulatory classes (regClassIDs 20 and 30), varied by fueltype, model
year group, engine technology, and "size weight fraction" as discussed in the MOVES2010a
energy updates report.57 The energy rates in MOVES2010a were simplified to be a single set of
energy rates for each regulatory class, fuel type and model year combination by weighting across
engine size, engine technology, and vehicle weight according to the default population in the
MOVES2010 sample vehicle population table. The resulting CO2 (g/mile) emission rates and fuel
economy values (miles per gallon) calculated from the energy rates using nationally representative
operating mode distributions and average speeds are shown in Figure 2-52 and Figure 2-53.
Because this approach uses highly detailed data, coupled with information on the vehicle fleet that
varies for each model year, model year variability was introduced into the energy rates used in
MOVES.
2.1.4.1.2 MHD, Urban Bus, and HHD
The data used to develop NOx rates was used to develop running-exhaust energy rates for the
MHD, Urban Bus, and HHD vehicles. The energy rates were based on the same data (Section
94

-------
2.1.1.1), STP structure and calculation steps as in the NOx analysis (Sections 2.1.1.3 and 2.1.1.4);
however, unlike NOx, we did not classify the energy rates by model year, regulatory class, or by
age, because neither variable had a significant impact on energy rates or CO2.
In MOVES, CO2 emissions were used as the basis for calculating energy rates. To calculate energy
rates (kJ/hour) from CO2 emissions (Equation 2-32), we used a heating value (HV) of 138,451
kJ/gallon and CO2 fuel-specific emission factor (fcoj) of 10,180 g/gallon58 for conventional diesel
fuel.
HV
^energy ^C02 T	Equation 2-32
JCO2
The energy rates for the MHD, Urban Bus, and HHD vehicle classes are shown in Figure 2-49.
Compared to other emissions, the uncertainties in the energy rates are smaller, in part because there
is no classification by age, model year, or regulatory class. Thus, the number of vehicles used to
determine each rate is larger, providing for a greater certainty of the average mean energy rate.
Operating mode-based energy consumption rates are the same across MHD, Urban Bus, and HHD
regulatory classes. However, the distribution of time spent in the operating mode varies between
these regulatory classes based on differences in their activity and tractive power demand. Thus, the
CO2 (g/mile) emission rates and fuel economy values (miles per gallon) calculated from the energy
rates using nationally representative operating mode distributions differ by regulatory class as
shown in Figure 2-52 and Figure 2-53.
6
5 -
~ • *
~i—1—1—1—1—1—1—1—1—1—1—1—1—1—1—1—1—1—1—1—1—1—1
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
Operating mode
Figure 2-49 Diesel running exhaust energy rates for MHD, HHD, and Urban Buses for 1960-2009 model years.
Error bars represent the 95 percent confidence interval of the mean
95

-------
2.1.4.2
2010-2013 Model Years
The MY 2010+ HDIUT dataset described in Section 2.1.1.1 and Table 2-2 included CO2 emissions
data, which was used to update the energy rates. The energy rates are derived using the CO2 rates
and the conventional diesel specific values for carbon content (0.0202 g/KJ) and a ITV of of
138,451 kJ/gallon, yielding a CO2 fuel-specific emission factor (fan) of 10,255 g/gallon. MOVES
uses these same values to calculate CO2 emissions from the energy rates of vehicles using
conventional diesel fuel - this methodology is described in the MOVES GHG and Energy
Consumption report3.
The 2010-2013 model year energy rates were calculated using the NOx FEL production volume and
model year group splits used for estimating the MY 2010+ NOx rates as described in Section
2.1.1.5. The energy rates for the 0.2 NOx FEL group are based only on MY 2010-2013 vehicles.
The energy rates for the 0.35 and 0.5 NOx FEL group were developed using vehicles sampled
between 2010-2016 model years. Figure 2-50 shows the mean HMD CO2 emission rates for the
NOx FEL Groups used to estimate the MY 2010-2013 emission rates. As shown, CO2 emission
rates are a strong function of STP operating mode, and there is significantly less variability in the
CO2 emission rates between the sampled vehicles compared to other pollutants, as evidenced by the
small confidence intervals. Even though there is little difference between the CO2 emission rates
among the different FEL groups, we used the FEL production volume and model years splits to
estimate the CO2 emissions, to be consistent with our analysis of the THC, CO, and NOx emission
rates.
m
32
IN
o
u
100
80 -
60
40 -
20 -
p HHD FEL 0.20, MY 2010-2013,
)¦ HHD FEL 0.35, MY 2010-2015,
'¦ HHD FEL 0.50, MY 2010-2015,

N=78
N=31
N=35
11 12 13 14 15 16 : 21 22 23 24 25 27 28 29 30 , 33 35 37 38 39 40
MOVES OpMode
Figure 2-50 Average HHD CO2 Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013 and
the 0.35 and 0.5 NO* FEL for MY 2010-2015. Error Bars are 95% Confidence Interv als of the Mean.
As shown in Table 2-29, the majority of vehicles within the NOx FEL groups of 0.35 and 0.5
include MY 2010-2013 vehicles, for both FflTD and MILD vehicles. As discussed in the next
subsection, we expect improved energy efficiencies in MY 2014 and later vehicles due to the
phase-in of the Phase 1 Fleavy-duty Greenhouse Gas Emission Standards. Due to the small sample
96

-------
of vehicles in the 0.35 and 0.5 NOx FEL groups, we assume that the MY 2010-2016 energy rates
are representative of both the MY 2010-2013 and the 2014-2016 model year groups.
Table 2-29 HDIU Vehicles with Valid CO2 Measurements By Regulatory Class, Model Year Group and NOx

NOx FEL Group
Total
Valid
Tested
Vehicles
Reg
Class
Model Year
Group
0.2
0.35
0.5
LHD
2010-2013
52
0
10
62
64
2014-2016
27
0
5
32
32
MHD
2010-2013
21
23
9
53
55
2014-2015
19
0
0
19
19
HHD
2010-2013
78
26
35
139
139
2014-2015
44
5
0
49
55
Using this method, the energy rates for each model year are unique based on NOx FEL based
production volume weighting, as can be observed in the model year variability among the MY
2010-2013 CO2 (g/mile) emission rates and fuel economy values (miles per gallon) shown in
Figure 2-52 and Figure 2-53.
2.1.4.3 2014-2060 Model Years
2.1.4.3.1 LHD45, MHD, Urban Bus, and HHD Energy Rates
In developing the MY 2014-2060 running energy rates for LHD45, MHD, Urban Bus, and HHD,
we also used the NOx FEL groups, model year groups, and production volume weights as discussed
in the previous section (Section 2.1.4.2). The MY 2014-2018 running energy rates were calculated
as a weighted average using the MY 2014-2015/2016 rates from the 0.2 NOx FEL group, and the
MY 2010-2016 rates from the 0.35 and 0.5 NOx FEL groups. Although the 0.35 and 0.5 NOx FEL
groups contain measurements from MY 2010-2013 vehicles (Table 2-29), applying the MY-
specific production volume weighting of the FEL groups means that the MY 2014-2018 emission
rates are primarily or entirely based on the data from the MY 2014-2015/2016 vehicles in the 0.2
NOx FEL group. The MY 2014-2018 running energy rates developed in this step are considered the
"baseline" - the reductions in energy rates expected from the Medium- and Heavy-Duty
Greenhouse Gas (GHG) Phase 1 Rule59 were applied to the "baseline" energy rates as described
below.
MOVES accounts for the improved fuel efficiency achieved by the HD GHG Rulemakings in two
ways. First, the running, start, and extended idle rates for total energy consumption are reduced to
be consistent with the HD GHG rules. Second, the truck weights and road-load coefficients are
updated to reflect the lower vehicle curb weights through lightweighting of materials, lower
resistance tires, and improved aerodynamics of the vehicle chassis. Vehicle weights and road-load
coefficients are discussed in the Population and Activity Report.6
The HD GHG Phase 1 rule59 was implemented starting with 2014 model year and increased in
stringency through model year 2018. The reductions in start and running energy rates reflect the
improvements expected from improved energy efficiency in the powertrain. The estimated
97

-------
reductions for heavy-duty diesel energy rates from the HD GHG Phase 1 rule are shown in Table
2-30.
However, the MY 2014-2016 running energy rates were not adjusted for the HD GHG Phase 1
rulemaking because the impact of Phase 1 is assumed to be included in the measurements from the
MY 2014-2015/2016 vehicles sampled from the HDIUT program. Instead, we renormalized the
Phase 1 GHG reductions using the MY 2014-2016 as the baseline using Equation 2-33.
Renormalized Phase 1 reductions in yeart
1 — (reductions in yeart)	Equation
1 — (Average reductions in 2014 thru 2016)	2-33
For example, the renormalized reductions for LHD and MHD in 2017-2020 are calculated as:
Renormalized Phase 1 reductions for LHD and MHD in 2017 thru 2020
1 — (9%) 91% Equation
— 1			- — 1	— 1 _ Qfioz. — 40A	1
1 — (5%) ~~ 95%	yb/o-4/o	2.34
We applied the renormalized reductions to estimate the MY 2017 and later running energy rates, as
shown in Table 2-30. As discussed in Section 2.2.4, because the start energy rates were not updated
with more data from model year 2014-2016 vehicles, the reduction in energy consumption from
starts due to HD GHG Phase 1 rule was modeled by directly using the reductions estimated from
the rule.
Table 2-30 Estimated Reductions in Diesel Engine Energy Consumption Rates from the HD GHG Phase 1
				Program60		
Regulatory
Class
Fuel
Model
Years
Estimated Reduction from the
MY 2013 Baseline
(applied to starts)
Renormalized Reductions to
MY 2014-2016 Energy
Rates
(applied to running)
HHD and
Urban Bus
Diesel
2014-2016
3%
-
2017-2020
6%
3%
LHD and
MHD
Diesel
2014-2016
5%
-
2017-2020
9%
4%
MOVES3 incorporated the Medium- and Heavy-Duty GHG Phase 2 rule.61 We updated MOVES4
to reflect Phase 2 as implemented. The Phase 2 program begins in 2021 and phases in through
model year 2027. These Phase 2 standards continue indefinitely after model year 2027. The
programs break the diverse truck sectors into three distinct categories, including:
•	Line haul tractors and trailers (combination trucks source types in MOVES)
•	Heavy-duty pickups and vans (passenger truck and light-commercial trucks)
•	Vocational trucks (buses, refuse trucks, motorhomes, single-unit trucks)
98

-------
The Phase 2 Rule set separate standards for engines and vehicles and ensured improvements in
both. It also set separate standards for fuel consumption, CO2, N2O, CH4 and HFCs.11
Because the Phase 2 rulemaking set different standards for vocational vehicles and tractor-trailers
and because single-unit vocational vehicles and tractor-trailers are mapped to the same regulatory
classes (MHD and HHD) under the default MOVES framework for emission rates, we modeled
changes in running energy rates due to Phase 2 using the EmissionRateAdjustment table. The
EmissionRateAdjustment table includes the following data fields, many of which are shared with
the EmissionRate table:
1)	polProcessID (primary key)
2)	sourceTypelD (primary key)
3)	regClassID (primary key)
4)	fuelTypelD (primary key)
5)	beginModelYearlD (primary key)
6)	endModelYearlD (primary key)
7)	emissionRateAdjustment
8)	dataSourcelD
Table 2-31 summarizes the energy rate reductions stored in the EmissionRateAdjustment table
which are applied to the running rates in MOVES4 for MY 2021 and later heavy-duty diesel
vehicles.
n HFCs are not modeled in MOVES, and the N20 and CH4 standards are not considered technology forcing on
emissions.
99

-------
Table 2-31 Estimated Reductions in Diesel and CNG Engine Energy Consumption Rates due to the HD GHG
		Phase 2 Program61		
Vehicle Source Type
(Source Type ID)
Fuel
Model years
Reduction from MY 2020
Energy Rates
Long-haul Combination
Truck
(62)
Diesel & CNG
2021-2023
6.9%
2024-2026
11.4%
2027+
15.3%
Short-haul Combination
Truck
(61)
Diesel & CNG
2021-2023
6.8%
2024-2026
11.3%
2027+
14.4%
Other Bus, School Bus,
Refuse Truck, Single-Unit
Short-Haul, Single-Unit
Long-Haul, Motorhomes
(41,43,51,52,53,54)
Diesel & CNG
2021-2023
7.8%
2024-2026
12.3%
2027+
16.0%
Transit Bus
(42)
Diesel & CNG
2021-2023
7.0%
2024-2026
11.8%
2027+
14.4%
Thus, for LHD45, MHD, HHD and Urban Bus, the running energy rates for MY 2021 and later are
estimated with a chain of calculations starting with the HDIUT-based estimates by operating mode
and regulatory class, then reduced by applying the HD GHG Phase 1 reduction in Table 2-30 and
further reduced by applying the HDGHG Phase 2 reductions listed in Table 2-31. The reductions
shown in Table 2-31 reflect a combination of improvements to the engine and other systems, but
exclude improvements to aerodynamics and tire rolling resistances. The projected improvements
due to aerodynamics and tire rolling resistance are reflected in new road load coefficients, as
described in the Population and Activity Report.6 The relative reductions in energy use from both
the Phase 1 and Phase 2 HD rulemakings as applied to the MY 2014-2016 energy rates developed
using HDIUT data are displayed in Figure 2-51.
100

-------
CU
CC
>
CUD
i_
QJ ¦
C
0.95
o
f\J
i
—
rH
o
r\j
>
U
c
o
r\j
¦a
c
Q_
(J
X
(J
<4—
o
0.85
" 7,
ai QJ
0.75
0.7
•	4
I	1
v

L



Phase 1: MY2014-2020
Phase 2: MY2021-2027+


—1
\


\











\

fcrd
































1 ¦	¦ ¦ ¦















2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029
Model Year
9	Long-haul Combination (MHD) "X™ Long-haul Combination (HHD)
9	Short-haul Combination (MHD) —X™Short-haul Combination (HHD)
0	Transit Bus (Urban Bus)	• • A* • Vocational (LHD)
0	Vocational (MHD)	"X~ Vocational (HHD)
Figure 2-51 HD GHG Phase 1 and Phase 2 rule reductions in relative running energy consumption rates for
LHD2b3, LHD45, MHD, HHD, and Urban Bus diesel vehicles from the MY 2014-2016 baseline
2.1.4.3.2
LHD2b3
LHD2b3 energy reductions are modelled slightly differently than the other heavy-duty vehicles.
Unlike the HD standards for tractors and vocational vehicles, the HD pickup truck/van standards
are evaluated in terms of grams of CO2 per mile or gallons of fuel per 100 miles. For simplicity, we
apply the diesel chassis-certified reductions to all LHD2b3 vehicles since most of the diesel
LHD2b3 vehicles are chassis-certified.0 The LHD engine-certified vehicles are subject to the light-
heavy duty reductions discussed in the previous section that are applied to LHD45 vehicles. In
addition, the fuel economy of medium-duty passenger vehicles (MDPVs) are covered by the Light-
duty GHG rule.62
Because MOVES includes energy rate measurements from LHD vehicles for model years 2014-
2016, we renormalized the Phase 1 reductions starting in MY 2017 so they could be applied to the
MY 2014-2016 rates using Equation 2-33. Example calculations for LHD2b3 diesel in MY 2018-
2020 are provided in Equation 2-35.
0 As discussed in Section 1.4, engine-certified LHD2b3 vehicles are re-classified in MOVES as LHD45 vehicles for
model year 2017 and later.
101

-------
Renormalized Phase 1 reductions for LHD2b3 in 2018 thru 2020 =
1 -15%	1 -15%	85%
= 1	=1	=1	= 1 — 88% = 12% Equation
± ^2.3% + 3% + 6%-j 1-3.8% 96.2%	2-35
Table 2-32 describes the expected changes in CO2 emissions for diesel chassis-certified LHD2b3
vehicles due to improved engine and vehicle technologies due to the HD GHG Phase 1 program.
Note that the impacts of the HD GHG Phase 1 program on gasoline LHD2b3 energy rates are
discussed in Section 3.1.3.2.1. Since nearly all HD pickup trucks and vans will be certified on a
chassis dynamometer, the CO2 reductions for these vehicles are not treated as separate engine and
road-load reduction components, but represented as total vehicle CO2 reductions and applied to all
LHD2b3 vehicles in MOVES. MOVES models the HD pickup truck/van standards by lowering the
energy rates stored in the emissionRate table. No change is made to the road-load coefficients or
weights of passenger or light-duty truck source types. Instead, the energy consumption rates for
LHD2b3 were lowered by the percentages shown in Table 2-32 for the corresponding model years.
Table 2-32 Estimated Total Vehicle Reductions in Energy Consumption Rates for LHD2b3 Diesel Vehicles due
Regulatory
Class
Fuel
Model years
Reduction from
MY 2013 Energy
Rates
Renormalized Reductions Applied to
MY 2014-2016 Energy Rates
(running process)
LHD2b3
Diesel
2014
2.3%
-
2015
3%
-
2016
6%
-
2017
9%
5%
2018-2020
15%
12%
Table 2-33 shows the projected improvements in CO2 emissions due to the HD GHG Phase 2
program for chassis-certified diesel and gasoline LHD2b3 vehicles. These reductions were applied
using the emissionRateAdjustment table for energy and the running process.
Note that we expect the Phase 2 requirements for LHD2b3 to be met via electrification starting in
MY2025. Therefore, we only model reductions in diesel energy rates through MY2024; diesel rates
for MY2025 and beyond are modeled the same as MY2024. Projected electrification rates are
described in the Population and Activity Report.6
Table 2-33 Estimated Total Vehicle Reductions in Energy Consumption Rates for LHD2b3 Diesel and Gasoline
due to the HD GHG Phase 2 Program
Regulatory Class
Fuel
Model years
Reduction from MY
2020 Emission Rates
LHD2b3
Gasoline
and Diesel
2021
2.50%
2022
4.94%
2023
7.31%
2024+
9.63%
102

-------
2.1.4.4
Model Year Trends
Figure 2-52 and Figure 2-53 display the CO2 (g/mile) emission rates and fuel economy values
calculated in MOVES from the energy rates using the carbon content and energy density
conversion factors for conventional diesel fuel as documented in the MOVES4 Greenhouse Gas
and Energy Report.3 The CO2 (g/mile) emission rates and fuel economy values are estimated using
nationally representative operating mode distribution and average speed values. The figures show
that, since model year 2010, there are decreasing trends in CO2 (g/mile) with corresponding
increases in fuel economy, due to the lower MOVES energy consumption rates as well as the lower
source mass values and improved road load coefficients estimated vehicles meeting both Phase 1
and Phase 2 Heavy-Duty Greenhouse Gas Standards. The energy rates by operating mode are
constant for model year 2027-2060. However, some of the small differences in CO2 (g/mile) and
fuel economy values observed within model year groups and regulatory classes are due to
differences in the nationally representative operating modes across model years due changing
fractions of regulatory classes among different source types.
1500
E
S 1000
Reg Class
-+¦ 41-LHD2b3
42-LHD45
46-MHD67
47-HHD8
48-Urban	Bus
49-Gliders
1980	2000	2020	2040
Model Year
Figure 2-52. Base running emission rates for CO2 from age 0-3 diesel heavy-duty vehicles averaged over a
nationally representative operating mode distribution.
103

-------
re
O
at
Q.
Reg Class
41-LHD2b3
¦	42-LHD45
46-MHD67
¦	47-HHD8
48-Urban	Bus
49-Gliders
1980
2000
2020
2040
Model Year
Figure 2-53. Fuel economy for age 0-3 diesel heavy-duty vehicles averaged over a nationally representative
operating mode distribution
2.1.5 Evaluation of Fleet-average Running Rates with Real- World
Measurements
As one evaluation of the MOVES diesel exhaust running rates, Table 2-34 compares preliminary
MOVES3 emission rates estimated in fuel-specific units (g/kg-fuel) to fuel-specific emission rates
estimated from a remote sensing and tunnel measurements. Haugen et al. (2018) conducted exhaust
plume measurements from 1,844 in-use heavy-duty diesel trucks at the Peralta weigh station near
Anaheim, CA in 2017, of which over 63% of the fleet were model year 2011 or later. Wang et al.
(2019), conduced sampling of the Ft. McHenry Tunnel in Baltimore, MD during winter and
summer of 2015. The model year distribution of the Ft. McHenry diesel fleet was not measured.
Wang et al. (2018) estimated the heavy-duty emission factors separately from the light-duty
vehicles using a linear regression which accounted for the fraction of the fleet is composed of
heavy-duty vehicles. The emission rates from both studies are compared to MOVES emission rates
estimated from a national scale run with a preliminary version of MOVES3 conducted for calendar
year 2016 for all on-road heavy-duty diesel vehicles. No effort was made to match the vehicle
operation of the studies or to match the fleet and fuel characteristics (model year distribution,
regulatory class distribution). As such, the comparison is only intended to be a rough comparison,
to assure that MOVES provide estimates that are in the range of feasible values measured from in-
use fleets.
Table 2-34 shows that CO emission factors compare quite well between the different studies. The
NOx values are comparable to the Peralta C A location, but significantly lower than the Ft.
McHenry location. THC are also below the Peralta, CA measurements. PM2.5 is lower than the Ft.
104

-------
McHenry estimates, but within the standard error of the winter measurements, and close to the 95%
confidence range of the summer measurements (approximately two times the standard error). Given
the expected differences in vehicle operation and fleet composition, the comparisons increased our
confidence that MOVES is estimating representative in-use running emission factors for heavy-
duty diesel vehicles.
Table 2-34. Comparison of MOVES Emissions with Remote Sensing and Tunnel Measurements (g/kg-fuel)

Peralta CA 2017
HDV (Haugen et
al. 2018)
Winter Ft.
McHenry
MD 2015
(Wang et al.
2019)
Summer Ft.
McHenry
MD 2015
(Wang et al.
2019)
MOVES3
National
Heavy-duty
Fleet 2016
THC
2.2 ±0.4
NA
NA
0.68
CO
5.9 ±0.9
4.6 ±2.0
7.5 ±2.6
5.2
NOx
12.4 ±0.6
29.6 ±4.7
17.9 ± 1.4
12.2
pm25
NA
0.81 ±0.89
0.61 ±0.11
0.36
Note: The error terms are the standard error of the mean based on individual vehicle measurements for the Peralta
location and sampling periods for the Ft. McHenry Tunnel.
2.2 Start Exh aust Emissions
The start process occurs when the vehicle is started and the engine is not fully warmed up. For
modeling purposes, we define start emissions as the increase in emissions due to an engine start.
Operationally, we estimate difference in emissions between a test cycle with a cold start and the
same cycle with a hot start.p
As explained in Section 1.2.2, we define eight stages which are differentiated by soak time length
(time duration between engine key off and engine key on) between a cold start (> 720 minutes of
soak time) and a hot start FTP (< 6 minutes of soak time). More details on how start emission rates
are calculated as a function of soak time, can be found later in this section and in the MOVES light-
duty exhaust emission rate report.9 The impact of ambient temperature on cold starts is discussed in
the Emission Adjustments MOVES report.63
The next subsections discuss the derivation of heavy-duty diesel start emissions by pollutant and
model year group. Start emissions are currently a small contributor to total exhaust emissions from
heavy-duty diesel vehicles. No T&M or other age effects are currently applied to the diesel start
emissions.
2.2.1 THC, CO, and NOx
The pre-2010 model year emissions are discussed in Section 2.2.1.1 and 2010+ model year
emission rates are discussed in Sections 2.2.1.2.
2.2.1.1 1960-2009 Model Years
For light-duty diesel vehicles, start emissions are estimated by subtracting FTP bag 3 emissions
from FTP bag 1 emissions. Bag 3 and Bag 1 are collected on the same dynamometer cycle, except
105

-------
that Bag 1 starts with a cold start, and Bag 3 begins with a hot start.p A similar approach was
applied for LHD vehicles tested on the FTP and ST01 cycles, which also have separate bags
measuring cold and hot start emissions over identical drive cycles. Data from 21 LHD diesel
vehicles, ranging from model years 1988 to 2000, were analyzed. No classifications were made for
model year or age due to the limited number of vehicles. The results of this analysis for THC, CO,
and NOx are shown in Table 2-35.
Table 2-35 Average Start Emissions Increases (g/start) for pre-2010 Model Year Light Heavy-Duty Diesel
Vehicles for Regulatory Class LHD2b3 and LHD45 (regClassID 41 and 42)
THC
CO
NOx
0.13
1.38
1.68
For pre-2010 model year HHD and MHD trucks, analogous data were unavailable. To provide at
least a minimal amount of information, we measured emissions from a 2007 Cummins ISB which
is used in both LHD and MHD vehicles on an engine dynamometer at the EPA National Vehicle
and Fuel Emissions Laboratory in Ann Arbor, Michigan. Among other idle tests, we performed a
cold start idle test at 1,100 RPM lasting four hours, long enough for the engine to warm up.
Essentially, the "drive cycle" we used to compare cold start and warm emissions was the idle cycle,
analogous to the FTP and ST01 cycles used for LHD vehicles. Emissions and temperature
stabilized about 25 minutes into the test. The emission rates through time are shown in Figure 2-54.
The biggest drop in emission rate over the test was with CO, whereas there was a slight increase in
NOx (implying that cold start NOx is lower than running NOx), and an insignificant change in THC.
50 -1
45
40 -
35
30 -
|> 25 -
20 -
0 -I	1	1	1	1	1
0.00	1.00	2.00	3.00	4.00	5.00
time [hrs]
Figure 2-54 Trends in the Stabilization of Idle Emissions from a Diesel Engine Following a Cold Start (from a
2007 Cummins ISB Measured on an Engine Dynamometer)
p As discussed in Section 1.2.2, ideally, bag 3 would not include a start, but only include running emissions.
Operationally, we use bag 3 with a hot-start because that is the available data, and we assume that the hot-start
emissions are small in comparison to the cold-start emissions, and thus have minimal impact on the cold-start estimate.
Our estimates of emissions by soak time in Sections 2.2.3.1 and 2.2.3.2 support this assumption; for pre-2010, hot-start
THC and CO emissions are less than 10% of the cold-start emissions, and NOx hot-start emissions are less than 20% of
cold-start emissions. For 2010+ emissions the hot-start emissions for THC, CO, and NOx are less than 1% of the cold-
start emissions.
106

-------
We calculated the area under each curve for the first 25 minutes and divided by 25 minutes to get
the average emission rate during the cold start idle portion. Then, we averaged the data for the
warm idle portion using the remaining portion of the test (215 minutes). We then calculated the the
difference between cold start and warm idle over a 25-minute period of the elevated cold starts as
shown in Equation 2-36.
Grams per Start =
= stabilization time x (cold start average rate — hot running average rate)
, 25	240	Equation 2-36
. v-1 emissions v-1 emissions \
= 25 minutes x
-I
25 L-> 215
U=0	t=25
The results are shown in Table 2-36. The measured THC increment is zero. The NOx increment is
negative since cold start emissions were lower than warm idle emissions.
Table 2-36. Cold-start Emissions Increases (g/start) in Grams on the 2007 Cummins ISB
THC
CO
NOx
0.0
16.0
-2.3
We also considered NOx data from University of Tennessee,64 which tested 24 trucks with PEMS
at different load levels during idling. Each truck was tested with a cold start going into low-RPM
idle with air-conditioning on. We again used Equation 2-36 to integrated the emissions over the
warm-up period to get the total cold start idling emissions. We calculated the warm idling
emissions by multiplying the reported warm idling rate by the stabilization time. We used the
stabilization period from our engine dynamometer tests (25 minutes). Then, we subtracted the cold
start-idle emissions from the warm idle emissions to estimate the cold start increment. We found
that several trucks produced lower NOx emissions during cold start (similar to our own work
described above), and several trucks produced higher NOx emissions during cold start. Due to these
conflicting results, and the recognition that many factors affect NOx emission during start (e.g., air-
fuel ratio, injection timing, etc.), we set the default NOx cold-start increment to zero. Table 2-37
shows our final MOVES inputs for HHD and MHD diesel start emissions increases from our 2007
MY in-house testing. Due to the limited data, the emission rate is constant for all pre-2010 model
years and ages.
Table 2-37. MOVES Inputs for Pre-2010 HHD and MHD Diesel Start Emissions (grams/start) for Regulatory
THC
CO
NOx
0.0
16.0
0.0
As discussed in the Emission Adjustments Report63, MOVES applies an additive adjustment to
diesel THC cold-start emissions for ambient temperatures below 72 F. Thus, despite a pre-2010
baseline THC start emission rate of zero, MOVES estimates positive THC start emissions from
heavy-duty diesel vehicles at ambient temperatures below 72 F. No temperature adjustments are
107

-------
applied to CO, PM2.5, or NOx diesel start emissions because no clear trend was found with the
data.
2.2.1.2 2010-2026Model Years
The cold start emissions for 2010 model year and later LHD, MHD, and HHD diesel engines have
been updated for MOVES3 based on new data. However, because of the small sample size and lack
of real-world data, notable uncertainty about real world heavy-duty diesel start emissions remains.
Similar to the approach taken for light-duty vehicles, the cold start emissions are defined as the
difference in emissions between a test cycle with a cold start and the same test cycle with a hot
start. Heavy-duty diesel engines are certified using the Heavy-Duty Diesel Engine Federal Test
Procedure (FTP) cycle65. The test procedure for certification requires that manufacturers run the
engine over the FTP cycle with a cold start and then repeat the cycle with a warm start. Starting in
model year 2016, EPA began collecting certification data that contained separate cold and hot
results for each engine certified. The data that was analyzed for MOVES3 includes the following
engine families from 2016 and 2017 model years shown in Table 2-38.
Table 2-38 Engine Data Analyzed to Estimate the Cold Start Emission Rates for HP Diesel Engines
Category
Number of Engines
Manufacturers
LHD
5
Ford, Isuzu, Hino, FPT
MHD
6
Ford, Hino, Cummins, Detroit Diesel
HHD
11
Cummins, PACCAR, Detroit Diesel,
Volvo, Hino
The certification data was used to determine the grams emitted per cold start using Equation 2-37.
Grams per Start
= [Cold FTP Emission Results (g/(hp — hr))
— Hot FTP Emission Results (g/(hp — hr))]	Equation 2-37
* FTP Cycle Work (hp — hr)
The amount of work (hp-hr) performed over the FTP cycle is required to convert the FTP emission
results in grams per horsepower-hour into grams, but it is not provided as part of the certification
data submitted by the manufacturers to EPA. Furthermore, the FTP cycle work is unique to each
engine and is generally calculated based on the engine's maximum speed, curb idle speed, and the
maximum torque curve. Therefore, we needed to develop a surrogate from the information that is
provided by manufacturers for certification for each engine. We determined that the rated power of
an engine correlates well to the FTP cycle work. This analysis was based on FTP cycle work and
rated power data from ten HD engines. As shown in Figure 2-55, the FTP cycle work is
approximately a linear function of the engine's rated power. For the calculation of cold start
emissions for each engine analyzed, the FTP cycle work (hp-hr) was estimated for the engine based
on its rated power using the equation in Figure 2-55 - 0.0599 (hr) times the rated power (hp) plus
4.4297 (hp-hr).
108

-------
FTP workvs Rated Power - Diesel
V = 0.0559X + 4.4297



	RJ
: = 0.8503
•




•

	






9




250	300	350	400	450	500	550
Rated Power (hp)
Figure 2-55: Relationship between HD Diesel Engine Rated Power and FTP Cycle Work
2.2.1.2.1 Heavy-heavy duty
Analysis of cold and hot start FTP certification data from eleven HHD diesel engines determined
the grams per start for THC, CO, NOx, and PM2.5. The average and standard deviation of the THC,
CO, and NOx emission levels of the eleven engines are shown in Table 2-39. The PM2.5 emissions
are summarized in Table 2-45. The sample included both MY2016 and MY2017 engines, ranging
in displacement between 7.7 and 14.9 liters, and in rated power between 260 and 605 HP. The
default cold start emissions values in MOVES are the mean values shown in the table.
Table 2-39: Cold Start Emissions for MY2010 and Later Heavy Heavy-Duty Diesel Engines
Grams per Start
THC
CO
NOx
Mean
0.08
6.6
8.4
Standard Deviation of Data
0.1
5.6
1.7
2.2.1.2.2 Medium-heavy duty
The certification data from six MHD diesel engines were used to develop the THC, CO, and NOx
grams emitted per start. The average and standard deviation of the emissions from the six engines
are shown in Table 2-40. The sample included MY2016 and MY2017 engines, ranging in
displacement between 5.1 and 8.9 liters, and in rated power between 230 and 380 HP. The default
values in MOVES are the mean values shown in the table.
Table 2-40 Cold Start Emissions for MY2010 and Later Medium Heavy-Duty Diesel Engines
Grams per Start
THC
CO
NOx
Mean
0.20
2.5
6.4
Standard Deviation of Data
0.2
2.7
1.8
2.2.1.2.3 Light-heavy duty
Analysis of five LHD diesel engines from the certification data determined the grams per start for
THC, CO, and NOx shown in Table 2-41. The sample included MY2016 and MY2017 engines,
109

-------
ranging in displacement between 3.0 and 6.7 liters, and in rated power between 161 and 330 HP.
The default values in MOVES are the mean values.
Table 2-41 Cold Start Emissions for MY2010 and Later Light Heavy-Duty Diesel Engines (LHD45 and
	LHD2b3)	
Grams per Start
THC
CO
NOx
Mean
0.005
2.47
6.77
Standard Deviation of Data
0.11
2.61
2.24
We are applying the new cold start THC, CO, and NOx emission rates from the 2016 MY and 2017
MY engines to all 2010 MY and newer engines. The latest tier of HD diesel emission standards
completed phase-in in 2010 MY and the aftertreatment systems on these engines are similar and
generally include both a diesel particulate filter and selective catalytic reduction system.
2.2.1.2.4 Incorporation of Tier 3 Standards for Light Heavy-Duty Diesel
The Tier 3 exhaust emission standards affect light heavy-duty diesel vehicles in the LHD2b3
regulatory class (regClassID 41). Reductions are applied to start rates for NOx only, phasing in
from MY2018 to MY2021 as previously described for running emissions in Section 2.1.1.5.5. No
reductions applies to THC and CO rates.
2.2.1.3 202 7-2060 Model Years
As noted in Section 2.1.1.6, the HD2027 standards include duty-cycle standards, off-cycle
standards and changes to warranty and useful life requirements. To account for the HD2027
standards, we updated NOx start emission rates to reflect the changes in the duty-cycle standards
using the method described in this section.
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 MOVES4 are not based on in-use data but are
based on emissions data from the FTP duty-cycle. Additionally, because the heavy-duty diesel start
emission rates in MOVES4 do not vary with age due to insufficient data, we did not estimate
changes due to the changes in warranty and useful life.
Because engines meeting the HD2027 standard are not yet in production, to update the NOx start
emission rates for MY2027+, we estimated the NOx cold start emission rate (g/start) from a CARB
Stage 1 HDD engine66 tested on the FTP duty-cycle cycle after different periods of use (aging).
Table 2-42 contains the NOx Cold and Hot FTP measurements in Columns (B) and (C) for different
aging periods. Column (E), "Cold - Hot," is calculated as the difference between Columns (B) and
(C). The cold start, Column (F), is then calculated by multiplying the difference in Column (E) by
the work performed on the FTP cycle, Column (D), as shown in Equation 2-38.
NOx Cold Start f—)
Vstart'
= [C°ld (hjvhr) ~ HOt(hjvhr)
T-.rrin W1 , N Equation 2-38
x FTP work (hp ¦ hr)
110

-------
Table 2-42 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 was 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 2-42 demonstrates that
the larger cold start measured with increased aged hours, and after the DPF 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
2-42) to represent the 12-hour cold-start (operating mode 108) emission rate.
To estimate the 12-hour cold-start NOx emission rate for HHD diesel vehicles subject to the
HD2027 standards, we interpolated the HHD 12-hour cold-start between the Stage 1 cold start
(3.14 g/start) and the MOVES baseline (MY2010-2026) 12-hour cold-start (8.4 g/start), and their
respective FTP duty-cycle standards using Equation 2-39 as shown in Figure 2-56 and Table 2-43.
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.
Ill

-------
Start ERFXPx HHD12 hour
/MOVES start HHD12 h0ur ~ Stage 1 start\
V Baseline FTP — Stagel FTP J
Where:
x (FTPX
Start ERFXPx hhd 12 hour ~
Stagel start =
Stagel FTP =
MOVES start HHD,i2hour=
Baseline FTP =
FTP„ =
Baseline FTP — Stagel FTP
- Baseline FTP) + MOVES start HHD12 hour
Equation 2-39
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).
1000 Post Ash Clean start emission rate from the CARB Stage 1 HHD
diesel engine = 3.14 g/start (Table 2-42)
Composite FTP level of the CARB Stage 1 engine = 0.02 g/hp-hr
MOVES3 baseline start emission rate (= 8.4 g/start) for MY 2027 heavy
heavy-duty diesel engine for a 12-hour soak (operating mode 108)
baseline FTP composite NOx standard = 0.2 g/hp-hr
composite FTP standard in the HD2027 standards
9
_ 8
4->
L_
03 7
4-> /
IS)
4-»
5 5



























¦a
o 4
(J
^ 3
o
T 2
r\l
T—1
Q 1
X
xo
c



























) 0.05 0.1 0.15 0.2 0.25
FTP Composite Standard (g/hp-hr)
Figure 2-56 Calculated relationship between the HHD NOx 12-hour cold-start and the composite FTP NOx
standards
112

-------
Table 2-43 HHD Cold Start Emissions for Baseline and HD2027 Standards
Scenario
Applicable Model
Years
Weighted
Average FTP
standard (g/hp-
hr)
Cold Start
emissions (g/start)
Baseline
Model Year 2010-2026
0.2
8.40
HD2027 Standards
Model Year 2027+
0.05
4.02
We assumed that the relative difference in cold start emission rates by regulatory class is the same
in the baseline and HD2027 standards. This calculation was combined with the estimate of
emissions by start operating mode and is described in Equation 2-40 in Section 2.2.3.
2.2.1.4 Model Year Summary
Figure 2-57 through Figure 2-59 display the cold start (operating mode 108) emission rates across
model years for heavy-duty diesel vehicles. The figures show the large difference in start emission
rates before and after model year 2010. Model year 2010 corresponds to the implementation of
(SCR) aftertreatment, as well as the different datasets and methodologies. The rates for LHD2b3
are lower starting in MY2018 due to the phase-in of Tier 3 standards. The HD NOx emission rates
change in MY2027 due to the HD2027 rule.
0.20'
0.15'
re
*—•
w

-------
1980
2000
2020
2040
Model Year
Reg Class
41-LHD2b3
¦	42-LHD45
46-MHD67
¦	47-HHD8
48-Urban	Bus
49-Gliders
Figure 2-58 Heavy-duty Diesel CO Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class
and Model Year
1980
2000
2020
2040
Model Year
Reg Class
41-LHD2b3
¦	42-LHD45
46-MHD67
¦	47-HHD8
48-Urban	Bus
49-Gliders
Figure 2-59 Heavy-duty Diesel NOx Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class
and Model Year
114

-------
2.2.2 Particulate Matter (PM2. s)
2.2.2.1 1960-2010Model Years
Data for particulate matter start emissions from heavy-duty vehicles are limited. Typically, heavy-
duty vehicle emission measurements are performed on fully warmed up vehicles. These procedures
bypass the engine crank and early operating periods when the vehicle is not fully warmed up.
Data for model year 2009-and-earlier vehicles was only available from engine dynamometer testing
performed on one heavy heavy-duty diesel engine, using the FTP cycle with particulate mass
collected on filters. The engine was manufactured in MY 2004. The cycle was repeated six times,
under both hot and cold start conditions (two tests for cold start and four replicate tests for hot
start). The average difference in PM2.5 emissions (filter measurement - FTP cycle) was 0.11 grams.
The data are shown in Table 2-44.
Table 2-44 Average PM2.5 emissions (grams) from MY 2004 HHP diesel engine tested on the FTP Cycle

PM2 5 emissions (grams)
Cold start FTP average
1.93
Warm start FTP average
1.82
Cold start - warm start
0.11
We use the difference between the cold start and warm start bags to represent the cold start (g/start
in MOVES.p We applied this value to 1960 through 2006 model year vehicles. For 2007 through
2009 model years, we applied a 90 percent reduction to account for the expected use of DPFs,
leading to a corresponding value of 0.011 g/start. The value is the same for all heavy-duty diesel
regulatory classes.
As introduced in Section 2.1.2.1.8, in MOVES, the PM2.5 emission rates are estimated as the
elemental carbon (EC) and non-elemental carbon PM (nonEC). We estimated the EC and nonEC
from the total PM2.5 starts rates by applying the EC/PM fraction of 46.4 percent from the PM2.5
speciation profile developed from the idle mode of the UDDS tests from the E55/59 program for
pre-2007 trucks.1 For all 2007+ trucks, we apply the EC/PM fraction of 9.98 percent from the
PM2.5 speciation profile developed from trucks equipped with diesel particulate filters.1
2.2.2.2 2010-2060 Model Years
The cold start emissions for 2010 model year and later LHD, MHD, and HHD diesel engines were
updated in MOVES3 based on new data. We updated the cold start particulate matter emission
rates based on the certification data and data analysis methods discussed in Section 2.2.1.2. The
resulting cold start emission rates for each HD diesel engine regulatory group are shown in Table
2-45. For LHD diesel vehicles, the certification data yielded zero PM2.5 start emissions. We
attribute the zero start to the uncertainty of the data (note the standard deviation shown in Table
2-45 is of similar magnitude to that of HHD and MHD). Instead of using the certification test data,
we used the data from MHD diesel to represent the LHD diesel PM2.5 emission rate in MOVES,
because of the overlap in engines and aftertreatment systems between the two categories.
115

-------
Table 2-45: Cold Start PM2.5 Emission Rates for Heavy-Duty Diesel Emissions for 2010+ MY
Grams per Start
HHD
MHD
LHD Test Data
LHD for MOVES3
Mean of Data
0.013
0.008
0.000
0.008
Standard Deviation of
Data
0.029
0.017
0.010

Note:
a Instead of using the test data, we used the data from MHD diesel to represent the LHD diesel PM2.5 emission rate in
MOVES as noted above.
We are applying the new cold start PM2.5 emission rates from the model year 2016 and 2017
engines to MY 2010 and newer engines because the PM standards are the same and all the MY
2010 and later engines generally include both a diesel particulate filter (DPF) and selective
catalytic reduction (SCR) system.
2.2.2.3 Model Year Summary
Figure 2-59 and Figure 2-60 display the cold start (operating mode 108) emission rates across
model years for heavy-duty diesel vehicles. As expected, large reductions are shown in model year
2007 with the implementation of diesel particulate filters. Further changes are due to the
incorporation of the 2010 and later certification data.
- 0.09'
03
•*->
W
D>
C3
1 0.06'
w
o
o
10
c\i
- 0.03'
TO
*—•
O
0.00'
Reg Class
41-LHD2b3
¦	42-LHD45
46-MHD67
¦	47-HHD8
48-Urban	Bus
49-Gliders
1980
2000
2020
2040
Model Year
Figure 2-60 Heavy-duty Diesel PM2.5 Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class
and Model Year. Urban Bus and HHD are equivalent. MHD, LHD45, LHD2b3 are equivalent.
116

-------
2.2.3 Adjusting Start Rates for Soak Time
The discussion to this point has concerned the development of rates for cold start emissions from
heavy-duty diesel vehicles. In addition, it was necessary to derive rates for additional operating
modes that account for shorter soak times. As with light-duty vehicles, we accomplished this step
by applying soak fractions.
In the MOVES input database, operating modes for start emissions are defined in terms of soak
time preceding an engine start. The "cold-start" is defined as a start following a soak period of at
least 720 minutes (12 hours) and is represented as opModeID=108. An additional seven modes are
defined in terms of soak times ranging from 3 min up to 540 min (opModelD = 101-107). Table
1-5 describes the different start-related operating modes in MOVES as a function of soak time. The
distribution of vehicle start activity among the start operating modes is described in the MOVES
Vehicle Population and Activity report.6
2.2.3.1 Adjusting Start Rates for Soak Time - MY 2009 and Earlier
The soak adjustment ratios we used for THC, CO, and NOx for MY 2009 and older HD diesel
vehicles are illustrated in Figure 2-61 below. Due to limited data, we applied the same soak ratios
that we applied to 1996+ MY light-duty gasoline vehicle as documented in the light-duty emission
rate report.9 The soak adjustments are taken from the non-catalyst soak adjustments derived in a
CARB report67 and reproduced in a MOBILE6 report.68
Soak Time (minutes)
Figure 2-61. Soak Adjustment Ratios Applied to Cold-Start Emissions (opModelD = 108) to Estimate Emissions
for shorter Soak Periods (operating modes 101-107). This figure is reproduced from the Light-Duty Emissions
Report9
For light heavy-duty vehicles (regulatory classes LHD2b3 and LHD45), the soak ratios apply to the
cold starts for THC, CO and NOx. For medium and heavy heavy-duty vehicles (regulatory classes
MHD, HHD, and Urban Bus), only the CO soak ratios are applied to the cold-start emissions,
because the base cold start THC and NOx emission rates for medium and heavy heavy-duty
117

-------
emission rates are zero (see Section 2.2.1.1). The start emission rates entered into MOVES for
2009 and older model year heavy-duty vehicles, derived from applying the soak ratios are
displayed in Table 2-46 for THC, CO, and NOx.
Table 2-46. Heavy-Duty diesel THC, CO, and NOx Start Emissions (g/start) by Operating Mode for 2009 and
Earlier Model Year Vehicles

THC
CO
NOx
opModelD
LHD1
Other HD2
LHD
Other HD
LHD
Other HD
101
0.0052
0
0.055
0.64
0.275
0
102
0.0273
0
0.276
3.2
0.760
0
103
0.0572
0
0.607
7.04
1.350
0
104
0.0780
0
0.869
10.08
1.481
0
105
0.0832
0
1.007
11.68
1.481
0
106
0.0949
0
1.090
12.64
1.468
0
107
0.1183
0
1.256
14.56
1.376
0
108
0.1300
0
1.380
16
1.298
0
Notes:
1	LHD refers to LHD2b3 and LHD45
2	Other HD refers to the medium heavy-duty, heavy heavy-duty, and urban bus regulatory classes
The PM2.5 start rates by operating mode are given in Table 2-47 below. They are estimated by
assuming a linear decrease in emissions with time between a full cold start (>720 minutes) and zero
emissions at a short soak time (< 6 minutes).
Table 2-47. Particulate Matter Start Emission Rates (g/start) by Operating Mode (soak fraction) for all HD
	Diesel vehicles through MY 2009	
Operating
Mode
1960-2006
MY
2007-2009
MY
101
0.0000
0.00000
102
0.0009
0.00009
103
0.0046
0.00046
104
0.0092
0.00092
105
0.0138
0.00138
106
0.0183
0.00183
107
0.0549
0.00549
108
0.1099
0.01099
2.2.3.2 Adjusting Start Rates for Soak Time - MY 2010 and Later
As described in the preceding section, the start rates are based on data collected from light-duty
vehicles in the 1990's. The question arose as to whether they could be considered applicable to
heavy-duty diesel vehicles with aftertreatment systems designed to meet the 2007/2010 exhaust
emissions standards. To address this question, we initiated a research program in 2016, with the
goal of examining the relationships between soak time and start emissions for a set of heavy-duty
vehicles. Two test programs were conducted to revise the 2010 MY and later soak curves for
118

-------
heavy-duty diesel vehicles in M0VES3 and later versions. The testing consisted of both chassis
and onroad testing of MY 2015 and MY 2016 vehicles.
The first test program included a MY 2015 day-cab tractor with a MY 2015 HHD diesel engine
tested on a heavy-duty chassis.69 The vehicle was relatively new and had 10,000 miles on the
odometer. The testing consisted of running two repeats of a transient drive cycle developed by the
National Renewable Energy Laboratory (NREL). The vehicle speed trace is shown below in
Figure 2-62. Prior to each soak test, the vehicle was first run through two of the NREL cycles.
Then the engine was shut off for a specified amount of time to reflect the soak periods shown in
Figure 2-62. At least two repeats were conducted for each soak period. The emission measurements
included dilute gaseous measurements and triplicate particulate matter filters.
Double NREL Cycle
60
U"5rHvO(NCQmff'),TOs£>'Hr-.
-------
SO |
7.0
T 6.0
J. 5 0
M
£ 4.0
1 3-0 1
E,„
lj 2.0 J
1.0 1
0.0 '
c




NOx















I
•

•


•

•









•
a










>










r










i





















t 10 20 30 40 50 50 70 80 90 100
Soak Time (hours)
Figure 2-63 MY 2015 Heavy-Duty Vehicle NOx Emissions by Soak Time
0.06
0.05
~ 0.04
S
£ 0.03
o
~ 0.02
lj
0.01
0.00 t
(



HC



	f	
<
i







9
•




•




•
•

*






•

•
• •










•1
i




















) 10 20 30 40 50 50 70 SO 90 100
Soak Time (hours)
Figure 2-64 MY 2015 Heavy-Duty Vehicle THC Emissions by Soak Time
120

-------
0.50
0.45
	0.40
OJ
~ 0.55
3 °-30
£ 0.25
"5 0-20
¦j= 0.15
m 0.10
0.05 J
0.00 '
t




CO
















•

*


•	


A t
1









<









•











•








•









1;










J.










¥










) 10 20 30 40 50 60 70 SD 90 100
Soak Time (hours)
Figure 2-65 MY 2015 Heavy-Duty Vehicle CO Emissions by Soak Time
a r.



PM


8.0
7.0
aj
~ 6.0
E 5.0
I '
.2 3.0 ,
2.0
1.0
0.0
(









*


































L.










y

•
»








/
•






• *



















X)
10 20 30 40 50 60 70 BO 90 It
Soak Time (hours)
Figure 2-66 MY 2015 Heavy-Duty Vehicle PM2.5 Emissions by Soak Time
In addition to the chassis testing, onroad testing was conducted using a portable emissions
measurement system (PEMS).70 The emissions data gathered by the PEMS in this test program
only included the gaseous emissions, not PM data. A MY 2016 work van with a diesel engine was
tested on the road. The vehicle was soaked and started within a laboratory under controlled
temperatures. All onroad testing occurred with ambient temperatures over 50 degrees F. Each test
began with 10 seconds of idle followed by driving a defined "soak route." A typical vehicle speed
121

-------
profile from the route is shown in Figure 2-67. The route consisted of approximately 700 seconds
of driving in a neighborhood/urban environment over approximately 2.7 miles.
Soak Route Speed Profile
0	200	400	600
T*ip Courier 4sectrrf4|
Figure 2-67 Onroad Soak Drive Route
The emission results, in terms of total emissions over the route, from the onroad tests are shown in
Figure 2-68 through Figure 2-70.
3.0
2.5

Total NOx Over Cycle





•


s
c 20
o
wi
^ 1.5
E
LjJ
75 1.0
4-1
o
H
0.5
0.0
{










•
•
a


•
•

•
•


>













) 5 10 15 20 25 30
Soak Time (hours)
Figure 2-68 MY 2016 Heavy-Duty Vehicle NOx Emissions by Soak Time
122

-------
0.8
0.7
"bD 0.6
l/l
J 0-5
1/1
Ln 0 4
E
^ 0.3
ra
£ 0.2
0.1
0.0 '
t

Total HC Over Cycle



























•










•

••
• •


• * .
A






f





0
) 5 10 15 20 25 31
Soak Time (hours)
Figure 2-69 MY 2016 Heavy-Duty Vehicle THC Emissions by Soak Time
a n
Total CO Over Cycle


SO
— 7.0
£ 6.0
o
S 5.0
Ln
E 4.0
LjJ
"ra 3.0
-M
,o
1- 2.0
1.0
0.0
{




•











a






w
•
~





A


m #
•

•



m






f






»





0
5 10 15 20 25 3
Soak Time (hours)
Figure 2-70 MY 2016 Heavy-Duty Vehicle CO Emissions by Soak Time
The soak emission adjustment ratios were calculated using a multi-step process based on the
chassis test and onroad test results. First, the total emissions over the route or drive cycle were
averaged for each soak period for each pollutant (NOx, THC, CO) for each vehicle. Then the start
emissions for each soak period were determined by subtracting the average total emissions from the
tests with the 3 minute soak time from the emissions from the specific soak period. The ratios for
soak period operating modes 102 through 108 were calculated based on the average start emissions
of the soak period divided by the average start emissions of the cold start (>12 hours) soak period.
The soak fractions for the operating mode 101 were determined by extrapolating the value from the
operating mode 102 result using the proportional difference in time between the midpoints of each
operating mode 101 and 102 soak times. In other words, soak fraction for operating mode 102 was
123

-------
multiplied by the ratio of 3 minutes divided by 18 minutes (the midpoint times of operating mode
101 and 102). The NOx, CO, and THC soak period ratio results for each vehicle are shown below
in Figure 2-71.
HD Diesel Truck Start Emission Ratio
1.20
1.00
.2 0.80
ts
cd
o 0.60
tn
& 0-40
0.20
0.00
0	100 200 300 400 500 600 700 800
Soak Time (min)
¦•¦•#¦¦¦ HD Van NO*	HD Van CO	HD Van HC
—Day Ccto NOx	—•--DayCabCO	—•—DayCabHC







/













£




*

1/


f i
/•'
/ ji
wr
P"	^
T""*1
^ 4t tusiz,4 ,V ^



4
A./
, . 4 . -J
If-'






Figure 2-71 Soak Emission Ratios from a MY 2015 HD Day-Cab and a MY 2016 HD Van
The 2010 MY and later heavy-duty diesel soak ratios for MOVES were determined by averaging
the results from the two trucks. The resulting soak adjustment ratios are shown in Table 2-48. The
soak adjustment ratios are applied to all heavy-duty diesel regulatory classes because the two trucks
tested cover the range of HD diesel regulatory classes.
124

-------
Table 2-48 HP Diesel Engine Soak Ratios for MY 2010 and Newer
Operating
Mode
Description
NOx
CO
THC
101
Soak Time < 6 minutes
0.01
0.00
0.00
102
6 minutes < Soak Time < 30 minutes
0.04
0.03
0.02
103
30 minutes < Soak Time < 60 minutes
0.13
0.06
0.05
104
60 minutes < Soak Time < 90 minutes
0.33
0.02
0.24
105
90 minutes < Soak Time < 120 minutes
0.40
0.12
0.36
106
120 minutes < Soak Time < 360 minutes
0.37
0.32
0.33
107
360 minutes < Soak Time < 720 minutes
0.62
0.38
0.55
108
720 minutes < Soak Time
1.00
1.00
1.00
For MY2027+ vehicles subject to HD2027 standards, we used Table 2-40 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. We assumed that the
relative difference in emission rates by regulatory class and by operating mode is the same for
MY2010-2026 and MY2027 and later meeting the HD2027 standards.
Start ERFXP=x reg class=y,soak=z
— Start ER[)uty cycle standard x,HHD,12
x /MOVES start regclass=y,soak=z\	Equation 2-40
X\ MOVES Start HHD,12-hour /
Where:
Start ERfxp= the start NOx emission rates for the HD2027 standards with FTP x (0.035 or 0.05) 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 = MOVES3 baseline start emission rate for MY 2027 for regulatory class y (LHD45,
MHD, and HHD), and soak length z
MOVES start HHD,i2-hour= MOVES3 baseline start emission rate for MY 2027 HHD diesel engine for a 12-hour soak
(operating mode 108)
For example, Figure 2-72 compares the estimated MOVES NOx start emission rates for HHD
diesel vehicles for MY2010-2026 and MY2027 and later.
125

-------
¦ Without HD2027
HD2027 Standards Rate for MY2027+
<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 2-72 Duty-cycle-based NOx start emissions for HHD Diesel comparing MY2010-2026 and MY2027+
rates.
The PM2.5 start rates by operating mode for MY 2010 and newer vehicles are presented in Table
2-49 below. They were updated in MOVES3 using a linear interpolation based on the new cold
start data (certification data discussed in Section 2.2.2.2) for Operating Mode 108. They are
estimated by assuming a linear decrease in emissions with time between a full cold start (>720
minutes) and zero emissions at a short soak time (< 6 minutes). This approach is consistent with the
approach taken for MY 2009 and older vehicles, as described in Section 2.2.3.1. We did not revise
the approach because we obtained PM2.5 data for only one of the trucks and it showed mixed soak
effect results.
126

-------
Table 2-49 PM2.5 Start Emission Rates (g/start) by Regulatory Class and Operating Mode (soak fraction) for all
MY 2010 and newer HD Diesel Vehicles
Operating
Mode
HHD and
Urban Bus
MHD
LHD2b3 and
LHD45
101
0.00000
0.00000
0.00000
102
0.00163
0.00100
0.00100
103
0.00325
0.00200
0.00200
104
0.00488
0.00300
0.00300
105
0.00650
0.00400
0.00400
106
0.00813
0.00500
0.00500
107
0.00975
0.00600
0.00600
108
0.01300
0.00800
0.00800
2.2.3.3 Adjusting Start Rates for Ambient Temperature
The ambient temperature effects in MOVES are used to estimate the impact ambient temperature
has on cooling the engine and aftertreatment system on vehicle emissions. The temperature effect is
the greatest for a vehicle that has been soaking for a long period of time, such that the vehicle is at
ambient temperature. Accordingly, the impact of ambient temperature should be less for vehicles
that are still warm from driving. The emission adjustments report discusses the impact of ambient
temperature on cold start emission rates (operating mode 108).63 The ambient temperature effects
for starts with warm and hot soaks (operating mode 101-107) are documented below and recorded
in the MOVES startTempAdjustment table.
Because the THC temperature effects in MOVES are modeled as additive adjustments, the
adjustment calculated for cold starts needs to be reduced for warm and hot starts. Due to lack of
data, we multiply the soak fractions described earlier in Figure 2-61 for pre-2007 trucks by the
additive cold temperature effect for the 12-hour cold start (operating mode 108) to obtain cold start
temperature adjustments for the warm and hot soaks starts (operating mode 101 through 107) for
all model years.q The additive cold start adjustment for THC emission factors are displayed in
Table 2-50, along with the soak fractions applied. These additive THC starts are applied to all
diesel sources in MOVES, including light-duty diesel (regulatory class LDV and LDT). There are
currently no diesel temperature effects in MOVES for PM2.5, CO, and NOx.
q The temperature effects from pre-2010 technology engines are applied to all model years. We plan to update the
temperature effects by operating mode for 2010 and later model year vehicles in future version of MOVES using the
data from 2010 and later engines.
127

-------
Table 2-50 THC Diesel Start Ambient Temperature Adjustment by Operating Mode
Operating
mode ID
Start Temp Adjustment
Soak fraction
101
-0.0153x(Temp-75)
0.38
102
-0.0152x(Temp - 75)
0.37
103
-0.0180x(Temp - 75)
0.44
104
-0.0201 x(Temp - 75)
0.5
105
-0.021 lx(Temp - 75)
0.52
106
-0.0254x(Temp - 75)
0.62
107
-0.0349x(Temp - 75)
0.86
108
-0.0406x(Temp - 75)
1
2.2.4 Start Energy Rates
The start energy rates (in units of kJ) were developed for MOVES200471, and updated in
MOVES2010 as documented in the MOVES2010a energy updates report.57 Figure 2-73 displays
the cold starts in grams of CO2 emissions calculated from the energy rates using the carbon content
for conventional diesel fuel as documented in the MOVES3 Greenhouse Gas and Energy Report.3
As shown, there is more detail in the pre-2000 energy rates. The spike in CO2 g/start for model
years 1984-1985 reflects variability in the data used to derive starts, which was consistent with the
more detailed approach used to derive the pre-2000 energy rates in MOVES2004. The only updates
to the start energy rates post-2000 is the impact of the Phase 1 Heavy-Duty GHG standards, which
began phase-in in 2014 and have the same reductions as the running energy rates as presented in
Table 2-30 and Table 2-32. It is worth noting that unlike the Phase 1 HD GHG standards, the
technologies projected for meeting the Phase 2 HD GHG standards are not expected to have an
impact on start energy rates. Therefore, the start energy rates are constant after MY 2018 (the first
year of full phase-in of the HD Phase 1 rule).
128

-------
400'
r
(5
¦4-<
w
s

-------
Table 2-52. Relative contribution of total energy consumption from each pollutant process by regulatory class
processID
processName
LHD<14K
LHD45
MHD
HHD
Urban
Bus
Gliders
1
Running
Exhaust
98.5%
99.3%
99.42%
98.85%
99.7%
98.63%
2
Start Exhaust
1.5%
0.7%
0.55%
0.10%
0.3%
0.05%
90
Extended Idle
Exhaust


0.03%
1.03%

1.27%
91
Auxiliary
Power Exhaust


0.00%
0.03%

0.05%
2.3 Exten ded Idling Exh aust Emissions
In the MOVES model, extended idling is idle operation characterized by long duration idle periods
(e.g., > 1 hour1) , typically overnight, including higher engine speed settings and extensive use of
accessories by the vehicle operator. Extended idling most often occurs during rest periods by long-
haul trucking operators where the truck is used as a residence (sometimes referred to as
"hotelling"). Operators idle to power accessories such as air conditioning systems or heating
systems. Heavy-duty engine and truck manufacturers recommend trucks not idle at low engine
speeds for extended periods, because it can "create engine wear and carbon soot buildup in the
engine and components."73 Additionally, idling for extended periods allows the vehicle's exhaust
to cool below the effective temperature required for emission aftertreatment systems in modern
trucks such as selective reduction catalysts and diesel oxidation catalysts. As a result, extended idle
is treated as a separate emission process in MOVES which uses a different emission rate than the
idling that occurs during the running emission process.
Extended idling does not include vehicle idle operation that occurs during normal road operation,
such as idling at a traffic signal or the "off-network" idle that might occur during a delivery.
Although frequent stops and idling can contribute to overall emissions, these modes are included in
the normal vehicle hours of operation. Extended idling is characterized by idling periods that last
hours rather than minutes.
In the MOVES model, long-haul combination trucks (sourceTypelD 62) is the only source type
assumed to have extended idling activity. These trucks are only associated with MHD, HHD and
Gliders regulatory classes. As an alternative to extended idling, long-haul truck operators can also
use auxiliary power units (APUs) or plug into facility's power (shore power) to power their cabin
and accessories during hotelling. The emission rates for auxiliary power units (APUs) are discussed
in Section 2.4, and the energy consumption rates for shore power are discussed in the Greenhouse
Gas and Energy Consumption technical report.3
r The default hotelling activity in MOVES3 is estimated from telematics data in which all idle events with duration
greater than one hr from long-haul combination trucks are assigned to extended idling.Error! Bookmark not defined.
s Glider extended idle emission rates are documented in Section 2.5
130

-------
Extended idle emission rates for THC, CO, NOx and PM2.5 were updated in MOVES3 for all model
years. Energy rates were updated for 2007 and later model years. Separate analyses were conducted
using different data sets to derive extended idle emission rates for pre-2007 (Section 2.3.1) and
2007 and later long-haul combination trucks (Section 2.3.2). For each range of model years,
MOVES applies different data and assumptions regarding the impact of accessory use, frequency
of high idle engine speed, and impacts of tampering and mal-maintenance to calculate extended
idle emission rates.
2.3.1 1960-2006 Model Years
The MOVES extended idling emission rates for pre-2007 model years were derived from data
collected in several distinct test programs under different types of idle conditions. For MOVES3
and later versions, weightings were adjusted from those in previous versions of MOVES to better
account for new information on typical extended idling engine idling speeds and loads. These
adjustments are described below in Section 2.3.1.2. Appendix D summarizes the data and
calculations for the pre-2007 model years.
2.3.1.1 Data Sources
The references included in this section provide more detailed descriptions of the data and how the
data were obtained:
•	Testing was conducted on 12 heavy-duty diesel trucks and 12 transit buses in Colorado by
McCormick et al.74 Ten of the trucks were Class 8 heavy-duty semi-tractors, one was a
Class 7 truck, and one of the vehicles was a school bus. The school bus data was not used to
calculate extended idle rates. The model years ranged from 1990 through 1998. Typical
Denver area wintertime diesel fuel was used in all tests. Idle measurements were collected
during a 20-minute time period. All testing was done at 1,609 meters above sea level (high
altitude).
•	Testing was conducted by EPA on five trucks in May 2002 (Lim et al.).75 The model years
ranged from 1985 through 2001. The vehicles were put through a battery of tests including
a variety of idling conditions.
•	A total of 63 trucks (nine in Tennessee, 12 in New York and 42 in California) were tested
over a battery of idle test conditions including with and without air conditioning (Irick et
al.).76 Not all trucks were tested under all conditions. Only results from the testing in
Tennessee and New York are described in the IdleAire report (Irick et al.)76
•	The California test data was collected on 42 diesel trucks in parallel with roadside smoke
opacity testing (Lambert)77. All tests conducted by the California Air Resources Board
(CARB) at a rest area near Tulare, California in April 2002 are described in the Lambert77
Clean Air Study. All analytical equipment for all testing at all locations was operated by
Clean Air Technologies.
•	Fourteen trucks were tested as part of the E-55/59 Coordinating Research Council (CRC)
study of heavy-duty diesel trucks with idling times either 900 or 1,800 seconds long.78
131

-------
•	The National Cooperative Highway Research Program (NCHRP)79 obtained the idling
portion of continuous sampling during transient testing to determine idling emission rates
on two trucks.
•	A total of 33 heavy-duty diesel trucks were tested in an internal study by the City of New
York (Tang et al.)80. The model years ranged from 1984 through 1999. One hundred
seconds of idling were added at the end of the WVU five-mile transient test driving cycle.
•	A Class 8 Freightliner Century with a 1999 engine was tested using EPA's onroad
emissions testing trailer based in Research Triangle Park, North Carolina (Brodrick).81 Both
short (10 minute) and longer (five hour) measurements were made during idling. Some
testing was also done on three older trucks.
•	Five heavy-duty trucks were tested for particulate and NOx emissions under a variety of
conditions at Oak Ridge Laboratories (Storey et al.).82 These are the same trucks used in the
EPA study (Lim et al.).
•	The University of Tennessee (Calcagno et al.) tested 24 1992 through 2006 model year
heavy-duty diesel trucks using a variety of idling conditions including variations of engine
idle speed and load (air conditioning).64
2.3.1.2 Analysis
We used the data sources referenced above to estimate the emission rates for particulate matter
(PM2.5 ), oxides of nitrogen (NOx), hydrocarbons (THC), carbon monoxide (CO) and carbon
dioxide (CO2). The data were grouped by truck and bus and by idle speed and accessory usage to
develop emission rates representative of extended idle emissions.
The important conclusion from the analysis was that truck operator behavior plays an important
role when assigning emission rates to periods of extended idling. Factors such as accessory use and
engine idle speed, which are controlled by operators, affect engine load and emission rates during
extended idling. The impacts of other factors, such as engine size, altitude, model year within
MOVES groups, and test cycle are negligible.
We first evaluated the studies on engine idle speed. NREL's review of owner's manuals found that
several heavy-duty engine manufacturers recommend use of fast idle (> 1000 rpm) if the engine
needs to idle for extended periods.83 In a 2004 UC-Davis survey (Lutsey et al. 2004), respondents'
average engine idle speed was 866 rpm, with small peaks around 650 and 1000 rpm.84 About one-
third of the respondents indicated they changed their idle speed from its usual setting, which is
consistent with the distribution of the responses where about one-third of the idle engine speeds
reported were 1000 rpm or faster. A 2015 study by Hoekzema (2015)85 suggested that even fewer
trucks operated in a high idle condition. Drivers surveyed for this study reported high idle operation
(> 1000 rpm) just 18 percent of the time during idling periods of an hour or more. Additionally,
Hoekzema (2015) cited similar studies representing 764 trucks that averaged engine speeds of 886
rpm during extended idle. Therefore, in MOVES3 and later versions, we reduced the amount of
high idle from 100 percent assumed in MOVES2014 to 33 percent, to better match the references
noted above.
The use of accessories (e.g., air conditioners, heaters, televisions, etc.) provides recreation and
comfort to the operator and increases load on the engine. There is also a tendency to increase idle
132

-------
speed during long idle periods for engine durability. The emission rates estimated for the extended
idle in MOVES assume both accessory use and engine idle speeds set higher than used for "curb"
(non-discretionary) idling. We classify the extended idling that does not employ high speed idle
without additional auxiliary loads as "curb idle."
Emissions data from the references in the data sources section (2.3.1.1) was classified into one of
three idle conditions. The first condition, which has a low engine speed (<1,000 rpm) and no air
conditioning is representative of curb idle. The second condition is representative of extended idle
with higher engine speed (>1,000 rpm) and no air conditioning. The third represents an extended
idle condition with higher engine speed (>1,000 rpm) and air conditioning. For the purpose of this
analysis, the load placed on the engine due to air conditioning is assumed to represent all forms of
accessory load that may be used during hotelling.
Note that some of the idle tests are of short duration. We believe it is reasonable to classify the
short-duration tests as curb idle in our calculations of extended idle emissions. We are using the
short-duration idle tests from the pre-2007 MY vehicles because idle emissions stabilize more
quickly than later model years because the pre-2007 vehicles lack the emission aftertreatment
technologies that can lose effectiveness as exhaust cools during longer idle periods.
For 1990 and earlier, we developed curb idle emission rates based on the analysis of the 18 heavy-
duty diesel trucks from 1975-1990 model years used in the CRC E-55/59 study and one MY 1985
truck from the Lim study. The curb idle rates were then adjusted using ratios from 1991-2006
trucks to estimate the elevated NOx emission rates characteristic of higher engine speed and
accessory loading of extended idle.
In particular, as summarized in the tables in Appendix D, data from 188 vehicles were used to
estimate curb idle NOx emission rates for 1991-2006 model year heavy-duty diesel trucks. The curb
idle NOx emission rate of 91 g/hr was calculated by weighting the average NOx emission rate from
each test by the number of vehicles tested. Four studies and results from 31 vehicles included
higher idle engine speed and air conditioner use, which resulted in a weighted idle NOx emission
rate of 227 g/hr. The ratio of the 1991-2006 MY NOx emission rate from curb idle to idle with high
engine speed and A/C was applied to the 1990 and earlier model year curb idle rate to get the
calculated 1990 and earlier NOx emission rate with high engine speed and A/C. A similar strategy
was applied to the THC, CO, and CO2 emission rates for 1990 and earlier model years.
For both the MY 1960-1990 and 1991-2006 vehicles, using the data summarized in Appendix E,
adjusted emission rates were calculated for each pollutant by weighting the overall "high speed
idle, A/C on" results by 0.33 and the "low speed idle, A/C off' (i.e., curb idle) results by 0.67 to
account for the fraction of idling at high and low engine speeds.
The NOx, THC, CO, and PM2.5 emission rates from this data analysis are primarily from diesel
HHD trucks. In MOVES2014, we calculated the MHD extended idle emission rates as half of the
corresponding HHD emission rates. However, a study by Khan et al. (2009)86 found that MHD and
HHD trucks had similar emission rates during extended idle. Consequently, MOVES applies the
same extended idle emissions rates to MHD and HHD, as shown in Table 2-53.
133

-------
MOVES stores PM2.5 emission rates according to elemental carbon (EC) and non-elemental carbon
(NonECPM), but the data sources used to calculate the extended idle emission rates reported only
total PM2.5. As mentioned in Section 2.1.2.1.8, an EC/PM fraction of 46.4 percent is applied for the
running exhaust idle operating mode (opModelD 1), and we also apply it to extended idle. The
resulting EC and NonECPM rates are also shown in Table 2-53.
No adjustment to the rates are made to account for tampering and mal-maintenance (T&M) because
the pre-2007 trucks do not have the exhaust aftertreatment technologies that are anticipated to see
large emission increases when they are tampered or mal-maintained. While the 188 trucks used for
these estimates may not fully represent real-world emission deterioration, they do include real-
world vehicles at a variety of ages and conditions and thus it would be "double-counting" to apply
the exhaust running T&M effects to these rates.
Table 2-53. Pre-2007 Extended idle emission rates (g/hour) in MOVES by pollutant for MHD and HHP
Model Year
Groups
NOx
THC
CO
pm25
EC
Non-ECPM
Pre-1991
69.3
49.8
50.8
5.39
2.50
2.89
1991-2006
136
25.6
55.0
2.48
1.15
1.33
2.3.2 2007-2026 Model Years
The extended idle emission rates for model years 2007 to 2026 are based on the following data
sources and analysis.
2.3.2.1 Data Sources
The extended idle emission rates for model year 2007 and later heavy-duty diesel combination
long-haul trucks (sourceTypelD 62) diesel emission rates in MOVES are based on two test
programs measuring extended idle emissions from HHD diesel trucks. The Texas Transportation
Institute (TTI) tested extended idle emission from 15 heavy-duty diesel tractors ranging from
model year 20051 to 2012.87 Another study conducted by California Air Resources Board (ARB)88
tested five tractors (engine model years 2007 and 2010). As discussed in the analysis section
(Section 2.3.2.2), the four MY 2005 and 2006 engines included in the TTI study are included in the
development of the 2007 and later model year emission rates for THC, CO, NOx, and energy
because there is no noticeable differences in the emission rate for these model years from
comparable MY 2007 and later engines. For PM2.5, these engines are only used for comparison and
to develop T&M adjustment factors.
The study (TTI or ARB), engine model year, engine manufacturer, odometer, the NOx certification
level, California Clean Idle certification, and engine aftertreatment are listed for each of the trucks
in Table 2-54. The last three columns in Table 2-54 are taken from the California Executive Order
4 Although 2005-2006 model year engine data was available at the time of the MOVES3 MY 2007+ analysis, we
lacked the time and resources to incorporate them into the pre-2007 emission rates.
134

-------
certification database.89 NOx certification level (g/bhp-hr) is the standard to which the engine was
certified. Some 2010 and later engines were certified above the 0.2 g/bhp-hr NOx 2010 federal
standard due to the emissions averaging, banking and trading (ABT) program, and EPA allowance
of nonconformance penalty (NCP) engines in 2012.90 In these cases, the family emission limit for
which the vehicle was certified is reported in Table 2-54. California Clean Idle Certification was
implemented in 2008 and allows engines that are certified to a 30 g/hr idle NOx standard to idle
beyond the 5-minute idle limit initiated in 2008 in California. The aftertreatment column in Table
2-54 indicates whether the engine was certified with an oxidation catalyst (OC), diesel particulate
filter or periodic trap oxidizer (DPF), and/or selective catalytic reduction (SCR) system.
Table 2-54. HHP Diesel Tractors Used to Update the MY 2007 and Later Extended Idle Emission Rates





Clean


Engine


NOx cert
Idle

Study
MY
Engine
Odometer
(g/bhp-hr)
Certified?
Aftertreatment
TTI
2005
Caterpillar
484,550
2.4
No
OC
TTI
2006
Cummins
505,964
2.4
No

TTI
2006
Volvo
640,341
2.4
No

TTI
2007
Cummins
406,740
1.2
No
OC, DPF
ARB
2007
Cummins
390,000
2.2
No
OC, DPF
ARB
2007
DDC
10,700
1.2
No
OC, DPF
TTI
2008
Cummins
353,945
2.4
Yes
OC, DPF
TTI
2008
Mack
82,976
1.2
Yes
DPF
TTI
2009
Mack
96,409
1.2
Yes
OC, DPF
TTI
2010
Mack
89,469
0.2
Yes
OC, DPF, SCR
TTI
2010
Navistar
73,030
0.5
Yes
OC, DPF
TTI
2010
Navistar
57,814
0.5
Yes
OC, DPF
TTI
2010
Navistar
10,724
0.5
Yes
OC, DPF
ARB
2010
Cummins
13,500
0.35
Yes
OC, DPF, SCR
ARB
2010
Navistar
70,000
0.5
Yes
OC, DPF
ARB
2010
Volvo
68,000
0.2
Yes
OC, DPF, SCR
TTI
2011
Mack
95,169
0.2
Yes
OC, DPF, SCR
TTI
2012
Mack
6,056
0.2
Yes
OC, DPF, SCR
TTI
2012
Mack
11,989
0.2
Yes
OC, DPF, SCR
TTI
2012
Mack
25,148
0.2
Yes
OC, DPF, SCR
The 15 trucks from the TTI program were tested in an environmental chamber under hot and cold
conditions to represent summer conditions in Houston, TX and winter conditions in the Dallas-Fort
Worth area. The test data we used in this analysis were the measurements taken after a twelve-hour
soak, where the vehicle had idled for at least one hour, and the vehicle had reached a 'stabilized'
idling condition. The vehicles were tested at the engine load required to run the heater or air
conditioning under the cold winter or hot summer conditions (see Table 2-55) but were not
commanded to be in the high idle state.
While the TTI tests included idling after different soak lengths and 'commanded high idle' for
engines capable of idling with an engine speed approximately 400 rpm higher than their standard
idle speed, we decided not to use the 'commanded high idle' emission rates for several reasons:
1) Six of the fifteen TTI trucks were not able to be commanded into high idle.
135

-------
2)	The 'stabilized' idling emission tests did contain some high idle that appears representative
of automatic engine control strategies for 2007 and later trucks. Two of the trucks included
high idle during the winter stabilized tests due to automatic engine control strategies. We
assume that for 2007 and later technology trucks, operators and manufacturers rely on
automatic engine control strategies rather than the vehicle operators to employ high idle
conditions". Because most of the engines did not use high engine speeds to power the
heater/air conditioner during the winter/summer conditions, we assume this engine
operation of MY 2007 and later trucks is also representative of in-use operation.
3)	The emissions impact of "commanded" high idle versus stabilized idle was not as
pronounced as observed in the pre-2007 trucks. For the TTI study, the high idle NOx rates
were only -36 percent higher than the stabilized emission rates. By using the stabilized
emission rates, we are using emission rates that are not much different than the
"commanded" high idle emission rates.
For these reasons, the summer and winter stabilized conditions were deemed to be the best estimate
of real-world extended idle emissions. The 'stabilized' idle emission rates (g/hr) for the winter and
summer conditions, are reported in Figure 2-74 through Figure 2-78.
Table 2-55. Ambient Test Conditions for the TTI Extended Idle Tests
Test ID
Temperature
Relative Humidity
Auxiliary Load
Hot (Summer)
100 °F (37.8 °C)
70%
Air conditioning
Cold (Winter)
30°F(-1.1°C)
N/A
Heating System
ARB tested five trucks on a chassis dynamometer on the ARB HHDDT 4-mode cycle, reporting
the g/hr results from the 10-minute 'Idle' mode. Before testing the 'Idle' mode, the vehicle was
first warmed on a pre-conditioning cycle, and then soaked for 10-20 minutes.91 Additional test
conditions were not reported by ARB, but we assumed that the ARB vehicles were tested at
moderate temperatures, with no auxiliary loading. Thus, we treated the ARB data as more
representative of an extended idling truck that did not require significant AJC or heating system
auxiliary loading on the engine, where the extended idling occurred shortly after active driving by
the main engine.
2.3.2.2 Analysis
In developing the extended idle emission rates, we averaged the emission rate from each of the
tests, within model year ranges that represent engine and aftertreatment technology groups that
have similar impacts on extended idle emissions. Where possible, we used all 35 tests (15 trucks x
2 conditions = 30 TTI tests, and 5 ARB tests). Because there were more TTI tests, the average
within each model year group is weighted significantly towards the TTI tests. We chose to weight
uAs discussed earlier, our assumptions for pre-2007 trucks are different.
136

-------
each test equally, because we believe the TTI data are more representative of real-world extended
idle conditions, because they were tested with auxiliary loads at non-standard 'lab' temperatures.
The individual test results and the average emission rates by model year group are presented in the
following figures (Figure 2-74 to Figure 2-78). Within each figure, the tests are distinguished
according to the test condition - 'hot' and 'cold' conditions represent the tests from the TTI test
program; 'lab' test condition are the tests from the ARB test program. Additionally, we indicate if
the test was from a truck equipped with SCR or not, which we found was the most useful
aftertreatment classifier to determine engine model year groups.
For CO2, CO, and NOx, we do not model any increase in emissions to account for deterioration,
including tampering, of the engines or emission control systems, because we did not observe strong
effects of the emission control on the extended idle emission rates for these pollutants - the
aftertreatment technology (oxidation catalyst, selective catalytic reduction systems) may not be
fully functional during the extended idle conditions, due to lower exhaust temperature occurring at
extended idle. On the other hand, for THC and PM2.5 emissions, we adjust the model year group
emission rates to account for deterioration of the aftertreatment systems, as discussed in more detail
below.
Figure 2-74 displays the CO2 individual test results. No trend with respect to aftertreatment or
model year is observed (nor was one expected). The emissions from cold tests tend to be higher
than the hot tests, which are both higher than the ARB laboratory tests. Two of the cold tests have
extended idle emission rates > 10,000 g/hr which is likely due to higher engine rpm for these
engines during the cold tests. TTI observed that some engines have an engine control strategy,
termed "cold ambient protection," which increases the idle engine speed at cold temperature to
warm the coolant temperature and protect against engine wear. We calculated an average CO2
extended idle emission rate for all 2007 and later trucks by using all the data and treating each test
equally across all model years.
The CO2 extended idle emission rate is used to derive the energy and fuel consumption extended
idle rate of 97,084 kJ/hr and 0.71 gallons-diesel/hr, respectively. We used the conversion factor of
0.0736 g C02/kJ and 10,045 g C02/gallon from B3.4 biodiesel (3.4% percent biodiesel blend)
highway diesel reported from the MOVES GHG and Energy Report.3
137

-------
C02
15000
a)
^10000
c
o
w
w
'E
CD
0)
2 5000



A
<
T
A
A
~
A
condition

•
hot

•
cold

•
lab
SCR

•
0

A
1
2006
2008
2010
2012
Figure 2-74. CO2 Emission Rates from the TTI and ARB Programs by Engine Model Year, and Average
Emission Rate (line) based on all the data.
Note:
Within "condition," "hot" refers to the summer conditions from the TTI tests, "cold" refers to the winter conditions
from TTI, and "lab" refers to the laboratory tests conducted by ARB. For SCR, 0 means the truck does not have a
selective catalytic reduction system (SCR), and 1 means the truck has SCR.
Figure 2-75 displays the CO individual test results. No trend is observed with respect to model year
or use of aftertreatment. The laboratory ARB tests are lower than the TTI tests, which could be due
to the lower fuel consumption of the tests. The CO emission rate is slightly lower than the emission
rate for 1990-2006 MY of 55 g/hr. Similar to CO2, a single average emission rate is calculated for
all the tests results and is applied to all 2007 and later model years.
138

-------
CO

t
~
2006
2008
2010
2012
2014
condition
•
hot
•
cold
•
lab
SCR
•
0
~
1
Figure 2-75. CO Emission Rates from the TTI and ARB Programs by Engine Model Year, and Average
Emission Rate (line) Based on All the Data
Figure 2-76 displays the NOx individual test results. We initially expected the data to show a
decrease in the extended idle emission rates beginning in MY 2008 to account for the California
Clean Idle Certification (all MY 2008 and later trucks were clean-idle certified). However, no
reduction was observed. We also expected to observe a decrease in 2012, with the full
implementation of SCR, but this was also not the case. Therefore, we calculated average NOx
emission rates for two model year groups (2005-2009) and (2010-2026) as represented by a solid
line in Figure 2-76. The MY 2005-2009 rates calculated in this analysis are applied to the 2007-
2009 model years. Given the variability of the data, the 2007-2009 average rate of 100 g/hr
compares well to the MY 1991-2006 rate of 136 g/hr shown in Figure 2-79.
139

-------
NOx
250
200
150
3
(J)
£=
o
C/5
Uo
a)
~o
50


A






•
A


•
•
•
•

•
\ A
\ •
\:

¦
• A
J *
• A



condition
[•jhot
•	cold
•	lab
SCR
•	0
A 1
2006
2008
2010
2012
Figure 2-76. NOx Emission Rates from the TTI and ARB Programs by Engine Model Year, and Average
Emission Rates for 2005-2009 and the 2010-2012 Engine Model Years (lines)
Figure 2-77 displays the THC individual test results. The results are displayed with the SCR
aftertreatment, rather than according to the use of an oxidation catalyst aftertreatment. The use of
SCR corresponded better to THC emissions than the reported use of an oxidation catalyst. We
believe the SCR aftertreatment classification is a surrogate for the combined engine control and
aftertreatment system used with SCR equipped trucks that have a large impact on THC emissions.
For example, with the use of SCR, engines can be calibrated to run leaner, which reduces engine-
out THC emissions. Additionally, SCR systems rely on oxidation catalysts, or catalyzed DPFs to
convert NO to NO2, which also reduces the THC tailpipe emissions.
We calculated average emission rates for three model year groups 2005-2009, 2010-2012 and 2013
and later model years. The 2005-2009 model year vehicles include a combination of DPF and non-
DPF equipped trucksv and are used to represent the 2007-2009 emission rates in MOVES. The
2010-2012 represents DPF equipped trucks, with some penetration of SCR equipped trucks. The
model year group representing 2013 and later model years was developed because starting in 2013,
Navistar began certifying a heavy heavy-duty diesel (HHDD) engine equipped with SCR
aftertreatment. In 2014 and 2015, Navistar and all other engine manufacturers certified all their
HHDD engines equipped with SCR aftertreatment.92 Therefore, emission rate for the 2013+ model
year group was estimated by averaging the rates of all the SCR equipped trucks in the data set,
even though the dataset did not include any data on 2013 and later model year engines.
v The 2005-2009 THC rates here are ~3 times smaller than the MOVES3 THC rates for MY 1990-2006 derived in
Section 2.3.1.2, which may be due to the small sample size of overlapping model year vehicles (3 MY 2005-2006
trucks) in the TTI study.
140

-------
THC
I
v>
c
o
w
c/>
O)
condition
[•jhot
• cold
• lab
'E
(D
SCR
• 0
A 1
A
A
0-
A
A
A
2006
2008
2010
2012
2014
Figure 2-77. THC Emission Rates from the TTI and ARB Programs by Engine Model Year, and Average
Emission Rates for 2005-2009,2010-2012, and 2013+ (SCR only) Engine Model Years (lines)
Figure 2-78 displays the PM2.5 individual test results. The ARB tests reported zero emission or
"Not Reported due to PM collection failure" for the five ARB tests, and thus, only the TTI data was
used to develop the PM2.5 extended idle emission rates. For the same reasons provided for the THC
results, the use of an SCR-equipped engine and aftertreatment systems should also have a
significant impact on the PM2.5 emissions. Additionally, and as expected, the implementation of
diesel particulate filters starting in 2007 model year had a significant impact on the PM2.5
emissions.
We grouped the individual emission tests into four model year groups: 2005-2006 (pre-DPF),
2007-2009 (DPF, pre-SCR), 2010-2012 (DPF and phase-in of SCR) and 2013 and later model
years (SCR only). Because the MY 2005-2006 PM2.5 emission rates are significantly different than
the MY 2007-2009 emission rates, they are grouped separately. The 2005-2006 rates from this
study are not used to update the pre-2007 PM2.5 emission rates.w The other model years and
aftertreatment groups are used to estimate the MOVES emission rates for MY 2007 and later. As
for THC, we used the results from the 2010 and later SCR equipped trucks to calculate PM2.5
emission rate for the 2013 and later model year group.
w The MY 2005-2006 PM2 5 emission rates measured from the TTI data are only ~3 times higher than the MY 2007-
2009 PM2.5 rates, and roughly ~10 times smaller than the PM2 5 rates for MY 1990-2006 (2.5 g/lir). We would expect a
larger decrease in PM2 5 emission rates with the use of DPF as discussed in Section 2.3.3. Differences could be due to
PM sampling methods, or variation in the truck emissions given the small sample size of 2005-2006 model year trucks
in the TTI study. As mentioned above, we did not update the pre-2007 PM2 5 emission rates in MOVES with
information from the TTI dataset due to limitations on time and resources.
141

-------
PM
0.5
0.4
3
 0.3
c
o
C/D
00.2
Q)
"O
0.1
0.0
. : I : .
condition

•
hot

•
cold

•
lab
SCR

•
0

~
1
2006
2008
2010
2012
2014
Figure 2-78. PM2.5 Emission Rates from the TTI Program by Engine Model Year, and average Emission Rates
Using for 2005-2006,2007-2009,2010-2012, and 2013+ (SCR only) Engine Model Years (lines)
PM2.5 emission rates in MOVES are composed of elemental carbon (EC) and non-elemental carbon
PM (nonEC). The TTI study measured total PM2.5 emissions, but not EC. We used the EC/PM
fractions from the sources listed in Table 2-56 to estimate the EC and PM2.5 emission rates.
Table 2-56. Baseline elemental carbon to PM2.5 fraction assumed for extended idling
Model Year


Group
EC/PM
Source


MOVES2014 Extended
Pre-2007
0.26
Idling35 x
2007-2009
0.10
ACES Phase I91
2010+
0.16
ACES Phase II93
2.3.2.3 Tampering and Mai-maintenance
As discussed in Section 2.3.1.2, we did not incorporate tampering and mal-maintenance effects on
the pre-2007 extended idle rates. For the 2007 and later extended idle rates, we incorporated the
effects of the effect of tampering and mal-maintenance (T&M) for two reasons:
1. The twenty vehicles used to estimate the extended idle emission rates did not appear to
include any tampered or mal-maintained vehicles with elevated emission rates. In addition,
14 of the 20 vehicles had odometer readings with less than 100,000 miles (Table 2-54).
x The pre-2007 EC/PM ratio for extended idling has subsequently been updated in MOVES3 to be 46.4% as discussed
in Section 2.3.1, but it was not updated for this analysis.
142

-------
2. The 2007 and later technology includes aftertreatment technology, including diesel
oxidation catalysts (DOC) and diesel particulate filters (DPF). We anticipate that the failure
of these after-treatment systems would significantly increase extended idle emissions if they
were tampered or mal-maintained.
We incorporated the T&M effects for extended idle exhaust using different data and methodology
than was used to derive the tailpipe exhaust emission rates for two reasons:
1.	Extended idle emissions in MOVES are stored in the EmissionRate table, and are not
distinguished by vehicle age, as the running and start exhaust emission factors. To fit the
current MOVES structure, we incorporated the effects of T&M into a single emission rate
by model year that applies to all vehicle ages.
2.	We are less confident in the application of the emission effects of T&M failures estimated
for running emissions in Appendix B to extended idling emission. For example, we do not
think failure of selective catalytic reduction (SCR) aftertreatments systems should impact
extended idling NOx emission rates as much as running exhaust emissions, because the SCR
systems is not fully operational during long idling periods. Instead, we estimated the effects
of T&M on 2007 and later extended idle emissions using pre-2007 extended idle emissions
as surrogate values for 2007 and later extended idle emission with failed aftertreatment
systems.
As shown in the figures above, the THC and PM2.5 emissions showed the largest reductions in
extended idle emissions with newer model year vehicles. We believe that the reductions are due
primarily to the continued effectiveness of the catalyzed diesel particulate filter even during
extended idling conditions. For the MOVES extended idle THC and PM2.5 emission rates, we
included an estimate of the impact of deterioration and failure of the diesel particulate filters in
calculating the 2007-2009, 2010-2012, and 2013+ model year group emission rates as discussed in
Appendix C, and displayed in Table 2-57. As shown, the MOVES EC/PM emission rates for MY
2007+ trucks are slightly higher than the 'Baseline' EC/PM fractions in Table 8-2, because the fleet
emissions are assumed to include some emission contribution from trucks with failed DPFs, which
have a higher EC/PM fraction.
Table 2-57. Extended Idle Emission Rates for 2007 and Later Model Year Heavy-Duty Vehicles
Model Year
Group
C02
fe/hr)
CO
fe/hr)
NOx
fe/hr)
THC
fe/hr)
PM2.5
fe/hr)
EC
fe/hr)
nonEC
fe/hr)
EC/PM
2007-2009
7151
39.3
100.5
8.5
0.087
0.012
0.076
0.13
2010-2012
7151
39.3
42.6
2.7
0.034
0.006
0.028
0.18
2013+
7151
39.3
42.6
1.6
0.021
0.004
0.017
0.20
2.3.2.4 MHD Regulatory Class
The extended idle emission rates for MHD are assumed to be the same as HHD for the following
two reasons. First, MHD trucks are estimated to account for only five percent of long-haul
143

-------
combination trucks in the US and therefore, they are a minor contributor to the emissions from
extended idling trucks. Second, Khan etal. 200994 evaluated extended idle emission rates of pre-
2007 MHD engines and did not observe a pronounced difference in extended idle emission rates
between MHD and HHD trucks. Taken together, these imply that any difference in emissions
modeled with unique MHD extended idling rates would be minimal, so without any extended
idling data on 2007 and later model year MHD trucks, we felt it was most defensible to keep the
MHD emission rates the same as the HHD emission rates.
2.3.3 202 7-2060 Model Years
For MY2027 and later vehicles subject to HD2027 standards, we anticipate that reductions in the
HHD and MHD NOx extended idle emissions rates will be driven by the idle standard, rather than
the duty-cycle standards in the rule. The duty-cycle standards do not contain high duration
extended idling (> 1 hour) that is representative of truck hotelling activity. We did not estimate any
change in extended idle emission rates due to the lengthened warranty or useful life periods
because MOVES extended idle rates do not vary by age.
First, we estimated extended idle emission rates that would comply with the off-cycle NOx/CCh
g/kg standard calculated in Table 2-13. We then used Equation 2-25 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 2-58.
Table 2-58 Calculation of HHD and MHD Extended Idle NOx g/hr Emission Rates for MY2027+
Model Year
Group
MOVES
Extended
Idle Rates
CO2 (kg/hr)
Idle Standard
(g/hr)
Idle
Standard
NOx/COi
(g/kg)
Idle-
standard
compliant
NOx
emission
rate
(g/hr)
2027-2028
7.191
9
1.17
8.42
2029+
7.191
8.7
1.13
8.14
2.3.4 Model Year Trends
Figure 2-79 through Figure 2-82 illustrate the extended idle emission rates in MOVES4 for
regClassIDs 46 and 47.
As shown, the NOx and the CO extended idle emission rates have a relatively small decrease
between the pre-2007 and the 2007+ model years. For THC and PM2.5, we observe large decreases
starting in MY 2007, which is consistent with our understanding of the effect of diesel particulate
filters. We observed a decrease by -29 times in extended idle PM2.5 rates between the pre-2007 and
post-2007 extended idle rates corresponding to the implementation of the DPFs, which is consistent
with the -27 times decrease in PM2.5 running exhaust emission rates from PM2.5 certification data
144

-------
as discussed in Section 2.1.2.1.7. Extended idle rates for NOx decrease in MY 2027 due to the
HD2027 rule, but other pollutant rates remain unchanged.
100
3
O
re
0£
ju
4—*
X
HI
X
O
50 ¦
Reg Class
46-MHD67
•- 47-HHD8
•- 49-Gliders
1980
2000
2020
2040
Model Year
Figure 2-79. Extended Idle NOx Emission Rates for HHD and MHD Diesel Vehicles by Model Year
50
40
D
O
S 30
a>
•*-*
re
Q£
4)
¦o
X
UJ
o
X
20
10
Reg Class
46-MHD67
•- 47-HHD8
•- 49-Gliders
1980
2000
2020
2040
Model Year
Figure 2-80. Extended Idle THC Emission Rates for HHD and MHD Diesel Vehicles by Model Year
145

-------
I
Reg Class
46-MHD67
*¦ 47-HHD8
•- 49-Gliders
1980
2000
2020
2040
Model Year
Figure 2-81 Extended Idle CO Emission Rates for HHD and MHD Diesel Vehicles by Model Year
Reg Class
46-MHD67
*¦ 47-HHD8
•- 49-Gliders
0-
1980
2000
2020
2040
Model Year
Figure 2-82. Extended Idle PM2.5 Emission Rates HHD and MHD Diesel Vehicles by Model Yearv
y Glider emission rates are intended to be the same for all model years from 2000 through 2060. However, for EC and
non-EC PM, the pre-MY 2007 values were updated without changing the post-MY 2007 values. We plan to update the
MY 2007 and later EC and non-EC PM rates for gliders in future versions of MOVES. Glider emissions prior to model
year 2008 are not used in the MOVES model since the gliders do not exist in the fleet.
146

-------
2.3.5 Extended Idle Energy Rates
The pre-2007 extended idle energy emission rates are unchanged from those originally developed
for MOVES2004 and are documented in the Energy and Emissions Report71, and are displayed in
Figure 2-83. The extended idle energy consumption rates are the same for regulatory class MHD
and HHD diesel vehicles. The extended idle energy rates for 2007+ trucks were updated in
MOVES3 and estimated using the CO2 emission rates presented in Table 2-57 and are also plotted
in Figure 2-83. The extended idle energy consumption rates are the same for regulatory class MHD
and HHD diesel vehicles.
1,5e+05
1.0e+05
CD
4—'
ns
i_
c
o
'(/)
(J)
'e

-------
Table 2-59: Projected APU Use during Extended Idling for Combination Long-Haul Tractor-Trailers
Vehicle Type
Model years
Diesel APU
Penetration
Battery APU
Penetration
Combination
Long-Haul
Trucks
2010-2020
9%
0%
2021-2023
30%
10%
2024-2026
40%
10%
2027+
40%
15%
The APU emission rates in MOVES are based on two studies that measured in-use APU emission
rates. The Texas Transportation Institute (TTI, 2014)95 tested two diesel APU systems with and
without diesel particulate filters at ambient temperatures of 100°F and 0°F. The exhaust emission
rates (THC, CO, CO2, and NOx) and the exhaust flow rates were measured using an ECOSTAR
gaseous portable emission measurement system. The PM mass was measured using a BG-3 partial
flow dilution and filter sampling system. Limitations of the TTI study are discussed in the HD
GHG Phase 2 MOVES documentation.101'2
The second study used to update APU emission rates was by Frey and Kuo (2009),96 who tested
two APU systems (APU ID 2 and 3), equipped with 2006 Kubota Z482 engines. The APU systems
were tested at a range of electric output loads to obtain the fuel consumption relationship with the
electric power demands, and the fuel-based emission rates. The study measured the in-use APU
electric loads from a fleet of 20 vehicles (10 trucks equipped with each APU system) for over a
year. They then used the relationship between electric power demand and the fuel-based emission
factors with the average energy use of the APU system to estimate average APU (g/hr) emission
rates of CO2, CO, NOx, THC, and PM for both a mild temperature (50-68 °F) scenario and a high
temperature (100 °F+) scenario. Frey and Kuo 2009 reported a PM emission rate, but the emission
rate is 'inferred from the literature' because their PM measurements were semi-qualitative.
An additional two studies were used as a source of data to compare and evaluate the APU emission
rates obtained from the studies mentioned above. TTI 201297 conducted testing of two APU
systems using their environmental chamber at both 100°F and 0°F. The APU systems (APU 4 and
5) manufacturer, engine make and model year were maintained confidential in the report. Storey et
al. 200398 tested a Pony Pack APU System (APU ID 6), equipped with a Kubota Z482 engine, in an
environmental chamber at both 90 °F and 0°F. This is one of the studies used by Frey and Kuo
200996 to determine the PM emission factor for the APU's tested in their study. The engine year,
engine displacement, and engine power were not reported in the TTI 2012 and Storey et al. 2003
studies. For this reason, these studies were used only as comparative data sets.
z Problems in testing meant only one of the APU systems could be used. Additionally, PM composition (EC/PM
fraction) was measured on tests with errors in the exhaust flow measurement. The PM emission rates determined
invalid for these tests were excluded and repeated, but the PM composition measurements from these tests were
considered valid and were not repeated.
148

-------
Table 2-60. APU Engines and Studies Used in This Analysis
APU
ID
Engine Model
Engine
Year
Displacement
(L)
Power
(HP/kW)
Tier
Study
1
Kubota Z482
2011
0.48
14.2/11
Tier 4
TTI201495
2
Kubota Z482
2006
0.48
10.9/8.1
Tier 2
Frey and Kuo
20 0996
3
Kubota Z482
2006
0.48
10.9/8.1
Tier 2
Frey and Kuo 2009
4
Confidential Information
TTI201297
5
Confidential Information
TTI2012
6
Kubota Z482




Storey et al. 200398
Table 2-61 contains the in-use emission rates measured from reviewed APU systems. As shown,
the emission and fuel rates for the APUs measured in the TTI 2014, and Frey and Kuo 2009 (APU
ID 1, 2 and 3) compare well with the APU emission rates reported from TTI 2009 Storey et al.
2003 (APU ID 4, 5, and 6). The impact of the DPF is clearly shown on the PM emission rates from
APU ID 1, as expected. However, there does not appear to be a substantial impact of the DPF on
the gaseous emissions (CO2, CO, NOx, and THC). Additionally, no notable emission effects are
observed with respect to the nonroad emission standard tier or engine model year.
The impact of ambient temperature can be observed within individual studies. For APU ID 2 and 3,
the CO2, and fuel consumption are higher at the hot ambient temperatures compared to the mild
conditions, which is expected. However, there is no consistent trend between hot and cold
conditions, when the APU is required to either cool or heat the tractor cabin. For APU ID 1 and 4,
the cold temperatures had higher CO2 emissions and fuel use. For APU ID 5 and 6, the hotter
temperatures had higher CO2 emissions and fuel use.
For CO, NOx, THC, and PM there are conflicting trends with respect to ambient temperature. For
APU 2 and 3, NOx and PM emissions are higher at the hot conditions compared to mild conditions,
consistent with the higher fuel use. However, CO shows lower emissions at hot conditions, and
THC shows a mixed trend. For the other studies, there is no consistent trend between the hot and
cold conditions.
149

-------
Table 2-61. In-Use APU Emission Rates
APU
C02
CO
NOx
THC
PM
Fuel
Ambient
Temperature
DPF
ID
fe/hr)
fe/hr)
fe/hr)
fe/hr)
fe/hr)
feal/hr)
condition
(°F)
present
1
4340
7.3
18.6
1.35
0.96
0.43
Cold
0
No
1
4270
5.1
20.0
0.73
0.02
0.43
Cold
0
Yes
1
2820
6.2
23.5
1.35
0.56
0.29
Hot
100
No
1
2800
5.2
23.7
1.52
0.03
0.28
Hot
100
Yes
2
3000
20.4
6.3
1.4
1
0.3
Mild
60a
No
3
2500
7.2
13.4
1.3
0.8
0.25
Mild
60
No
2
3900
13.9
11.5
1.5
1.3
0.38
Hot
100
No
3
3600
6.3
20.2
1
1.2
0.36
Hot
100
No
4
3100
5.8
19
1.3
1.23
0.3
Hot
100
No
5
3600
7.3
24
0.8
0.58
0.35
Hot
100
No
4
4000
3.9
22
1.2
0.75
0.39
Cold
0
No
5
2800
24
14
2.4
0.98
0.28
Cold
0
No
6
2146
25
8.7
7.8
0.48
0.22
Cold
0
No
6
2351
10.8
11.4
4.2
1.00
0.24
Hot
90
No
Note:
aFrey and Kuo 2009 report the mild condition for auxiliary loads on the trucks is for ambient temperatures ranging
from 10-20 °C (50-68 F)
Because the only notable trend in the APU emissions data was the large decrease in PM emission
rates with the use of a DPF, we developed "no DPF" baseline MOVES emission rates using the
"no DPF" results from TTI, 2014 and Frey and Kuo, 2009 (APU ID 1, 2, and 3). We first averaged
the emission rates within the cold, hot, and mild conditions as shown in Table 2-62.
Table 2-62. Average APU Emission Rates from non-DPF APU IDs 1,2, and 3 according to Cold, Hot, and Mild
Ambient Conditions
C02
CO
NOx
THC
PM
Fuel
Ambient
Temperature
DPF
fe/hr)
fe/hr)
fe/hr)
fe/hr)
fe/hr)
feal/hr)
condition
(T)
present
4340
7.27
18.59
1.35
0.96
0.43
Cold
0
No
3440
8.80
18.41
1.28
1.02
0.34
Hot
100
No
2750
13.80
9.85
1.35
0.90
0.28
Mild
60
No
Next, we calculated a fleet-average APU emission rate. Similar to our treatment of the extended
idle emission rates, we equally weighted the different ambient conditions. For APUs, we weighted
each ambient condition (Cold, Hot, and Mild) equally in developing the fleet-average emission rate
shown in Table 2-64.
We estimated elemental carbon (EC) fraction of PM from composition measurements made on
APU ID 1 as reported in Appendix J. For each test, we calculated the elemental carbon/total carbon
ratio, and then averaged the ratio across all cold and hot tests, separately for the DPF and the non-
DPF tests as shown in Table 2-63. We assumed that total carbon (TC) is a reasonable
approximation of the total PM2.5 emissions from the APU, and we used the EC/TC ratio from the
non-DPF tests as the source of the EC/PM fraction to derive the EC and nonEC emission rates in
Table 2-64.
150

-------
Table 2-63. Average Elemental Carbon/Total Carbon Ratio for APU ID 1 without and with a Diesel Particulate
	Filter (DPF)	

EC/TC ratio
APU 1 non-DPF
0.138
APU 1 DPF
0.073
Table 2-64. Fleet-Average Non-DPF Equipped APU Emission Rates in MOVES
C02
(g/hr)
CO
(g/hr)
NOx
(g/hr)
THC
(g/hr)
PM25
(g/hr)
EC
(g/hr)
NonEC
(g/hr)
EC/PM
Fuel
(gal/hr)
3510
10.0
15.6
1.3
0.96
0.13
0.83
0.14
0.35
The HD GHG Phase 2 rule implements a phase-in standard that requires APUs installed in new
tractors to meet lower PM standards from MY 2021 through MY 2024 (beyond the Tier 4 nonroad
standards).99 The APU PM standards along with the current Tier 2 and Tier 4 nonroad standards for
nonroad diesel engines 8
-------
2-57). We used the EC/PM split measured from the DPF-equipped APU (Table 2-63) to estimate
the EC and nonEC emission rates.
From the in-use testing of APU ID 1, we did not observe a meaningful impact on the CO2, CO,
NOx, and THC emissions with the use of the DPF. Thus, for the model year 2024 and later APUs,
we maintained the same emissions rates as were used in the 2010-2020 model year group. The
emission standard adjusted APU emission rates by model year group are shown in Table 2-66.
Table 2-66 APU Emission Rates in MOVES with APU PM Controls in the HP GHG Phase 2 Program
Model Year
C02
CO
NOx
THC
PM2.5
EC
NonEC
EC/PM
Fuel

(g/hr)
(g/hr)
(g/hr)
(g/hr)
(g/hr)
(g/hr)
(g/hr)

(gal/hr)
2010-20201
3510
10.0
15.6
1.3
0.96
0.13
0.83
0.14
0.35
2021-2023
3510
10.0
19.5
1.3
0.32
0.044
0.28
0.14
0.35
2024-2050
3510
10.0
15.6
1.3
0.021
0.0015
0.019
0.073
0.35
Note:
1 The default APU allocation in MOVES assigns APU usage beginning in model year 2010. If MOVES users specify
APU usage in years previous to 2010, it will use the 2010-2020 APU emission rate.
152

-------
2.5
Glider Vehicle Emissions
"Glider vehicles" or "Gliders" refer to vehicles with old powertrain (engine, transmission and/or
rear axle) combined with a new chassis and cab assembly. Most gliders are Class 8 heavy heavy-
duty vehicles. They typically use model year 2001 or older remanufactured engines that do not
have to use emissions controls such as DPF or SCR needed to meet the stringent PM and NOx
standards starting MY 2007+. 102
Starting with MOVES3, we model the emission impacts of the glider vehicles as a separate
regulatory class (regClassID 49) because their population became significant starting with model
year 2008 as described in the Population and Activity Report.6
For modeling purposes, all glider vehicles are presumed to be combination trucks (sourceTypelD
61 and 62) running on diesel fuel. EPA's in-house glider vehicle emission testing data103 suggest
that glider emissions have similar THC, NOx, PM2.5, and CO2 running exhaust emission rates to the
MOVES model year 2000 heavy heavy-duty vehicles (regClassID 47), while CO from glider
vehicles is higher. Based on this analysis, the MOVES running, start, and extended idling exhaust
rates for gliders of all model yearsaa are set equal to those of the model year 2000 heavy heavy-duty
vehicles.
For example, Figure 2-84 shows a comparison of the running exhaust emission rates (for age 0-3
group) of regClass 47 (heavy heavy-duty) vs. regClass 49 (glider vehicles) for selected pollutants
and model year groups. The rates for the two regulatory classes are identical for model year 2000.
For later model years, however, the emissions rates for regular heavy heavy-duty vehicles are
significantly lower due to more stringent emission standards, whereas the rates for glider vehicles
stay the same at the model year 2000 levels.
The auxiliary power unit (APU) exhaust emission rates of the glider vehicles, on the other hand, are
set equal to those of regular (non-glider) heavy heavy-duty vehicle fleet. This is consistent with our
assumption that glider vehicles have the same vehicle characteristics as regular heavy heavy-duty
vehicles for non-powertrain components, and thus, have the same APU, aerodynamics, rolling
resistance, brake and tire wear.
aa Glider emission rates are intended to be the same for all model years from 2000 through 2060. However, for EC and
non-EC PM, the pre-MY 2007 values were updated without changing the post-MY 2007 values. As discussed in the
activity report, glider activity in MOVES begins in model year 2008.
153

-------
regClassID B47B49
3000-
2000-
1000-
o-
s
-2 ioo-
ro
m
w 50-
03
DO
TO
d)
E 30-
20-
10-
0-
.. __ ¦¦ II
MY2000
nil IIIMLIIII
II II II II .. II II II
.. ¦¦ .. II II II II .. II II II
MY2008
.. ¦! il ll ll ll .. ¦! il ll ll ll I
¦¦ ll I
.. II II II I
.nil...ill
., 11
.. II II II I
J.LLjjjjJ
..I .I J J.
MY2010+
_¦ .1 ll ll ll ll .. .1 .1 all ll ¦
. ... I I I I . I I I l
	I I I I ..ill l
.ill
.III
t—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—r
opModelD
Figure 2-84. Comparison of the running exhaust emission rates (0-3 age group) of HHD (regClassID 47) vs. Gliders (regClassID 49) for selected
pollutants (NOx, ECPM, NonECPM) and model year groups
154

-------
3 Heavy-Duty Gasoline Exhaust Emissions
The discussion of heavy-duty gasoline vehicles first covers running exhaust emissions (Section
3.1), followed by start emissions (Section 3.2). Within each emission process, we discuss the
derivation of the emission rates by pollutant and model year group. As gasoline-fueled vehicles are
a small percentage of the heavy-duty vehicle fleet, the amount of data available on their emissions
is more limited.
3.1 Running Exh aust Emissions
3.1.1 THC, CO and NOx
The heavy-duty gasoline running rates were analyzed in three stages. The MY 1960-2007 emission
rates were originally developed in MOVES2010. In MOVES2014, we updated the MY 2008-2009
heavy-duty gasoline rates to account for the Tier 2 and 2007 heavy-duty rulemakings. In MOVES3,
we updated the MY 2010-2026 emission rates based on the more recent testing data. In MOVES4,
we revised the MY2027+ emission rates to account for the impact of HD2027 rule. The analysis of
PM2.5 running exhaust emission rates are discussed separately in Section 3.1.2 because it used
separate data and analyses than for the gaseous pollutants.
3.1.1.1 1960-2007Model Years
The heavy-duty gasoline emission rates for model year 2007 and earlier were carried over from
previous versions of MOVES. They are based on analysis of four medium heavy-duty gasoline
trucks from the CRC E-55 program and historical data from EPA's Mobile Source Observation
Database (MSOD)104, which has results from chassis tests performed by EPA, contractors and
outside parties. The heavy-duty gasoline data in the MSOD is mostly from pickup trucks which fall
mainly in the LHD2b3 regulatory class. Table 3-1 shows the total number of vehicles in these data
sets. In the real world, most heavy-duty gasoline vehicles fall in either the LHD2b3 or LHD45
class, with a smaller percentage in the MHD class. There are very few HHD gasoline trucks now in
use.
Table 3-1 Distribution of Vehicles in the Data Sets by Model Year Group, Regulatory Class and Age Group
Model year
group
Regulatory
class
Age group
0-5
6-9
1960-1989
MHD

2
LHD2b3

10
1990-1997
MHD

1
LHD2b3
33
19
1998-2002
MHD
1

LHD2b3
1

Similar to the HD diesel PM, THC, and CO analysis described above, the chassis vehicle speed and
acceleration, coupled with the average weight for each regulatory class, were used to calculate STP
155

-------
(Equation 1-6). To supplement the available data, we examined engine certification data as a guide
to developing model year groups for analysis. Figure 3-1 shows averages of certification results by
model year.
30
25 -¦
20 -
-Q
K 15 '
o
o 10

I CO
! Nox
~ HC
if Hi

S £
I " --
-- -¦= ~
f * ir
1.4
1.2
- 1
-	0.8 h
Q.
-	0.6 £
o
o,»
0.2
- 0
-0.2
1990	1995	1990	1995	2000
Model year
2005
2010
Figure 3-1 Brake-Specific Certification Emission Rates by Model Year for Heavy-Duty Gasoline Engines
Based on these certification results, we decided to classify the data into the coarse model year
groups listed below.
•	1960-1989
•	1990-1997
•	1998-2007
Unlike the analysis for HD diesel vehicles, we used the age effects present in the data itself. We did
not incorporate external tampering and mal-maintenance assumptions into the HD gasoline rates.
Due to the sparseness of data, we used only the two age groups listed in Table 3-1, and applied the
same age effects to all the heavy duty regulatory classes.
3.1.1.1.1 LHD
The emission rates for LHD (LHD2b3 and LHD45, regClassID 41 and 42, respectively) were
analyzed by binning the emission measurements using the STP with a fixed mass factor of 2.06
(Table 1-3). Figure 3-2 shows all three pollutants vs. operating mode. In general, emissions follow
the expected trend with increasing STP, though the trend is most pronounced for NOx. As expected,
NOx emissions for light heavy-duty gasoline vehicles are much lower than for light heavy-duty
diesel vehicles.
156

-------
4000"
30001
2000 ~
1000 -
o -
, 150-
™ 100
c
.2 50"
(/>
CO
o-
500 ~
4 A ~	A
, i I ^ "4 1 4
itt
iuH *
400-
300-
200
100 -
0-
• A
A • • • 1
. i • i 1
iPm.
• A
i A .
* ~ *

~
~

*
* ¦ ¦ 1
« ¦ 1
~ *
Model.Years
~ 1960-1989
A 1990-1997
¦ 1998-2007
1—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—r
0 1 11 1213141516212223242527282930333537383940
opModeD
Figure 3-2. Emission Rates by Operating Mode for MY Groups 1960-1989,1990-1997, and 1998-2007 at Age 0-3
Years for LHD2b3 and LHD45 Vehicles
Table 3-2 displays the multiplicative age effects by operating mode for LHD gasoline vehicles. The
relative age effects are derived from the sample of vehicle tests summarized in Table 3-1. The
multiplicative age effects are used to estimate the aged emission rates (ages 6+) years from the base
emission rates (ages 0-5) for THC, CO, and NOx. These multiplicative age effects apply to all
model year groups between 1960 and 2007. As discussed earlier, we derived multiplicative age
effects from the pooled data across the three model year groups and regulatory classes due to the
limited data set.
157

-------
Table 3-2 Relative Age Effect on Emission Rates between Age 6+ and Age 0-5 for LHD Gasoline Vehicles in
Model Years 1960-2007
OpModelD
THC
CO
NOx
0
2.85
1.45
1.67
1
2.43
1.79
1.85
11
3.12
1.66
1.88
12
2.85
2.05
1.69
13
3.55
2.68
1.48
14
3.43
2.84
1.46
15
3.37
3.03
1.26
16
3.76
3.88
1.06
21
2.78
1.67
1.42
22
2.64
1.64
1.36
23
2.96
1.67
1.32
24
2.83
1.62
1.21
25
3.23
2.79
1.43
27
3.21
3.20
1.21
28
3.20
4.04
1.11
29
3.00
3.90
1.05
30
2.55
2.56
1.05
33
1.95
2.00
1.77
35
2.67
2.20
1.59
37
2.80
2.24
1.42
38
2.46
2.06
1.34
39
2.46
2.30
1.27
40
2.47
2.59
1.17
Figure 3-2 illustrates the emissions trends by age group for the 1998-2007 model year group. Since
we did not use the tampering and mal-maintenance methodology as we did for diesels, the age
trends reflect our coarse binning with age. For each pollutant, only two distinct rates exist - one for
ages 0-5 and another for age 6 and older.
158

-------
2500
2000-
1500 "
1000
500 -
o-
I 100-
I
» 75-
)
¦ 50
; 25-
0-
200-
150-
100 -
50 J
o-
A A
A A
A A
~ . *
• * i
A A A
• + • + •
~ ~ ~ •
A ~
~ i * *
~
~	A
~ ~
~ ~
* i
A
~
1 * ~
• ~
A A
A
A •
~ *
- A
A * *
*	A »
•	*
A ~ ~
A A t *
• *
age
A 6+
~ 0-5
t—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—r
0 1 11 12 13 14 1 5 1 621 22 23 24 25 27 28 29 30 33 35 37 38 39 40
opModelD
Figure 3-3. Emission Rates by Operating Mode and Age Group for MY 1998-2007 Vehicles in Regulatory Class
LHD2b3 and LHD45
3.1.1.1.2 MHD and HHD
Like the LHD rates described above, the 2007 and earlier MHD and HDD gasoline rates are based
on emissions data from the mix of LHD2b3 and MHD vehicles outlined in Table 3-1. The same
model year groups were used to classify the emission rates: 1960-1989, 1990-1997, and 1998-2007.
Also, we used the same relative increase in emission rates for the age effect. The only difference
from the analysis of LHD emission rates is that the regulatory class MHD and HHD emission rates
were analyzed using STP operating modes with a fixed mass factor of 17.1. The resulting MHD
and HHD emission rates for THC, CO, and NOx for each model year group are presented in Figure
3-4.
159

-------
30000
20000i
10000
0
600
3
o
.c
3 400
CD
i_
C
.2 200
to

-------
Table 3-3 Relative Age Effect on Emission Rates between Age 6+ and Age 0-5 for MHD and HHD Gasoline
Vehicles in All Model Years 1960-2060
OpModelD
THC
CO
NOx
0
2.85
1.45
1.67
1
2.43
1.79
1.85
11
3.12
1.66
1.88
12
3.36
3.12
1.13
13
3.53
3.16
1.11
14
3.53
3.16
1.11
15
3.53
3.16
1.11
16
3.53
3.16
1.11
21
2.78
1.67
1.42
22
3.08
2.59
1.23
23
2.97
3.31
1.05
24
1.80
1.54
1.03
25
1.80
1.54
1.03
27
1.80
1.54
1.03
28
1.80
1.54
1.03
29
1.80
1.54
1.03
30
1.80
1.54
1.03
33
2.45
2.41
1.33
35
2.16
2.41
1.19
37
2.16
2.41
1.19
38
2.16
2.41
1.19
39
2.16
2.41
1.19
40
2.16
2.41
1.19
Figure 3-5 displays the resulting emission rates by operating mode bin and age group for the
LHD45, MHD, and HHD gasoline vehicles, which were calculated by applying the multiplicative
age effects in Table 3-3.
161

-------
3
o
.c

-------
Table 3-4 Useful Life FTP Standards from the Tier 2 Rulemaking105 and the HP 2007 Rule107

MDPV
8.5k- 10K
10K-14K

(Tier 2 Bin 5)
(Class 2B)
(Class 3)
Units
g/mile
g/mile
g/mile
Fully Phased in MY
2009
2009
2009
THC
0.09 NMOG
0.195 NMHC
0.230 NMHC
CO
4.2
7.3
8.1
NOx
0.07
0.2
0.4
This section documents the THC, CO and NOx emission rates for regulatory class LHD2b3
vehicles in model years 2008 and 2009. In conducting this analysis, we lacked any modal data on
LHD vehicles and therefore, we ratioed the modal emission rates measured from light-duty
vehicles by the difference in standards.9 By MY 2008, the certification results demonstrated that
LHD2b3 were nearing the emission levels of light-duty vehicles certified to the Tier 2 Bin 8
standard.35 Consequently, we relied on the MOVES2014 analysis of in-use Tier 2 Bin 8 vehicles
conducted for the light-duty emission rates.9 We applied this analysis to derive MY 2009 emission
rates, then calculated MY 2008 rates by interpolating between MY 2007 and MY 2009.
Although the light-duty rates are based on VSP, rather than STP, adapting them for the LHD2b3
rates was deemed an acceptable approximation because the gasoline LHD2b3 gasoline vehicles are
chassis-certified to distance-based standards (g/mi). Accordingly, the vehicle emissions rates are
less dependent on the individual power and weight of the vehicle, and should scale approximately
to the g/mile emission standards.bb
Based on these assumptions, we scaled modal rates for Tier 2 Bin 8 vehicles by the ratio of FTP
standards to the calculated aggregate LHD2b3 standards documented in the MOVES2014 heavy-
duty exhaust report.35'cc Table 3-5 displays the aggregated LHD2b3 standards, Bin 8 FTP standard
and the ratio between the standards by pollutant.
Table 3-5 Aggregate LHD2b3 Standard Ratios against Bin 8 Modal Rates

Aggregate LHD2b3
FTP standard
Bin 8 FTP
standard
Aggregate
LHD2b3/Bin 8
NMOG
0.18
0.1
1.8
CO
7.49
3.4
2.2
NOx
0.22
0.14
1.6
We took an additional step to "split" these ratios into "running" and "start" components, such that
the running rates increased twice as much as the start rates, while maintaining the same simulated
value for the FTP composite. This split ratio is consistent with typical emission reduction trends,
where running emissions are reduced about twice as much as start emissions.9 The "split" ratios for
bb This approximation needs to be revisited in the future now that we have updated the mass of LHD vehicles in
MOVES3 to range from 3.5 to 7.8 metric tonsError! Bookmark not defined., which differs from the fSCau value of
2.06 metric tons.
00 As documented in MOVES2014 documentation, this analysis assumed that 5% of the gasoline LHD2b3 engines were
engine-certified, but, actually, all gasoline fueled LHD2b3 vehicles are chassis-certified. However, the engine-
certification standard has a small impact on the calculated aggregated standard conducted for MOVES2014.
163

-------
running and start, which were applied to the light-duty Tier 2 Bin 8 vehicle emission rates are
shown in Table 3-6.
Table 3-6 Ratio Applied to Light-Duty Tier 2 Bin 8 Emission Rates to Estimate Regulatory Class LHD2b3
Emission Rates for 2008-2009 MY

THC
CO
NOx
Running
2.73
2.73
1.95
Start
1.37
1.37
1.00
We also adopted the light-duty deterioration effects and applied them to the MY 2009 regulatory
class LHD2b3 (regClassID 41) emission rates. The light-duty emission rates have age effects that
change with each of the 6 age groups in MOVES, as shown in Table 3-7.
Table 3-7 Multiplicative Age Effect used for Running Emissions for Regulatory Class LHD2b32009 Model Year
ageGroupID
THC
CO
NOx
3
1
1
1
405
1.95
2.31
1.73
607
2.80
3.08
2.21
809
3.71
3.62
2.76
1014
4.94
4.63
3.20
1519
5.97
5.62
3.63
2099
7.20
6.81
4.11
After applying the steps described above (scaling the emission factors by ratio of FTP standards,
and applying light-duty deterioration trends), we restricted the scaled data so that the individual
emission rates by operating mode were never higher than MY 1998-2007 regulatory class LHD2b3
rates. This step essentially "capped" the emission rates, such that none of the modal rates for MY
2009 are higher than their counterparts for MY 2007 and earlier. MY 2008 rates are interpolated
between MY 2007 and MY 2009 emission rates as discussed later.
This final step "capped" the model year 2009 emission rates in the highest operating modes, as
shown in Figure 3-6.. For THC, emission rates in operating modes 28-30 and 38-40 were capped
for some or all age groups by the pre-2007 emission rates. For CO, emission rates in 12 of the 23
running operating modes (1, 16, 23-24, 27-30, 35-40) were capped by the pre-2007 rates. None of
the NOx emission rates were impacted by this step. Figure 3-6. shows the regulatory class LHD2b3
model year 2008-2009 emission rates for CO, THC, and NOx. In the figure, rates "capped" by the
pre-2007 rates exhibit the uncharacteristic "stairstep" deterioration trends. Even with the "capping"
effects, the rates for regulatory class LHD2b3 (regClassID 41) are higher than those for light-duty
trucks (regClassID 30), with a few exceptions. The few exceptions are some of the age-dependent
THC and or CO emission rates in operating modes 1, 30, 38, 39, and 40. However, the majority of
emission rates are considerably higher for the heavy-duty (LHD2b3) than for the light-duty trucks.
Similarly, when the FTP is simulated from the resulting rates, estimated composites are
substantially higher for LHD2b3 than for light-duty trucks.
The Light-duty Tier 2 standards shown in Table 3-4 phase-in at a rate of 50 percent in MY 2008
and are considered fully phased in by MY 2009.106 For estimating emission rates in MOVES, we
164

-------
used the same assumptions to estimate the MY 2008 emission rates. The MY 2008
emission rates are interpolated between the MOVES 2007 and 2009 emission rates
mode and age group.
2007
2008
2009
running
by operating
2500-
0-	
0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24
Vehicle age, years
Figure 3-6. Age Effects for CO, THC, and NOx Emission Rates for Regulatory Class LHD2b3 (regClassID 41)
Vehicles in Running Operating Modes for MY 2007,2008 and 2009
3
O
JZ
o>
d>
-«—»
(0
i—
c
o
{/)
(/)
£
LU
2000
1500
1000
500
0
100
75
50
25
0
200
150
100
opModelD
—
0
—
25
—
1
—
27
—
11
—
28
—
12
—
29

13

30

14

33
—
15
—
35
—
16
—
37
—
21
—
38
—
22
—
39
—
23
—
40
—
24


Due to limited data on LHD45 vehicles, we applied the LHD2b3 emission rates developed in the
previous section to the LHD45 emission rates. The LHD2b3 and LHD45 emission rates are
identical for model years 1960-2017.
3.1.1.2.1 MHD and HHD
Of the onroad heavy-duty vehicles GVWR Class 4 and above, a relatively small fraction are
powered by gasoline: about 15 percent are gasoline, as opposed to 85 percent diesel.dd The
percentage of gasoline-fueled vehicles decreases as the GVWR class increases. Since these vehicles
are a small portion of the fleet, there is relatively little data on these vehicles, and therefore, the
current 2008 and 2009 model year emission rates are from MOVES2010.107 The rates are modeled
by applying a 70 percent reduction to the MY 2007 running rates starting in MY 2008, which is
dd Negligible portions are run on other fuels. The figures are aggregated from data supplied by Polk.
165

-------
consistent with the emission standard reduction with the "Heavy-Duty 2007 Rule."108'66 The 2008
and 2009 model year emission rates have two age groups (0-5, and 6+) and the same relative
multiplicative age effects as the pre-2007 emission rates (Table 3-3).
3.1.1.3 2010-2060 Model Years"
In MOVES3, we updated the THC, CO, and NOx emissions rates for MY 2010 and later vehicles
for all gasoline heavy-duty regulatory classes. The initial update is described here. Additional
analysis to account for the Tier 3 and HD2027 rules are described in Sections 3.1.1.3.1 and
3.1.1.3.2 below. The updated rates are based on analysis of real-world PEMS-based emissions
measurement data from two engine-certified and one chassis-certified heavy-duty gasoline vehicles
(Table 3-8) with model years between 2015-2017. As explained in the PM2.5 section (Section
3.1.2.2), we also conducted chassis-dynamometer laboratory testing on these vehicles, but used the
PEMS gaseous emissions data because it better represents emissions in the real-world.
The Ford and Isuzu vehicles used the most popular engine configurations for recent model year
heavy-duty gasoline Class 4 vehicles. Each of the HD gasoline vehicles had three-way catalyst
(TWC) technology to control THC, CO, and NOx emissions. However, one key difference
compared to light-duty gasoline vehicle TWC configuration is that the engine-certified HD
gasoline vehicles do not use a close-coupled TWC. There might also be differences in catalyst
precious metal loading and in-cylinder combustion control for maximum TWC efficacy. The
reason for these differences is that engine-certified and chassis-certified gasoline spark-ignited
vehicles have to meet different standards.
Table 3-8 Summary of MY 2015-2017 Heavy-Duty Gasoline Vehicles with Real-World PEMS-based Emissions
Measurement Data
Vehicle
Engine
Test Weight
(lbs)
Cert-
ification
Make
Model
MY
Odometer
(miles)
GVWR
(lbs)
GCWR
(lbs)
Family
Displ
(L)
Low
High
Isuzu
NPR
2015
48,000
14,500
20,500
FGMXE06.0584
6.0
8,620
12,940
Engine
Ford
E450
2016
31,000
14,500
-
GFMXE06.8BWZ
6.8
9,320
13,080
Engine
RAM
3500
2017
32,000
13,300
19,900
HCRXD06.45W0
6.4
14,557
18,020
Chassis
The testing was conducted by US EPA over various test cycles in the Ann Arbor, Michigan area.
The test matrix covered a range of vehicle operation that included:
1.	Two idling tests of 15- or 30-minutes duration
2.	Seven on-road driving routes that cover the full range of power demand by including
transient low- and medium-speed urban driving to steady-state high-speed highway
driving
3.	Soak times ranging from zero minutes (hot start) to 720 minutes (cold start)
4.	Vehicle weight at low or high (Table 3-8)
ee The engine-certified standards for heavy-duty gasoline were reduced by 93% (THC), 80% (NOx), and 61% (CO)
with the MY 2008 standard.
ff The additional rate adjustments for MY2018-2060 LHD2b3 and MY2027-2060 LHD45/MHD/HHD vehicles are
described in subsections 3.1.1.3.1 and 3.1.1.3.2 respectively.
166

-------
5.	Air conditioning on or off
6.	Cabin windows down or up
A total of 202 tests across vehicles and operation modes was available for data analysis. These tests
covered about 412,000 seconds of post-QA operation. We removed the effect of warm and cold
starts from the operation since the running emissions are intended to be just the hot running
operation; details are discussed in Appendix 1.1. After removal of vehicle operation related to start
emissions, the final data set used for just the hot running emissions rates update was about 390,000
seconds.
The following steps were used to calculate the operating mode-based emission rates for each age
and regulatory class of LHD, MHD, and HHD:
1.	Assign operating modes as per the method described above for diesel vehicles and
calculate the average rate per operating mode per test per vehicle
2.	Calculate the average operating mode-based rate per vehicle (using only vehicle specific
tests)
3.	Estimate emission rates for operating modes with limited or missing data.
4.	Calculate the operating mode-based emission rate as the production weighted average of
the three test vehicles.
5.	Adjust emission rates by vehicle age.
In Step 1, the operating modes (Table 1-4) were assigned to the 1-hz data using the STP equation
(Equation 1-6) with road-load coefficients for single-unit short-haul truck (sourceType 52) for the
2014-2020 model year range as defined in the sourceusetypephysics table in MOVES3 database.
The coefficients for single-unit short-haul trucks are the same for all the regulatory classes within
this sourcetype (LHD2b3, LHD45, MHD, and HHD). The road-load coefficient values used are:
rollingTermA = 0.596526 [kW.sec/m]
rotatingTermB = 0 [kW.sec2/m2]
dragTermC = 0.00160302 [kW.sec3/m3]
For vehicle mass, we used the actual test weight (Table 3-8). Road-grade was not available, so it
was set to zero. The entire data set was analyzed with the new fscale values (Table 1-3) of 5
(LHD2b3 and LHD45), 7 (MHD), and 10 (HHD). The selection of these new fscale values was based
on the diesel HDIUT dataset and is described in Appendix G.
In Step 2, we averaged according to operating mode for each vehicle. In Appendix 1.2, we
compared the emission rates among the three vehicles by operating mode. Significant differences
are observed between the vehicles, however no consistent differences were noted across operating
modes and pollutants between the two engine-certified vehicles and the chassis-certified vehicle.
In Step 3, we estimated emission rates for high power operating modes with limited or missing data
from regulatory class MHD and HHD, due to the larger fscaie values used for these operating
modes. In these cases, we aggregated the data across the nearest high-power operating modes with
sufficient data, and set the emission rates to be equivalent across the aggregated bins. Additional
details and examples are discussed in Appendix 1.3.
167

-------
In Step 4, we calculated a weighted average of the emission rates from the three vehicles using the
production volumes of each of the tested engines. Ideally, the emission rates for each regulatory
class (LHD2b3, LHD45, MHD, and HHD) would be estimated from test data collected from
vehicles of that regulatory class, or estimated separately for the engine-certified (LHD45 and
heavier) and chassis-certified vehicles (LHD2b3). However, due to the small sample size
(including only one LHD2b3 vehicle), we used the same weighting of the three vehicles for all the
regulatory classes. The production volumes of the RAM 3500 vehicle are only a minor fraction of
the combined production of the Ford and Isuzu engine volumes. As such, the production weighting
is most representative of LHD45 emission rates.gg
Because we use the same production volume weighting for all the regulatory classes, the base
emission rates for MY 2010+ LHD2b3 and LHD45 are identical. However, the LHD2b3 rates are
further modified by applying the Tier 3 reductions phased-in from MY 2018 to 2022 (Section
3.1.1.3.1). The only difference between the LHD45, MHD, and HHD emission rates is the /scale
used to estimate the emission rate by operating mode, and methods used to estimate high-power
operating modes conducted in Step 4 (Details in Appendix 1.3).
In Step 5, we applied the MHD/HHD age effects shown in Table 3-3 to all gasoline heavy-duty
regulatory classes, including LHD2b3 and LHD45. We did not use the LHD2b3/LHD45 specific
age effects shown in Table 3-2. Both of these age effects tables are based on the same data set
(Table 3-1) with the difference being only the fscale used while assigning the data to operating
modes. Applying the LHD or LD (Table 3-7) age effects to rates developed using HD data and
different fscale ranges could over- or under-estimate the increases in emissions from aging. Ideally,
LHD2b3 emission rates and age effects would be derived from chassis-certified heavy-duty
gasoline vehicles.
3.1.1.3.1 LHD2b3 2018-2060 Model Years
The LHD2b3 vehicles are subject to the Tier 3 light-duty standards starting in MY 2018,hh To
calculate emission rates for MY 2018 and later, we applied reductions representing the Tier 3
phase-in for MY 2018-2022 for LHD2b3 vehicles (as shown in Table 3-9) to the emission rates
representing MY 2010-2017 estimated from the above. The reductions for each model year during
the phase-in were estimated by extracting the corresponding MOVES rates for MY 2007-2022, and
calculating the fractions relative to MY 2017. The basis and rationale for the Tier 3 reductions for
gasoline LHD2b3 vehicles developed for the Tier 3 rulemaking are documented in the
MOVES2014 heavy-duty exhaust report.109
The LHD2b3 MY 2018+ rates contain the same heavy-duty gasoline age effects as were applied to
the MY 2010-2017 rates (Table 3-3). The resulting emission rates for THC, CO and NOx are shown
in Figure 3-7 through Figure 3-9.
gg Sales of Class 2b gasoline trucks are much larger than for Class 3, 4, 5, and 6.
14 All LHD2b3 chassis-certified complete vehicles are subject to Tier 3. All LHD2b3 gasoline fueled vehicles are
chassis-certified complete vehicles.
168

-------
Table 3-9 Tier 3 Reductions by Model Year for Gasoline LHD2b3
Model Year
THC
CO
NOx
2018
35%
38%
41%
2019
44%
48%
52%
2020
53%
59%
63%
2021
62%
68%
74%
2022-2060
71%
78%
85%
3.1.1.3.2 LHD45, MHD, HHD 2027-2060Model Years
In order to account for the HD2027 standards in MOVES4, we revised the running exhaust
emission rates for NOx, THC, CO, (and PM2.5 described later in 3.1.2.3) for MY2027+ heavy-duty
(LHD45, MHD, HHD) gasoline vehicles using the methodology described in this section. Unlike
the HD2027 standards for diesel vehicles, the rule does not include off-cycle standards for gasoline
vehicles. Due to the relatively simple treatment of aging for MOVES HD gasoline vehicles, we did
not estimate any impact from the lengthened warranty and useful life periods provisions on the
emission rates.
The FTP duty-cycle standards shown in Table 2-9 apply to both heavy-duty compression-ignition
engines and heavy-duty spark-ignition engines. 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 3-10 shows the estimated
reduction in NOx emission rates, which is consistent with the ratio of the MY2010 FTP emission
standards and the HD2027 final FTP standards shown in Table 3-10.
In addition, we also estimated emission rate reductions due to the HD2027 standards for HC and
CO. 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 available data
from production HD Spark-ignition (SI) engines and from the heavy-duty gasoline technology
demonstration program to estimate our modeled emissions levels.110 We assumed a 65 percent
reduction in THC emissions would occur at a NOx NOx standard of 0.1 g/hp-hr. We assumed
additional decreases in THC emissions to reflect tighter final NOx standards in MY 2027. We
derived Equation 3-1 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 3-1 to estimate the reductions in THC emissions
using the NOx levels for the control scenarios (Table 3-10).
169

-------
^gasoline,THC,NOx FTP
= 1
NOx FTP Standard \ ^	^ Equation 3-1
g	1 X _ ^gasoline,THC,0.1 NOx FTP )
bhp ¦ hr J
* ( NOx FTP Standard\	^i-r,/\
= 1 - (	51=Z=	) x CI - 65%)
\	bhphr /
Where:
Rgasoiine.mc, NO,: /."//- = 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 3-10
NOx FTP Standard = HD2027 NOx FTP standards
We assumed a 60 percent reduction in CO for MY2027+ engines (see Table 3-10) based on EPA
testing.
Table 3-10 Running Emission Rate Reductions From Heavy-duty Gasoline Vehicles Due to HD2027 Standards,
Rgasoiine, Across All Heavy-duty Gasoline Regulatory Classes and Operating Modes
Regulatory
ClassA
FTP/SET
NOx
standard
(g/hp-hr)
NOx
THC
CO
LHD, MHD,
HHD
0.035
82.5%
87.8%
60%
A We applied the same standards to represent the SI engines modeled by the LHD, MHD, and HHD regulatory classes,
unlike the final standards for compression-ignition engines
Then Equation 3-2 was used to revise the MOVES emission rates to account for the HD2027
standards. Since spark-ignition engines are not subject to the HDIUT program, 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.
ERpinal Standards (l Rgasoiine) ^ ER.M0VES_baseline	Equation 3-2
Where:
FRcontroi = MOVES running exhaust emission rates for HD2027 standards based on the reduction in the FTP duty-cycle
standard
Rgasoiine = percent emission reductions in heavy-duty gasoline emissions from Table 3-10
ERmoves baseline = MOVES running exhaust emission rates before the rate revision
170

-------
3.1.1.4	Inspection and Maintenance Program Effects for LHD2b3 Gasoline Vehicles
In M0VES3.1, we updated the meanBaseRatelM values for THC, CO, and NOx running exhaust
emission rates to better reflect our understanding how state and local inspection and maintenance
(I/M) programs reduce gaseous emissions from LHD2b3 gasoline vehicles.
As background, in addition to the meanBaseRates described in Section 1, the MOVES
emissionRateByAge table includes a field, meanBaseRatelM, that is used to estimate emissions
under a relevant I/M program. These calculations are explained in more detail in the MOVES
Adjustments report.63
In MOVES3.0.4 and earlier versions of MOVES, we set the values of meanBaseRatelM for all HD
exhaust rates to equal the associated meanBaseRate—essentially assuming no benefit from HD I/M
programs. For MOVES3.1 and later, we reconsidered this choice for gasoline LHD2b3 trucks
(regClass 41). Given the similarity of the engine technology and aftertreatment systems between
gasoline LHD2b3 and light-duty trucks (regClass 30), and the similarity in the way these trucks are
tested in contemporary I/M programs, we updated the gasoline LHD2b3 HC, CO and NOx running
exhaust values for meanBaseRatelM to reflect the same proportional reduction (that is
meanBaseRatelM/meanBaseRate) that we model for each operating mode bin and age for the light-
duty trucks. For more detail on the meanBaseRate and meanBaseRatelM values for light-duty
trucks, see the MOVES3 LD report.
3.1.1.5	Model Year Trends
Figure 3-7 through Figure 3-9 display the THC, CO, and NOx non-IM running exhaust emission
rates by model year and regulatory class (HHD and Urban Bus). The emission rates are estimated
in grams per mile (g/mile) using nationally representative operating mode distributions and average
speeds. The model year groups used to estimate the emission rates are evident: 1960-1989, 1990-
1997, 1998-2007, 2008-2009, and 2010-2060. Note that not all the changes in the gram per mile
emission rates are due to changes in the operating mode specific emission rates. For example, the
MY 1995-1997 operating mode specific emission rates are the same as the 1990-1995 emission
rates for all regulatory classes. However, there is an observed spike in the HHD gram per mile
THC emission rate, which is attributed to a shift in the distribution of HHD gasoline activity among
different source types in MOVES.
Figure 3-7 shows that the THC emission rates follow decreasing trends with model year that
correspond with tighter emission factors. The drop in emission rates in model year 2008-2009 is
attributed to the different methodology used to develop those rates discussed earlier. Even though
the increasing trend between 2008-2009 and 2010 and later model year groups may not be intuitive
from a technical perspective, we have increased confidence that the MY 2010 and later THC
emission rates represent the real-world emissions since they were developed based on in-use testing
of MY 2010 and later vehicles.
Emission rates also change starting in MY2027 due to the implementation of the HD2027
standards.
171

-------
Reg Class
*- 41-LHD2b3
»- 42-LHD45
46-MHD67
•- 47-HHD8
1980
2000
2020
2040
Model Year
Figure 3-7. Base running emission rates for THCfrom age 0-3 gasoline heavy-duty vehicles averaged over a
nationally representative operating mode distribution.
The CO emission rates are shown in Figure 3-7 The CO emission rates for LHD2b3 vehicles (the
largest regulatory class of heavy-duty gasoline) follow a generally decreasing trend with model
year. The trends for LHD45 and MHD show unexpected variation across model years, including an
increase in CO emission rates for LHD45 and MHD vehicles. We have the most confidence in the
most recent model year data, and the variability in the model year trends reflects uncertainty in the
earlier heavy-duty gasoline emission rates.
172

-------
Reg Class
»- 41-LHD2b3
»- 42-LHD45
46-MHD67
•- 47-HHD8
1980
2000
2020
2040
Model Year
Figure 3-8. Base running emission rates for CO from age 0-3 gasoline heavy-duty vehicles averaged over a
nationally representative operating mode distribution.
Figure 3-9 shows that the NOx emission rates follow decreasing trends with model years that
correspond with tighter emission standards.
Reg Class
»- 41-LHD2b3
»- 42-LHD45
46-MHD67
¦- 47-HHD8
1980
2000
2020
2040
Model Year
Figure 3-9. Base running emission rates for NOx from age 0-3 gasoline heavy-duty vehicles averaged over a
nationally representative operating mode distribution
173

-------
3.1.2 Particulate Matter (PM2. s)
The available studies from which to develop PM2.5 emission for heavy-duty gasoline are
particularly limited. This includes limitations on second-by-second data from which to develop
operating mode specific rates, as well as studies representative of in-use and fleet average
emissions. At the same time, heavy-duty gasoline is a relatively small contributor to the total PM2.5
emissions inventory when compared to heavy-duty diesel and light-duty gasoline. As a result, the
limited analysis conducted for MOVES2010 has been carried over into MOVES3 for the 2009 and
earlier vehicles as discussed in Section 3.1.2.1. For MOVES3, we have updated the 2010 and later
model year heavy-duty gasoline emission rates to be based on heavy-duty diesel rates as discussed
in Section 3.1.2.2.
3.1.2.1 1960-2009 Model Years
For MOVES, the MY 1960-2009 heavy-duty gasoline PM2.5 emission rates were calculated by
multiplying the MOVES2010b light-duty gasoline truck PM2.5 emission rates by a factor of 1.40, as
explained below. Since the MOVES light-duty gasoline PM2.5 emission rates comprise a complete
set of factors classified by particulate sub-type (EC and nonECPM), operating mode, model year
and regulatory class, the heavy-duty PM2.5 emission factors are also a complete set. No change to
the PM emission rates is made between MY 2003 and 2009, because the HD 2007 Rule PM
standards are not expected to change in-use emissions for heavy-duty gasoline vehicles. As
presented in the next subsection, the simulated age 0-3 HD gasoline MY 1960-2009 emission rates
on the UDDS is -6.6 mg/mile, while the standard for 2008+ spark-ignition vehicles is 20
mg/mile108
3.1.2.1.1 Data Sources
The factor of 1.4 used to convert light-duty gasoline PM rates to heavy-duty rates was developed
based on PM2.5 emission test results from the four heavy-duty gasoline trucks tested in the CRC
E55-E59 test program. The specific data used were collected on the UDDS test cycle. Each of the
four vehicles in the sample received two UDDS tests, conducted at different test weights. Other
emission tests using different cycles were also available on the same vehicles but were not used in
the calculation. The use of the UDDS data enabled the analysis to have a consistent driving cycle.
The trucks and tests are described in Table 3-11.
Table 3-11 Summary of Data Used in HD Gasoline PM Emission Rate Analysis
Vehicle
MY
Age
Test cycle
GVWR
flbl
PM2 5 mg/mi
1
2001
3
UDDS
12,975
1.81
2001
3
UDDS
19,463
3.61
2
1983
21
UDDS
9,850
43.3
1983
21
UDDS
14,775
54.3
3
1993
12
UDDS
13,000
67.1
1993
12
UDDS
19,500
108.3
4
1987
18
UDDS
10,600
96.7
1987
18
UDDS
15,900
21.5
174

-------
The table shows the four vehicles, two of which are quite old and certified to fairly lenient
standards. A third truck is also fairly old at twelve years and certified to an intermediate standard.
The fourth is a relatively new truck at age three and certified to a more stringent standard. No
trucks in the sample are certified to the Tier 2 or equivalent standards.
Examination of the heavy-duty data shows two distinct levels: vehicle #1 (MY 2001) and the other
three vehicles. Because of its lower age (3 years old) and newer model year status, this vehicle has
substantially lower PM emission levels than the others, and initially was separated in the analysis.
The emissions of the other three vehicles were averaged together to produce these mean results:
Mean for Vehicles 2 through 4: 65.22 mg/mi Older Group
Mean for Vehicle 1:	2.71 mg/mi Newer Group
3.1.2.1.2 LHD
To compare these rates with rates from light-duty gasoline vehicles, we simulated UDDS cycle
emission rates based on MOVES2010b light-duty gas PM2.5 emission rates (with normal
deterioration assumptions) for light-duty gasoline trucks (regulatory class LDT). The UDDS cycle
represents standardized operation for the heavy-duty vehicles.
The simulated light-duty UDDS results were then compared to the results from the four heavy-duty
gas trucks in the sample. Emission rates from the following MOVES model year groups and age
groups for light-duty trucks were used:
•	MY group 1983-1984, age 20+
•	MY group 1986-1987, age 15-19
•	MY group 1991-1993, age 10-14
•	MY group 2001, age 0-3
The simulated PM2.5 UDDS emission factors for the older light-duty gas truck group using
MOVES2010b are 38.84 mg/mi (ignoring sulfate emissions which are on the order of lxlO"4
65 22—
mg/mile for low sulfur fuels). This value leads to the computation of the ratio: —:—— = l 679.
38.84^f
mile
The simulated PM2.5 UDDS emission rates for the newer light-duty gas truck group are 4.687
mg/mi using MOVES2010b. Ignoring sulfate emissions, which are in the order of lxlO"5 mg/mile
2 71—
for low sulfur fuels, this value leads to the computation of the ratio: —:—= 0 578 .
4.687^f
mile
The newer model year group produces a ratio which is less than one and implied that large trucks
produce less PM2.5 emissions than smaller trucks. This result is counter-intuitive and is the likely
result of a very small sample and a large natural variability in emission results.
Thus, all four data points were retained and averaged together by giving the older model year group
a 75 percent weighting and the newer model year group (MY 2001) a 25 percent weighting. This is
consistent with the underlying data sample. It produces a final ratio of:
175

-------
Ratiofinal = RatioolderWtFrac + Rationewer(l — WtFrac)
= 1.679x0.75 + 0.578x0.25 = 1.40
Equation 3-3
We then multiplied this final ratio of 1.40 by the light-duty gasoline truck PM2.5 rates to calculate
the input emission rates for heavy-duty gasoline PM2.5 rates.
This approach is similar to how the LHD THC, CO, and NOx emissions for MY 2008 and 2009
were estimated by using the light-duty gasoline truck emissions as the basis, with VSP-based light-
duty rates applied as STP-based LHD2b3 emission rates.This assumption was deemed an
acceptable approximation because the LHD2b3 gasoline vehicles are chassis certified to distance-
based standards (g/mi). Accordingly, the vehicle emissions rates are less dependent on the
individual power and weight of the vehicle, and should scale approximately to the the g/mile
emission standards.11
3.1.2.1.3 MHD and HHD
For MHD and HHD regulatory classes, the emission rates are based on a fscale of 17.1. The LHD
emission rates are based on the light-duty truck rates, with an fscaie of 2.06.
We used an indirect approach to derive MHD and HHD PM2.5 emission rates from the LHD
emission rates. We assume that the relationship of total hydrocarbon (THC) between emission rates
based on an fSCaie of 2.06 and 17.1 is a reasonable surrogate to map PM2.5 emission rates from an
fscaie of 2.06 and 17.1 because both pollutants are products of incomplete fuel combustion and
unburned lubricating oil. For the mapping, we first calculated the emission rate ratio for THC
emissions for each operating mode between regulatory class MHD (regClassID 46) and LHD2b3
(regClassID 41). We then multiplied this ratio by the EC and nonEC PM2.5 emission rates in
regulatory class LHD2b3 (regClassID 41) to obtain EC and nonEC emission rates based on the
17.1 fscaie used in the heavier regulatory classes (RegClassID 46 and 47). An example of the
regulatory class LHD2b3 EC emission rates, 17.1/2.06/caie THC ratios, and the calculated 17.1
/scale based EC emission rates are displayed in Table 3-12. No reductions are made between 2003
and 2009, because the 2007 HD rule is not anticipated to cause reductions in heavy-duty gasoline
PM2.5 emissions.
11 This approximation needs to be revisited in the future now that we have updated the mass of LHD vehicles in
MOVES3 to range from 3.5 to 7.8 metric tonsError! Bookmark not defined., which differs from the fscau value of
2.06 metric tons.
JJ For example, the LHD gasoline PM2.5 age 0-3 emission rates for model year 2016 are on average 5.5 mg/mile and 7
mg/mile for LHD2b3 and LHD45, respectively, using nationally representative operating mode distributions (See
Figure 3-11). In contrast, the MHD gasoline PM2 5 rates are lower than the comparable MHD diesel PM2 5 emission
rates.
176

-------
Table 3-12. Derivation of MHD and HHD Elemental Carbon Emission Rates from LHD2b3 Rates using/scafe
17.1/2.06 THC emission ratios. Using Model Year 2001 as an Example


fscale


LHD2b3 EC
17.1/2.06
MHD and HHD EC
opModelD
emission rates
THC
emission rates

(mg/hr)
emission
ratios
(mg/hr)
0
0.59
1.000
0.59
1
0.54
1.000
0.54
11
0.60
1.000
0.60
12
0.79
2.263
1.78
13
1.38
3.677
5.08
14
2.62
5.095
13.37
15
5.55
5.443
30.22
16
64.52
5.427
350.13
21
8.38
1.000
8.38
22
2.92
1.154
3.37
23
2.08
2.173
4.52
24
2.92
2.825
8.24
25
10.94
4.842
52.95
27
20.50
7.906
162.10
28
126.42
8.796
1,112.05
29
523.16
6.471
3,385.32
30
2,366.75
7.102
16,809.50
33
26.59
2.121
56.40
35
10.76
4.780
51.42
37
13.29
4.010
53.28
38
43.61
8.979
391.56
39
75.73
9.522
721.06
40
74.96
5.300
397.26
The resulting PM2.5 emissions by regulatory class for LHD, MHD and HHD are shown in Figure
3-10. In general, PM2.5 emission rates are of similar magnitude for each regulatory class between
model year 1980 and 2009. There is significant variation in the model years, with some unexpected
trends (e.g., LHD45 has higher emission rates than HHD and MHD for most of these model years).
These unexpected trends and variation in the emission rates across model years and regulatory class
reflect uncertainties in deriving the pre-2010 emission rates heavy-duty gasoline from light-duty
gasoline data and THC surrogate values.
3.1.2.2 2010-2026Model Years
The real-world PEMS-based emissions measurement data from two engine-certified and one
chassis-certified heavy-duty gasoline vehicles used to update the MY 2010 and later THC, CO, and
NOx emission rates (Section 3.1.1.2.1) did not include PM2.5. Lacking appropriate PM data by
operating mode, we populated the MY 2010+ HD gasoline PM2.5 rates by copying MY 2010+ HD
diesel rates.
177

-------
This decision was supported by analysis of laboratory chassis tests. Gravimetric filter-based PM2.5
emissions measured from the three HD gasoline vehicles (described in Section 3.1.1.2.1) over
various chassis-dynamometer tests are shown in Table 3-13 The average PM2.5 rate over all
vehicles and test cycles is 1.35 mg/mi. The average PM2.5 emission rate for MY 2016 age 0-3, LHD
diesel (comparable to the tested gasoline vehicles) using nationally representative operating mode
distributions and average speeds is 1.4 mg/mile (See Figure 2-34). Since those numbers were
comparable given the uncertainty of the PM2.5 emission rates, and no modal HD gasoline PM2.5
data was available, we decided to use the HD diesel PM2.5 rates for HD gasoline. These rates also
include the tampering and mal-maintenance age effects for model year 2010-2026 (see Appendix
B.8).
Table 3-13 PM2.5 Emissions for Lab-Based Cycles for HD Gasoline Vehicles1
Vehicle
FTP
HWFET
LA92
Supercycle
Average
2015 ISUZU NPR
1.74
0.75
1.69
2.73
1.64
2016 Ford E450
0.53
0.55
1.55
2.51
1.17
2017 RAM 3500
1.68
0.40
1.43
1.35
1.34
Average
1.36
0.57
1.53
2.24
1.35
Note:
1 The vehicles are described in section 3.1.1.2.1.
The draft diesel LHD2b3 and LHD45 PM2.5 rates were copied to the gasoline LHD2b3 and LHD45
rates, respectively, from a MOVES version used for the preliminary HD2027 Rule analysis.108
Since the diesel MHD rates were notably higher than the diesel LHD and HHD rates, the diesel
HHD rates were used for gasoline MHD and HHD. Note that after this analysis, the heavy-duty
diesel PM2.5 emission rates in MOVES were updated to account for the updated HDIUT sample
and model year split described in Section 2.1.2.2. For this reason, the zero-mile PM2.5 emission
rates from heavy-duty gasoline are constant for 2010 and later model years, whereas the heavy-duty
diesel PM2.5 emission rates are reduced starting in model year 2013, and the heavy-duty gasoline
rates for 2013 and later are generally higher than the comparable heavy-duty diesel rates.JJ
Gasoline engine rates for LHD45 and LHD2b3 are higher than the HHD rates for all 2010 and later
years, whereas this trend is only seen in the initial model years (2010-2013) of the heavy-duty
diesel PM2.5 rates from the HDIUT program (Figure 2-34). We intend to update the HD gasoline
rates in a future MOVES version to incorporate any new HD gasoline emissions data, or at least to
be consistent with the updated MOVES HD diesel rates.
The gasoline rates were copied from the diesel rates as PM2.5, and then allocated to EC and
nonECPM using gasoline-specific fractions based on the Kansas City study of light-duty cars and
trucks as described in the MOVES3 Speciation Report.1 Because the diesel EC (9.98 percent) and
nonEC (90.02 percent) split of PM2.5 differs from gasoline EC (14 percent) and nonEC (86
percent), the EC and nonECPM emissions rates stored in the MOVES database are also quite
JJ For example, the LHD gasoline PM2 5 age 0-3 emission rates for model year 2016 are on average 5.5 mg/mile and 7
mg/mile for LHD2b3 and LHD45, respectively, using nationally representative operating mode distributions (See
Figure 3-11). In contrast, the MHD gasoline PM2 5 rates are lower than the comparable MHD diesel PM2 5 emission
rates.
178

-------
different than the diesel emission rates. Figure 3-12 shows the EC and nonECPM emission rates for
gasoline LHD45 vehicles by model year for age 0-3 vehicles.
The MOVES heavy-duty gasoline PM2.5 zero mile emission rates are constant for MY 2013
through 2026. There are differences between the 2010-2012 and 2013-2026 model year groups due
to different tampering and mal-maintenance assumptions applied to the diesel emission rates (see
Section 2.1.2.3), which primarily impact the ages 4-5 and older ages. The Tier 3 rulemaking sets
PM FTP emission standards for Class 2b and Class 3 of 8 mg/mile and 10 mg/mile, respectively,
which began phase-in starting with model year 2018 vehicles.52 We did not model reductions in the
gasoline PM2.5 emission rates with the phase-in of Tier 3, because the data on the tested heavy-duty
gasoline vehicles (Table 3-13) suggests that the heavy-duty gasoline vehicles are well in
compliance with the Tier 3 standard. In addition, the diesel rates on which the gasoline rates are
based also are well in compliance with the Tier 3 standards as discussed in Section 2.1.1.5.5.
3.1.2.3 202 7-2060 Model Years
In order to account for the HD2027 standards in MOVES4, we revised the PM2.5 running exhaust
emission rates for MY2027+ heavy-duty (LHD45, MHD, HHD) gasoline vehicles using the
methodology described in 3.1.1.3.2 for other criteria pollutants.
To meet the PM standards (5 mg/hp-hr for MY2027+), manufacturers are expected to improve fuel
control and limit the need for catalyst protection. Therefore, we assumed a 50 percent reduction in
PM2.5, consistent with the 50 percent more stringent PM standard and revised the MOVES
emission rates using Equation 3-2.
179

-------
3.1.2.4 Model Year Trends
Figure 3-10 and Figure 3-11 display the PM2.5 rates by model year and regulatory class for age 0-3
age group estimated in grams per mile (g/mile) using nationally representative operating mode
distributions and average speeds.











[<*V\




Im r. .1 a\
u 1A/



Reg Class
»- 41-LHD2b3
»- 42-LHD45
46-MHD67
¦- 47-HHD8
1980
2000
2020
2040
Model Year
Figure 3-10. Base running emission rates for PM2.5 from age 0-3 gasoline heavy-duty vehicles averaged over a
nationally representative operating mode distribution
180

-------
0.015-
0.010-
!
3
i
3 0.005-
o
0.000-
Reg Class
*- 41-LHD2b3
»- 42-LHD45
46-MHD67
¦- 47-HHD8
2010
2020
2030
2040
Model Year
Figure 3-11 Base running emission rates for PM2.5 from age 0-3 gasoline heavy-duty vehicles for MY2007-2040
averaged over a nationally representative operating mode distribution
Figure 3-12 shows the PM2.5 emission rates separated into elemental carbon (EC) and non-
elemental carbon (nonEC) fractions for age 0-3 HHD gasoline vehicles using nationally
representative operating mode distributions and average speeds. The EC/PM fractions are
dominated by the nonEC across all model years.
181

-------
0.20
~ 0.15'
a>
0.10'
a- 0.05'
0.00'
EC
NonEC
1980
1990
2000
2010
2020
2030
Model Year
Figure 3-12 Heavy Duty Gasoline Running Exhaust PM2.5 Emission Rates by Elemental Carbon and Non-
Elemental Carbon (nonEC) Fractions for the 0-3 Age Group by Model Year and Regulatory Class using
Nationally Representative Operating Mode Distributions
3.1.3 Energy
3.1.3.1 1960-2009 Model Years
3.1.3.1.1	LHD
The energy rates for gasoline LHD (LHD2b3 and LHD45 regulatory classes) pre-2009 energy rates
are unchanged from MOVES2010a. In MOVES2010a, the energy rates for LHD2b3 and LHD45,
along with the light-duty regulatory classes, were consolidated across weight classes, engine size
and engine technologies, as discussed in the MOVES2010a energy updates report57.
3.1.3.1.2	MHD and HHD
The energy rates for gasoline MHD and HHD pre-2009 energy rates are unchanged from
MOVES2014. The rates were developed using the same data set we used to develop the THC, CO,
and NOx exhaust emission rates. Similar to the analysis for the diesel running exhaust energy rates,
we made no distinction in rates by model year, age, or regulatory class. To calculate energy rates
(kJ/hour) from CO2 emissions, we used a heating value (HV) of 122,893 kJ/gallon and CO2 fuel-
specific emission factor (fcoj) of 8,788 g/gallon for gasoline (see Equation 3-20). STP was
calculated using Equation 1-6. Figure 3-13 presents the gasoline running exhaust energy rates in
MOVES for these regulatory classes.
182

-------
6
5
~i	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
Operating mode
Figure 3-13. Gasoline Running Exhaust Energy Rates for MHD (1960-2009) and HHD (1960-2009)
A linear extrapolation to determine rates at the highest operating modes in each speed range was
performed analogously to diesel energy and NOx rates (see Section 2.1.1.4.2).
3.1.3.2 2010-2060 Model Years
The real-world PEMS-based emissions measurement data from two engine-certified and one
chassis-certified heavy-duty gasoline vehicles used to update the THC, CO, and NOx emission rates
(Section 3.1.1.2.1) included CO2 emissions data which was used to update the energy rates. The
energy rates are derived using the measured CO2 values and the conventional gasoline specific
values for carbon content (0.0196 g/KJ) and oxidation fraction (1.0) and the molecular mass of CO2
(44), and atomic mass of Carbon (12). These values are described in the MOVES GHG and Energy
Rates report.3
When calculating the operating mode-based energy rates for high-power operating modes with
limited or missing data, we extrapolated using STP values using the method described in Section
2.1.1.4.2.
For LHD2b3, the energy rates are identical for MY 2010-2013. For LHD45, MHD, and HHD, the
energy rates are identical for MY 2010-2015.
183

-------
3.1.3.2.1
LHD2b3 2014-2060 Model Years
The LHD2b3 gasoline energy rates are reduced to incorporate the impacts of the Phase 1 and Phase
2 Heavy-duty Greenhouse Gas rules. The LHD2b3 gasoline rates are adjusted from the 2010-2013
model year rates using the gasoline reductions documented in Table 3-14 (Phase 1) and Table 2-33
(Phase 2) in Section 2.1.4.3.
Table 3-14 Estimated Total Vehicle Reductions in Energy Consumption Rates for LHD2b3 Gasoline Vehicles
Regulatory
Class
Model years
Reduction from MY
2013 Energy Rates
LHD2b3
2014
1.5%
2015
2%
2016
4%
2017
6%
2018-2020
10%
The HD GHG Phase 1 reductions for the affected model years are incorporated into the energy
rates in the emissionRate table in the MOVES database. The adjustments for HD GHG Phase 2 are
applied at run-time using the values in the emissionRateAdjustment table in the MOVES database.
3.1.3.2.2 LHD45, MHD, andHHD 2016-2060Model Years
Updates to the energy rates were made to the heavy-duty gasoline energy rates for model years
2016-2020 based on the Phase 1 Medium and Heavy-Duty Greenhouse Gas Rule111 discussed in
Section 2.1.4.3 and shown in Table 3-15.
Table 3-15 Heavy-Duty Gasoline Reductions due to the Heavy-Duty GHG Phase 1 Rule
Regulatory Class
Model
Years
CO2 Reduction From 2013
Baseline
LHD45, MHD, HHD
2016-2020
5%
The energy rates for 2021 model year and beyond were updated in MOVES3 to reflect the CO2
emission reductions expected from the Heavy-Duty GHG Phase 2 rule, as shown in Table 3-16,
which have separate reductions for vocational and combination trucks.
As noted above, the HD GHG Phase 2 reductions to energy rates are not incorporated into the
energy rates in the emissionRate table in the MOVES database, but are applied at run-time using
the values in the emissionRateAdjustment table in the MOVES database.
184

-------
Table 3-16 Heavy-Duty Gasoline Reductions due to the Heavy-Duty GHG Phase 2 Rule112
Source Type
(SourceTypelD)
Regulatory Class
Model
Years
CO2 Reduction from
2017 Baseline
Other Bus, School Bus,
Refuse Truck, Single-Unit
Short-Haul, Single-Unit
Long-Haul, Motorhomes
(41,43,51,52,53, 54)
LHD45
2021-2023
6.9%
2024-2026
9.8%
2027+
13.3%
Other Bus, School Bus,
Refuse Truck, Single-Unit
Short-Haul, Single-Unit
Long-Haul, Motorhomes
(41,43,51,52,53, 54)
MHD and HHD
2021-2023
6.9%
2024-2026
9.8%
2027+
13.3%
Short-haul Combination
Trucks(61)
MHD and HHD
2018-2020
0.6%
2021-2023
7.4%
2024-2026
11.9%
2027+
15.0%
3.1.3.3 Model Year Trends
Figure 3-14 and Figure 3-15 display the CO2 (g/mile) emission rates and fuel economy values
calculated from the energy rates using the carbon content and energy density conversion factors for
conventional gasoline1^ as documented in the MOVES3 Greenhouse Gas and Energy Report.3 The
CO2 (g/mile) emission rates and fuel economy values are estimated using nationally representative
operating mode distribution and average speed values. Figure 3-15 displays the significant decrease
in fuel economy in model year 2010 due to the updated data and analysis incorporated for model
year 2010 trucks. The large change in fuel economy is not anticipated to be real, but an artifact of
the using the updated data and analysis.
The LHD emission rates show substantial variability in the early model years (pre-1985) - we do
not expect LHD vehicles to have lower fuel economy than MHD and HHD for these years, but
have not revisited these emission rates due to the small number of pre-1985 gasoline vehicles
remaining in the onroad fleet. As discussed in Section 2.1.4.1, the detailed methodology used in
MOVES2004 (which modeled different emission rates according to vehicle weights, engine
technologies, and engine sizes) introduced variability into the energy rates within the current
MOVES regulatory class emission rates for pre-2010 LHD.
The figures display that, since model year 2010, there are decreasing trends in CO2 (g/mile) with
corresponding increases in fuel economy, due to the lower energy rates as well as lower source
kk Using the energy content of conventional gasoline (E0), the fuel economy is ~4% higher than is estimated using the
energy content of E10 gasoline. Note that E10 is estimated to be the dominant gasoline fuel sold in 2008 and later.
MOVES has the same carbon content for both fuels, so there is no estimated impact on the CO2 g/mile.
185

-------
mass values and improved road load coefficients estimated with the Phase 1 and Phase 2 heavy-
duty greenhouse gas rulemaking. The energy rates by operating mode are constant for model year
2027-2060. However, some small differences in CO2 (g/mile) or fuel economy values observed
within model year groups and regulatory classes with the same energy rates are due to differences
in the nationally representative operating modes, which are different across model years due to
changing fractions of regulatory classes among different source types.
1980	2000	2020	2040
Model Year
Figure 3-14. Base running emission rates for CO2 from age 0-3 gasoline heavy-duty vehicles averaged over a
nationally representative operating mode distribution
2000
1500
aT
|
5
o>
I 1000
CM
O
O
Reg Class
41-LHD2b3
42-LHD45
46-MHD67
— 47-HHD8
186

-------
Reg Class
*- 41-LHD2b3
»- 42-LHD45
46-MHD67
•- 47-HHD8
1980
2000
2020
2040
Model Year
Figure 3-15. Fuel economy for age 0-3 gasoline heavy-duty vehicles averaged over a nationally representative
operating mode distribution
3.2
Start Emissions
Representative in-use data on vehicle start emissions for heavy-duty gasoline vehicles is even less
common than running data. While some data was available (Table 3-18, Table 3-23), the MOVES
analysis also relies on deterioration patterns from light-duty vehicles, as well as ratios to the
relevant engine emission standards. For LHD2b3 gasoline vehicles, manufacturers comply with
chassis (g/mile) emission standards. For the larger regulatory classes, engine emission standards
apply. We used the engine emission standards to estimate differences in emissions between the
LHD2b3 regulatory class and the heavier regulatory classes. Most of this analysis has been carried
over from MOVES2010b and MOVES2014, but the cold start emissions for LHD45, MHD, and
HHD gasoline engines of 2008 model year and later have been updated for MOVES3 based on
recent certification data.
The heavy-duty spark ignition engine emissions standards113 for the Federal Test Procedure (FTP)
are shown in Table 3-17. Note that the standards for model years 1990 through 2004 for CO and
THC vary by weight class, but not by model year, whereas those for NOx vary by model year, but
not by weight class. Also, for model years 2005-2007, a single standard is applied for
NMHC+NOx, but by 2008, separate but lower standards are again in effect. Note also that by
model year 2008, the standards for the three gaseous pollutants are the same across regulatory
class.
187

-------
Table 3-17 FTP Standards (g/hp-hr) for Heavy-Duty Gasoline Engines for Model Years 1990-2008+113
Model-Year
Group
GVWR < 14,000 lb
(LHD2b3)
GVWR > 14,000 lb

CO
NMHC1
NOx
CO
NMHC1
NOx
1990
14.4
1.1
6.0
37.1
1.9
6.0
1991-1997
14.4
1.1
5.0
37.1
1.9
5.0
1998-2004
14.4
1.1
4.0
37.1
1.9
4.0
2005-2007
14.4
1.01
37.1
1.02
2008+
14.4
0.14
0.20
14.4
0.14
0.20
Note:
1 Non-methane hydrocarbons standard expressed as NMHC + NOx
3.2.1 THC, CO, and NOx
The heavy-duty gasoline vehicle start emissions for MOVES regulatory class LHD2b3 and LHD45
vehicles are discussed in Section 3.2.1.1. Section 3.2.1.2 discusses the development of the rates for
MOVES regulatory class MHD and HHD gasoline vehicles. In Section 3.2.1.3, we summarize and
compare the two sets of start emission rates for THC, CO and NOx. Soak time adjustments are
detailed in Section 3.2.3.
3.2.1.1 LHD2b3
For LHD2b3, the gaseous emission rates for MY 1960-2004 are based on data analysis of test data,
and the MY 2005+ emission rates are based on ratioing the pre-2005 rates based on the emission
standards.
3.2.1.1.1 1960-2004 Model Years
To develop start emission rates for MY 1960-2004 heavy-duty gasoline-fueled vehicles, we
extracted data available in EPA's Mobile-Source Observation Database (MSOD).104 These data
represent aggregate test results for heavy-duty spark-ignition (gasoline powered) engines measured
on the Federal Test Procedure (FTP) cycle. The GVWR for all trucks was between 8,500 and
14,000 lbs, placing all trucks in the LHD2b3 regulatory class. The 1960-2004 LHD2b3 start rates
are unchanged from LHD2b3 start emission rates in MOVES2010b.
Table 3-18 shows the model-year by age classification for the data. The model year groups in the
table were designed based on the progression in NOx standards between MY 1990 and 2004.
Standards for CO and THC are stable over this period, until MY 2004, when a combined NMHC+
NOx standard was introduced. However, no measurements for gasoline HD trucks were available
for MY2004 and later.
Start emissions are not dependent on power, and therefore, the emission rates do not need to be
calculated differently to distinguish different fscale values as was done for running exhaust rates. As
188

-------
discussed later, start emission rates are separated by regulatory classes to account for differences in
the emission standards and/or available test data.
Table 3-18 Availability of Emissions Start Data by Model-Year Group and Age Group for LHD2b3 Vehicles
Model-year Group
Age Group (Years)
Total
0-3
4-5
6-7
8-9
10-14
1960-1989



19
22
41
1990


1
29

30
1991-1997
73
59
32
4

168
1998-2004
8




8
Total
81
59
33
52
22
247
3.2.1.1.2 Estimation of Mean Rates
As with light-duty vehicles, we estimated the "cold-start" as the mass from the cold-start phase of
the FTP (bag 1) less the "hot-start" phase (Bag 3). As a preliminary exploration of the data, we
averaged by model year group and age group and produced the graphs shown in Appendix F.
Sample sizes were small overall and very small in some cases (e.g., 1990, age 6-7) and the
behavior of the averages was somewhat erratic. In contrast to light-duty vehicle emissions, strong
model-year effects were not apparent. This may not be surprising for CO or THC, given the
uniformity of standards throughout. This result was more surprising for NOx, but model year trends
are no more evident for NOx than for the other two. Broadly speaking, it appeared that an age trend
may be evident.
If we assume that the underlying population distributions are approximately log-normal, we can
visualize the data in ways that illustrate underlying relationships. As a first step, we calculated
geometric mean emissions, for purposes of comparison to the arithmetic means calculated by
simply averaging the data. Based on the assumption of log-normality, the geometric mean (xg) was
calculated in terms of the logarithmic mean (xi) as shown in Equation 3-4.
Xg = Q^1	Equation 3-4
This measure was not appropriate for use as an emission rate, but was useful in that it represents
the "center" of the skewed parent distribution. As such, it was less strongly influenced by unusually
high or outlying measurements than the arithmetic means. In general, the small differences between
geometric means and arithmetic means suggest that the distributions represented by the data do not
show strong skew in most cases. Because evidence from light-duty vehicles suggested that
emissions distributions should be strongly skewed, this result implied that these data are not
representative of "real-world" emissions for these vehicles. This conclusion appeared to be
reinforced by the values in Figure F-3 which represent the "logarithmic standard deviation"
calculated by model-year and age groups. This measure (si), is the standard deviation of natural
logarithm of emissions (pci). The values of si were highly variable, and generally less than 0.8,
showing that the degree of skew in the data was also highly variable as well as generally low for
emissions data; e.g., corresponding values for light-duty running emissions are generally 1.0 or
189

-------
greater. Overall, review of the geometric means confirmed the impression of age trends in the CO
and THC results, and the general lack of an age trend in the NOx results.
Given the conclusion that the data as such are probably unrepresentative, assuming the log-normal
parent distributions allowed us to re-estimate the arithmetic mean after assuming reasonable values
for si. For this calculation, we assumed values of 0.9 for CO and THC and 1.2 for NOx. These
values approximate the maxima seen in these data.
The re-estimated arithmetic means were calculated from the geometric means, by adding a term
that represents the influence of the "dirtier" or "higher-emitting" vehicles, or the "upper tail of the
distribution," as shown in Equation 3-5.
s2
~ ~ p-k-	Equation 3-5
xa ~
For purposes of rate development using these data, we concluded that a model-year group effect
was not evident and re-averaged all data by age group alone. Results of the coarser averaging are
presented in Figure 3-16 with the arithmetic mean (directly calculated and re-estimated) and
geometric means shown separately.
190

-------
0 I I I I I I I I I I I I I I I I I I I I I I I I I I I
0	2	4	6	8 10 12 14
Age (years)
16
0	2	4	6	8	10 12 14
Age (years)
8 i
Age (years)
Figure 3-16. Cold-start FTP Emissions for Heavy-Duty Gasoline Trucks, Averaged by Age Group Only (g
Geometric Mean, a= Arithmetic Mean Recalculated from xi and si)
191

-------
We then addressed the question of the projection of age trends. As a general principle, we did not
allow emissions to decline with age. For THC and NOx, we assumed the emission rates stabilized at
the maximum level reached at the 6-7 and 8-9 age groups, respectively as shown in Table 3-19. For
CO emissions, we kept the age trends as they were, since there was only a slight decrease in CO
emissions after the maximum was reached in the 8-9 age group.
3.2.1.1.3 Estimation of Uncertainty
We calculated standard errors for each mean in a manner consistent with the re-calculation of the
arithmetic means. Because the (arithmetic) means were recalculated with assumed values of si, it
was necessary to re-estimate corresponding standard deviations for the parent distribution 5, as
shown in Equation 3-6.
S = ^|x|es2(es2 — 1)	Equation 3-6
After recalculating the standard deviations, the calculation of corresponding standard errors was
simple. Because each vehicle is represented by only one data point, there was no within-vehicle
variability to consider, and the standard error could be calculated as s/yfn. We divided the standard
errors by their respective means to obtain CV-of-the-mean or "relative standard error." Means,
standard deviations and uncertainties are presented in Table 3-19 and in Figure 3-17. Note that
these results represent only "cold-start" rates (opModelD 108). Soak time adjustments other start
opModes are detailed in Section 3.2.3.
192

-------
Table 3-19. Cold-Start Emission Rates (g) for Heavy-Duty Gasoline Trucks, by Age Group (Italicized Values
Age Group
n
Pollutant


CO
THC
NO*
Means
0-3
81
101.2
6.39
4.23
4-5
59
133.0
7.40
5.18
6-7
33
155.9
11.21
6.12
8-9
52
190.3
11.21
7.08
10-14
22
189.1
11.21
7.08
Standard Deviations
0-3

108.1
6.82
8.55
4-5

142.0
7.90

6-7

166.5
11.98
12.39
8-9

203.2
11.98
14.32
10-14

202.0
11.98
14.32
Standard Errors
0-3

12.01
0.758
0.951
4-5

18.49
1.03
1.18
6-7

28.98
2.08
2.16
8-9

28.18
2.08
1.99
10-14

43.06
2.08
1.99
193

-------
300
250
3200
& 150
3 100
*so
(a) CO
















k











10	15
Age (years)
20
25
3
a
£
2
o
(J
Q.
L
(b

; ino











	^
















10	15
Age (years)
20
25
12
10 -
3
£
(c) NQx

































10	15
Age (years)
20
25
Figure 3-17. Cold-Start Emission Rates for Heavy-Duty Gasoline Trucks, with 95 Percent Confidence Intervals
The steps described so far involved reduction and analysis of the available emissions data. In the
next step, we describe approaches used to impute rates for model years not represented in these
data. For purposes of analysis, we delineated four model year groups: 1960-2004, 2005-2007,
2008-2017 and 2018 and later. The rates above were used for the 1960-2004 model year group. We
describe the derivation of rates for the remaining groups below.
194

-------
3.2.1.1.4
2005-2007Model Years
For the 2005-2017 model year emission rates, we applied reductions to the 1960-2004 emission
rates, by comparing the standards between the two model year ranges. For CO, the approach was
simple. We applied the age zero values in Table 3-19 to the 2005-2007 model year group. The
rationale for this approach is that the CO standards do not change over the full range of model
years considered.
For THC and NOx, we imputed values for the 2005-2007 and 2008-2017 model-year groups by
multiplying the age zero values for the 1960-2004 emission rates in Table 3-19 by ratios expressed
in terms of the applicable standards. Starting in 2005, a combined THC+NOx standard was
introduced. It was necessary for modeling purposes to partition the standard into THC and NOx
components. We assumed that the proportions of NMHC and NOx would be similar to those in the
2008 standards, which separate NMHC and NOx while reducing both.
We calculated the THC value by multiplying the 1960-2004 value by the fraction fac as shown in
Equation 3-7.
0.14 g/hp-hr V10g/hp.hr)
v (0.14 + 0.20) g/hp-hr J"	Equation3-7
HC ~	1.1 g/hp-hr	~
This ratio represents the component of the 2005 combined standard attributed to NMHC. We
calculated the corresponding value for NOx as shown in Equation 3-8.
0.20 g/hp-hr \0g/hp_hr
, v(0-14 + 0 20) g/hp-hr,	
4.0 g/hp-hr
Equation 3-8
For these heavy-duty rates, we neglected the THC/NMHC conversions, to which we gave attention
for light-duty.
3.2.1.1.5 2008-2017Model Years
For the 2008-2017 model years, the approach to projecting rates was modified to adopt two
refinements developed for light-duty rates. First, start emission rates for the LHD2b3 gasoline
vehicles were estimated from composite rates by applying the "start split-ratio" shown in Table 3-6
to a set of rates representing light-duty trucks in Tier-2/Bin 8. Second, we updated the deterioration
effects for start NOx exhaust from MOVES2014114' by applying the ratios shown in Table 3-20.
195

-------
Table 3-20. Deterioration Reduction Ratio for 2008-2017 gasoline LHD2b3 NOx Starts
ageGroupID
Deterioration Reduction
Ratio
3
1.00
405
0.85
607
0.79
809
0.73
1014
0.62
1519
0.62
2099
0.62
These ratios were initially developed for LDT in a draft version of MOVES3115 but the LD age
effects were further updated prior to MOVES3 release.11
For THC and CO, the multiplicative age effects are unchanged from the effects from
MOVES2014.116 The resulting multiplicative age effects for start emission rates for LHD2b3
vehicles used in MOVES3 for model years 2009-2017 are shown in Table 3-21. The start emission
rates for model year 2008 are estimated by averaging the MY 2007 and 2009 emission rates across
all age groups and operating modes assuming a phase-in of 50% of the Tier 2 standards and the HD
2007 Rule in MY 2008 as we assumed for LHD2b3 gasoline running emissions as discussed in
Section 3.1.1.2.1. The relative age effects for LHD2b3 MY 2008 and MY 2009-2017 are shown
Figure 3-21.
Table 3-21 Multiplicative Age Effect Used for Start Emissions for Gasoline LHD2b3 Vehicles for 2009-2017
Model Years Adopted from the Deterioration Effects for Light-Duty Trucks
ageGroupID
THC
CO
NOx
3
1
1
1
405
1.65
1.93
1.47
607
2.20
2.36
1.74
809
2.68
2.54
2.01
1014
3.30
3.00
2.00
1519
3.66
3.35
2.26
2099
4.42
4.06
2.56
Using these deterioration rates for starts results in start emission rates for MY 2010+ gasoline
LHD2b3 vehicles having a higher relative deterioration than running emission rates (compared to
Table 3-3).mm We recognize this is inconsistent with our knowledge of light-duty start
deterioration.9 We plan to address this data gap with data collected on LHD2b3 in future versions
of MOVES.
11 In MOVES3, we incorporated additional updates to the start deterioration rates (including for NOx) for LDT as
documented in the MOVES3 light-duty exhaust emission rate report.9
mm The updated MY 2010 and later heavy-duty running gasoline rates (including LHD2b3) use the heavy-duty age
effects as discussed in Section 3.1.1.2.1 and 3.1.1.3.1.
196

-------
3.2.1.1.6 Incorporating Tier 3 Standards: 2018 and Later Model Years
Emission rates representing the phase-in of Tier-3 standards for the start-exhaust process were
developed for MOVES2014 as described in gasoline running emissions section of the
MOVES2014 heavy-duty exhaust report.109 Like the MY 2008-2017 rates, the LHD2b3 Tier 3 start
rates are based on light-duty truck emission rates scaled to higher emission standards for the
LHD2b3 regulatory class. The reduction in start emissions due to Tier 3 is relatively lower than the
reductions in running emissions presented in Section 3.1.1.3.1.
The LHD2b3 start rates during and following the Tier 3 phase-in have relatively lower
deterioration than the start rates for the model years preceding the onset of the phase-in (MY 2008-
2017) as documented in the MOVES2014 light-duty exhaust report. 116'nn For MOVES3, we
adjusted the NOx start emission rates by applying the deterioration ratios in Table 3-20 to the
MOVES2014 NOx start rates. The multiplicative age effects for LHD2b3 cold start rates for THC,
CO and NOx after the complete phase-in of Tier 3 phase-in model year 2022 are shown below in
Table 3-22 The age effects of the phase-in years of Tier 3 (MY 2018-2021) are a weighted average
of the MY 2010-2017 and the MY 2022 start emission rates using the phase-in assumptions
documented in the MOVES2014 heavy-duty exhaust report.109
Table 3-22 Multiplicative Age Effect Used for Start Emissions for Gasoline LHD2b3 Vehicles for 2022-2060
Model Years
ageGroupID
THC
CO
NOx
3
1
1
1
405
1.54
1.73
1.38
607
1.94
1.97
1.57
809
2.26
1.96
1.74
1014
2.78
2.33
1.72
1519
3.09
2.59
1.95
2099
3.73
3.15
2.21
We do not model any impact of the HD2027 rule on the Gasoline LHD2b3 start emissions because
HD2027 standards affect only engine-certified gasoline light-heavy-duty vehicles (modelled as
MOVES3 regulatory class LHD45).
3.2.1.1.7 Inspection and Maintenance Program Effects for LHD Gasoline
Vehicles
In MOVES3.1, we updated the meanBaseRatelM values for THC, CO, and NOx start exhaust
emission rates to better reflect the I/M programs for LHD2b3 gasoline vehicles.
Due to a lack of data and analysis, in MOVES3.0.4 and earlier versions of MOVES, we assumed
no benefit from I/M programs for starts. However, as explained in for running emissions in Section
3.1.1.3.2, we updated the gasoline LHD2b3 HC, CO and NOx start values for MeanBaseRatelM in
1111 In MOVES3, the deterioration effects for all model year light-duty vehicles were updated on updated data and
analysis. The light-duty Tier 3 emission rates no longer have different deterioration values. We plan to update the
LHD2b3 start deterioration effects to be consistent with the light-duty vehicles in an upcoming version of MOVES.
197

-------
M0VES3.1 and later to reflect the same proportional reduction (that is
meanBaseRatelM/meanBaseRate) that we model for each operating mode bin and age for the light-
duty trucks. For more detail on the meanBaseRate and meanBaseRatelM values for light-duty
trucks, see the MOVES3 LD report.9
3.2.1.2 LHD45, MHD, and HHD
The start emission rates from LHD45, MHD, and HHD gasoline vehicles differ from the rates for
LHD2b3. The following two subsections document the emission rates for 1960-2007 model years
(Section 3.2.1.2.1) and 2008+ model years (Section 3.2.1.2.2).
3.2.1.2.1 1960-2007Model Years
Since bag data were lacking for MY 1960-2007 vehicles in classes LHD45 and MHD, we
estimated cold start values relative to the LHD2b3 start emission rates.
For CO and THC, we estimated rates for the heavier vehicles by multiplying them by ratios of
standards for the heavier class to those for the lighter class. The value of the ratio for CO based on
1990-2004 model year standards is shown in Equation 3-9.
37.1 g/hp - hr
fco - ——		— - 2.58	Equation 3-9
14.4 g/hp - hr
The corresponding ratio for THC for 1990-2004 model year vehicles is 1.73, as shown in Equation
3-10.
. 1.9g/hp-hr
/hc = ———	— = 1.73	Equation 3-10
1.1 g/hp - hr
The ratios derived in the previous two equations (2.58 and 1.73) were applied to estimate the start
emission rates for 1960-2004 and 2005-2007 model year groups for the LHD45, MHD, and HHD
gasoline vehicles (Table 3-25 ). Note that the ratios for CO and THC do not vary by model year
group because the standards do not; See Table 3-16.
For MY 1960-2007, NOx start emission rates for medium and heavy-duty vehicles are equal to the
LHD2b3 start emission rates, because the same standards apply to all the HD regulatory classes.
The approaches for all three regulatory classes in all model years are summarized in Table 3-25 .
3.2.1.2.2 2008-2060 Model Years
The cold start emissions for 2008 model year and later LHD45, MHD, and HHD gasoline engines
have been updated for MOVES3 based on new data. Similar to the approach taken for light-duty
vehicles and for diesel vehicles (see Section 2.2.1.2), the cold start emissions are calculated as the
difference in emissions between a test cycle with a cold start and the same test cycle with a hot
start. Heavy-duty gasoline engines are certified using the Heavy-Duty Gasoline Engine Federal
Test Procedure (FTP) cycle.117 The test procedure for certification requires that manufacturers run
the engine over the FTP cycle with a cold start and then repeat the cycle with a warm start. Starting
in model year 2016, EPA began collecting certification data that contained separate cold and hot
198

-------
results for each engine certified. The data that was analyzed for this MOVES3 update includes the
following engine families from the 2016 and 2017 model years shown in Table 3-23.
Table 3-23 Engine Data Analyzed to Revise the Cold Start Emission Rates for HP Gasoline Engines
Category
Number of Engines
Manufacturers
LHD45,
MHD, HHD
Gasoline
3
Ford, GM, Powertrain
Integration
The certification data was used to determine the grams emitted per cold start using Equation 3-11.
Grams per Start
= [Cold FTP Emission Results (g/(hp - hr))	Fmiatinn 3 11
- Hot FTP Emission Results(g/(hp - hr))]	4
* FTP Cycle Work (hp — hr)
The amount of work (hp-hr) performed over the FTP cycle is not provided as part of the
certification data submitted by the manufacturers to EPA. We only had cycle work data from one
19.3 hp-hr HD gasoline engine. While we acknowledge that FTP cycle work is unique to each
engine because it is created based on the engine's maximum speed, curb idle speed, and the
maximum torque curve, we estimated cycle work for all HD gasoline engines using our one engine
data source.
The analysis of cold and hot start FTP emissions data from three FID gasoline engines determined
the grams per start for TFtC, CO, NOx, and PM2.5. The mean and standard deviation of the TFtC,
CO, NOx, and PM2.5 emission levels for the three engines are shown in Table 3-24. The MY 2016
and 2017 engines ranged in displacement between 5.4 and 7.2 liters, and ranged in rated power
between 297 and 332 FTP. The new default cold start emissions values for MOVES3 are the mean
values shown in Table 3-24. The TFtC, NOx and PM2.5 cold start emissions for FID gasoline
engines are higher compared to MOVES2014, while the CO emissions are lower.
Table 3-24 Cold Start Emissions for MY 2008 and Later Heavy-Duty Gasoline Engines
Grams per Start
THC
CO
NOx
PM25
Mean
5.57
31.5
1.88
0.084
Standard Deviation
0.6
6.36
1.04
0.049
We applied the same relative age deterioration for the 2008+ model years starts for TFtC, CO and
PM2.5 as was used for the previous model year groups (which is based on the gasoline LFID2b3
1960-2004 model years). For NOx, we applied the relative age deterioration as was used for
LFID2b3 vehicles for MY 2008 and later vehicles shown in Table 3-21 and Table 3-22.00 The start
rates for TFtC, CO, and NOx for this model year group for each age are graphed in Figure 3-21.
00 In a future update to MOVES, we intend to update the HD gasoline deterioration to be consistent with the updates
made to the LD rates in MOVES3, as well as apply a consistent approach for HD gasoline emissions deterioration for
both start and running deterioration and for all pollutants.
199

-------
We do not model any impact of the HD2027 rule on the Gasoline LHD45, MHD or HHD start
emissions due to the lack of sufficient data to model the impact.
3.2.1.3 Summary
Table 3-25 summarizes the data and methods used to estimate THC, CO, and NOx start emission
rates from heavy-duty gasoline vehicles as discussed in Sections 3.2.1.1 and 3.2.1.2. Figure 3-18
through Figure 3-20 displays the cold start (operating mode 108) emission rates across model years
for heavy-duty gasoline vehicles.
Table 3-25 Summary of Cold Start Emission Rates for Heavy-Duty Gasoline Vehicles
Regulatory
Class
Model Year
Group
CO
THC
NOx

1960-2004
Data analysis, values from
Table 3-19
LHD2b3
2005-2007
Data analysis, values
from
Table 3-19
Reduce in proportion
to standards from 1960-2004
2008 -2017
Section 3.2.1.1.5 Based on Tier 2 Bin 8 LDT rates and deterioration

2018 +
Section 3.2.1.1.6. Based on LDT rates, adjusted to account for Tier 3
standards and assumed lower deterioration. No change for HD2027
rule.

1960-2004
Increased
in proportion
to standards from LHD2b3
Same values as
LHD2b3
LHD45,
MHD,
HHD
2005-2007
Increased in proportion
to standards from LHD2b3
Same values as
LHD2b3
2008 +
Updated based on FTP certification data,
deterioration based on the 1960-2004
LHD2b3 data. No change for HD2027 rule.
Updated based on FTP
certification data,
deterioration based on
the 2008 + LHD2b3
vehicles. No change
for HD2027 rule.
Note:
Soak time adjustments are detailed in Section 3.2.3.
The outcomes of the methods described in the table above are summarized graphically in Figure
3-18 through Figure 3-20 for cold-start emissions. The decline in start emissions with the adoption
of more stringent standards begins with the reduction in model year 2005 and ends at the
completion of the phase-in of Tier 3 standards for LHD2b3 vehicles in model year 2022. Note that
there is a slight increase in THC start emissions for LHD45 vehicles in model year 2008, which is
the first model year using the new start certification data discussed above in Section 3.2.1.1.5.
200

-------
9-
6-
3-
0-





	




1





Reg Class
¦- 41-LHD2b3
•- 42-LHD45
46-MHD67
•- 47-HHD8
1980
2000
2020
2040
Model Year
Figure 3-18 Heavy-duty Gasoline THC Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory
Class and Model Year. LHD45, MHD and HHD are equivalent.
200-
t
re
re
0£
t
re
2 100-
Reg Class
»- 41-LHD2b3
•- 42-LHD45
46-MHD67
•- 47-HHD8
1980
2000
2020
2040
Model Year
Figure 3-19 Heavy-duty Gasoline CO Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory
Class and Model Year. LHD45, MHD and HHD are equivalent.
201

-------
1980
2000
2020
2040
Model Year
Figure 3-20 Heavy-duty Gasoline NOx Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory
Class and Model Year. LHD45, MHD and HHD are equivalent.
Reg Class
41-LHD2b3
— 42-LHD45
46-MHD67
47-HHD8
2004
2007
2008
2017
I







J	
y

J

2022
0 4 8 12 16 20 240 4 8 12 16 20 240 4 8 12 16 20 240 4 8 12 16 20 240 4 8 12 16 20 24
Vehicle age, years
Reg Class - LHD2b3 - LHD45
Figure 3-21 Heavy-duty Gasoline Cold-Start Rates (opModelD 108) vs. Vehicle Age for Select Model Years
LHD45, MHD and HHD are equivalent.
202

-------
3.2.2 Particulate Matter (PM2. s)
Data on PM2.5 start emissions from heavy-duty gasoline vehicles were unavailable, so these
emissions were extrapolated as described below.
3.2.2.1	LHD2b3
For LHD2b3 vehicles, we used the multiplicative factor from the running exhaust emissions
analysis of 1.40 (derived in Equation 3-3 in Section 3.1.2.1.2) to scale up start emission rates from
light-duty trucks (LDT) for model years 1960-2003.
For 2004+ model years, the LHD2b3 start emission rates are 1.4 times the model year 2003 LDT
emission rates. We project constant start emissions using the 2003 model year emission rates rather
than scaling to the LDT PM2.5 rates with the 2004 and later model years because the LD rates
increase due to the updated data on emission rates and sales penetration of gasoline direct injection
technology, and subsequently, decrease beginning in model year 2018 with the implementation of
the Tier 3 Vehicle Emissions and Fuel Standards Program. We are not confident that such patterns
will apply to HD gasoline due to limited data regarding heavy-duty PM2.5 rates and uncertainty
regarding (a) the expected penetration of gasoline direct injection technology in heavy-duty
gasoline vehicles and (b) the impact of Tier 3 on HD gasoline PM2.5 emissions (see Section
3.1.2.2). We do not model any impact of the HD2027 rule on the Gasoline LHD2b3 start
emissions because HD2027 standards affect only engine-certified gasoline light-heavy-duty
vehicles (modelled as LHD45 in MOVES4 regulatory class).
The start PM2.5 emission rates for heavy-duty gasoline vehicles exhibit the same relative effects of
soak time, and deterioration as the LDT PM2.5 start emission rates.
3.2.2.2	LHD45, MHD, and HHD
Due to a lack of PM2.5 start data, we use the same PM2.5 emission rates for LHD2b3 for all heavy-
duty gasoline for MY 1960-2007. For MY 2008 and later, for LHD45, MHD, and HHD, we
updated the PM2.5 start emissions data using certification data presented in Table 3-24. This causes
the start emissions to increase significantly for LHD45, MHD, and HHD between MY 2007 and
MY 2008 as shown in Figure 3-22. We do not model any impact of the HD2027 rule on the
gasoline LHD45, MHD or HHD start emissions due to the lack of sufficient data to model the
impact.
We continue to apply the same age adjustments to the start PM2.5 as the LDT emissions due to lack
of data. We caution there is considerable uncertainty in the start heavy-duty gasoline PM2.5
emission rates, especially for pre-2007 model years.
3.2.2.1 Model Year Summary
Figure 3-22 displays the cold start emission rates across model years for heavy-duty gasoline
vehicles. For the LHD45, MHD and HHD vehicles, we have more confidence in the emission rates
203

-------
from the 2010 and later model year groups since they are based on certification results from these
engines.
1980
2000
2020
2040
Model Year
Figure 3-22. Heavy-duty Gasoline PM2.5 Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory
Class and Model Year. LHD45, MHD and HHD are equivalent

CJ
5>
2
o
o
u>
cm'
0.06
Reg Class
41-LHD2b3
42-LHD45
46-MHD67
47-HHD8
3.2.3 Soak Time Adjustments
To estimate the start emissions at various soak lengths, we apply the same soak fractions to the cold
start emissions that we apply to 1996-2003 MY light-duty gasoline vehicle as documented in the
light-duty emission rate report9 and shown in Figure 2-61. These are the same adjustments used for
heavy-duty gasoline vehicles in MOVES2014.
To evaluate these adjustment ratios for MOVES, we considered recent start emission rate data from
one heavy-duty gasoline truck. The data was gathered using PEMS using the procedure and
methods discussed in Section 2.2.3.2. The vehicle tested was a 2012 MY box truck with a gasoline
engine. Figure 3-23 shows the results from the testing as compared to the MOVES adjustments.
Because the trend in the soak time effects is similar to the values used in MOVES2014, and
because we only had new data from one truck, MOVES4 retains the start emission adjustment
ratios used in MOVES2014.
204

-------
HD Gasoline Truck Start Emission Ratio
100 200 300 400 500 600 700 S00
Soak Time (min)
—•— New Data NOx —•— New Data CO —•— New Data HC
—•—MOVES NOx —•— MOVES CO —•— MOVES HC
ro
4-1
on
3:
O
U
o
4-1
o
4—>
ro
QC
c
o
V>
—
E
LU
1.6
1.4
1.2
1.0
o.s
0.6
0.4
0.2
0.0
Figure 3-23 HD Gasoline Start Emission Ratio Compared to Recent Data
3.2.4 Start Energy Rates
The heavy-duty gasoline start energy rates were originally derived in MOVES2004, and updated in
MOVES2010a as described in the corresponding reports.57 Figure 2-32 displays the CO2 (g/mile)
emission rates for cold start (operating mode 108) calculated from the energy rates using the carbon
content of conventional gasoline as documented in the MOVES3 Greenhouse Gas and Energy
Report.3 As shown, there is substantial variability in the start rates between 1974 and 2000. As
discussed in Section 2.1.4.1, the detailed methodology used in MOVES2004 (which modeled
different emission rates according to vehicle weights, engine technologies, and engine sizes)
introduced variability into the energy rate within the current MOVES regulatory class emission
rates.
Table 3-26 displays the relative contribution of running and start operation to total energy
consumption from the heavy-duty gasoline regulatory classes from a national run for calendar year
2016. Like diesel vehicles, starts from gasoline vehicles are estimated to be a relatively small
contributor to the total energy demand of vehicle operation. Due to the small contribution to the
205

-------
total energy inventory, we have not prioritized updating the heavy-duty gasoline start emissions
rates.
Table 3-26 Relative Contribution of Total Energy Consumption from Each Pollutant Process by Regulatory
processID
processName
LHD2b3
LHD45
MHD
HHD
1
Running Exhaust
97.8%
99.2%
99.0%
99.2%
2
Start Exhaust
2.2%
0.8%
1.0%
0.8%
The HD gasoline start energy rates are reduced for shorter soak times using the same factors for
diesel vehicles, as presented in Table 2-51. The energy rates also increase with cold temperatures
using the temperature effects documented in the 2004 Energy Report.71
The start energy rates include the projected impact of the Phase 1 Heavy-Duty GHG standards,
which began phasing-in in 2014 and have the same reductions as the running energy rates, as
presented in Table 2-30 and Table 2-32. As discussed in Section 2.2.4, the start energy rates are not
projected to change due to the HD GHG Phase 2 standards.
400
re
to
5 300'
CD
¦«-»
re
DC
t
re
55
2
o
o
04
O
O
200
100'
1



'
rvw



vV




















Reg Class
•- 41-LHD2b3
»- 42-LHD45
46-MHD67
»- 47-HHD8
1980
2000
2020
2040
Model Year
Figure 3-24 Heavy-Duty Gasoline Cold Start CO2 Emission Rates (g/start) by Model Year and Regulatory Class
(OpmodelD 108)
206

-------
4 Heavy-Duty Compressed Natural Gas Exhaust Emissions
While natural gas lacks the ubiquitous fueling infrastructure of gasoline, compressed natural gas
(CNG), propane, and liquefied natural gas have grown as transportation fuels for public transit,
government, and corporate fleets. Such fleets typically utilize centralized, privately-owned
refueling stations. Fleet vehicles are operated as back-to-base, which means the vehicles return to
the same base location each day for refueling. Within this segment, some of the most prevalent use
of in CNG vehicles has occurred among city transit bus fleets and in solid waste collection or
refuse truck fleets.118 Figure 4-1 displays the fraction of heavy-duty CNG fueled-vehicles by source
type and model year estimated in the default national activity database in MOVES3.6
30.00%
o 25.00%
o
•
U
cz
£ 20.00%
.o
15.00%
ft
"3
-•—Intercity Bus
-e— Transit Bus
-•—School Bus
Refuse Truck
-•—Single Unit Sliort-haul Truck
-e— Single Unit Long-haul Truck
-•—Combination Short-haul Truck
0.00%
1980
1990
2000	2010
Model Year
2020
2030
Figure 4-1 Fraction of Heavy-Duty CNG fueled-vehicles in MOVES3 by Source Type by Model Year
MOVES2014 modelled only CNG transit buses. In MOVES3, we allow the modeling of CNG fuel
for most heavy-duty source types. Long-haul combination trucks (source type 62) are still diesel-
only in MOVES because of the difficulties in accurately modeling hotelling for non-diesel vehicles.
We hope to improve this in a future version of MOVES.
The CNG transit buses are mapped to the urban bus regulatory class. The CNG vehicles in other
heavy-duty source types are mapped to the HHD regulatory class ). However, the base emission
rates for the two regulatory classes are identical. Thus, any differences in CNG emissions between
source types is due to differences in population and activity.
Much of the analysis for CNG emissions, especially for older model years, is unchanged from
MOVES2014. Important updates in MOVES3 and later versions include:
207

-------
Two new model year groups, 2007-2009 and 2010+, to replace the 2007+ emission rates in
MOVES2014
•	Emissions rates for MY2010+ based on real-world CNG vehicle emissions data.
•	For pre-2010 model years, we still estimate emissions using vehicle certification data, but
we now use all HD CNG engine emissions data within a model year group. In
MOVES2014, the certification emission rate was limited to engine families classified as
urban bus.
No updates were made to these rates in MOVES4 since the average NOx FTP emission level for
MY 2010-2017 CNG engine families is already close to the HD2027 0.1 g/hp-hr standard and any
further reductions due to the rule are expected to be small.
As noted above for diesel and gasoline vehicles, MOVES methane emissions are not estimated
using emission rates. Rather, methane is estimated in relation to THC, using ratios stored in the
MethaneTHCratio table in MOVES. The ratios are categorized by fuel type, pollutant process,
source type, model-year group, and age group. MOVES multiplies the THC rate by the
corresponding ratio from the "methanethcratio" table to calculate the CFUrate. The methane
fraction from CNG vehicles is 89% and 96% for model year groups 1960-2001 and 2002-2060
respectively, as documented in the Speciation report.1
These emission rates are dependent on vehicle age, and thus are stored in the emissionRateByAge
table.
Total energy consumption is age independent, and therefore, stored in the EmissionRate table.
Some of the published studies did not report total energy consumption directly, so it was necessary
to compute energy from a stoichiometric equation based on the carbon content in the emitted
pollutants or from reported values of miles per gallon equivalent of diesel fuel. In the former case,
we used 0.8037 as the carbon fraction coefficient for non-methane hydrocarbons (NMHC) when
the bus was equipped with an oxidation catalyst and 0.835 without due to high ethene levels, using
speciation profiles from Ayala et al. (2003)119 discussed later in this section. All other conversion
factors to energy were taken from Melendez et al. (2005).123
On a similar note, MOVES does not report particulate matter (PM2.5) as a single rate; it reports one
rate for PM from elemental carbon (EC) of 2.5 microns or less, and another rate for non-elemental
carbon of 2.5 microns or less. These separate rates for PM (EC) and PM (NonEC) from the
emissionRateByAge table are added together for a total PM2.5 rate used for comparison to the
measurements.
208

-------
4.1 Running Exhaust Emission Rates
The pre-2010 running emission rates are relatively unchanged from MOVES2014PP, and are based
on cycle average rates as discussed in Section 4.1.1. The running exhaust emission rates for model
year 2010 and later CNG vehicles using second-by-second in-use emission measurements from
heavy-duty vehicles are discussed in Section 4.1.2.
4.1.1 1960-2009 Model Years
Ideally, MOVES modal emission rates would be developed through analysis of second-by-second
data of vehicles of the appropriate regulatory class, model year, and age. Unfortunately, such data
are not readily available for all model years.
In particular, data at multiple ages that can be used to determine emission deterioration, and
second-by-second data that can be used to establish STP trends was very limited for MY 2009 and
earlier CNG vehicles. Thus, for MOVES, we applied STP and age trends from MHD gasoline
vehicles to cycle-based certification results. The following sections describe the available data and
the methods to calculate the adjustment ratios.
4.1.1.1 CNG Chassis Dynamometer Measurements
Chassis data was collected from programs that were conducted at several research locations around
the country on heavy-duty chassis dynamometer equipment. In our analysis for MOVES2014, we
compiled 34 unique dynamometer measurements. Data from newer studies such as Clark et al.
(2007)130 would provide further validation and refinement to the rates discussed in this report,
however they have not been incorporated here.
The data considered consisted of distance-specific running emissions rates for each of the
following pollutants and total energy:
1.	oxides of nitrogen (NOx)
2.	carbon monoxide (CO)
3.	particulate matter (EC + non-EC)
4.	total hydrocarbons (THC)
5.	methane (CH4)
6.	total energy consumption
This data was collected on two driving cycles, the Central Business District (CBD) and
Washington Metropolitan Area Transit Authority (WMATA).
The CBD cycle is defined as a driving pattern with constant acceleration from rest to 20 mph, a
short cruise period at 20 mph, and constant deceleration back to rest, repeated for 600 seconds (see
pp The only change was to limit the certification data used to derive the 2007-2009 model year emission rates to the
2007-2009, rather than 2007-2017 as discussed in Section 4.1.1.2.3.
209

-------
Figure 4-2).120 The WMATA cycle was developed using GPS data from city buses in Washington,
DC, and has higher speeds and greater periods of acceleration than the CBD cycle (see Figure 4-3).

CBD








































































0	100	200	300	400	500	600
Time, s
Figure 4-2 Driving Schedule Trace of the Central Business District (CBD) Cycle121
60
£ 50
&
2 20
¦S
> 10
0
0	500	1000	1500	2000
Time (sec)
Figure 4-3 Driving Schedule Trace of the Washington Metropolitan Area Transit Authority (WMATA) Cycle123
Table 4-1 shows a summary of the number of unique CNG bus measurements by driving cycle for
each study. Navistar published a similar study of CNG and diesel buses in 2008, and this analysis
shares many of the same sources.122 All of the vehicles were in service with a transit agency at the
time of testing. The number of unique measurements are typically equal to the number of vehicles
tested and the measurements were typically reported as averages based on multiple runs with the
same vehicle and configuration over a specific driving cycle with the exception of measurements
reported by Ayala et al. (2002)125 and Ayala et al. (2003).119 In the Ayala et al. (2002) study the
2000 model year CNG bus was tested and then retested after approximately two months of service,
which we treated as independent measurements. Ayala et al. (2003) retested the same 2000 CNG
bus as in their previous study; however, the bus had accumulated an additional 35,000 miles and
was serviced by the OEM to be equipped with an oxidation catalyst that was later removed for
baseline testing. Ayala et al. (2003) conducted duplicate tests under each vehicle/aftertreatment
configuration, which we considered four independent measurements.
210

-------
Table 4-1. Summary of External Emissions Testing Programs by Driving Cycle and Number of Unique
	 Measurements and their Corresponding Model Years 	
Paper/Article
Lead Research Unit
Driving
Cycle(s)
Model Year
(Number of
Measurements)
Melendez 2005123
National Renewable Energy Laboratory
(NREL)
WMATA
2001 (4),
2004 (3)
Ayala 2003119
California Air Resources Board (CARB)
CBD
2000	(4),
2001	(2)
LeTavec 2002124
Atlantic Richfield Company (ARCO)
CBD
2001 (1)
Ayala 2002125
CARB
CBD
2000 (2)
Lanni 2003126
New York Department of Environmental
Conservation
CBD
1999 (3)
McKain 2000127
West Virginia University (WVU)
CBD
1999 (3)
Clark 1997128
wvu
CBD
1996(10)
McCormick
1999129
Colorado School of Mines
CBD
1994 (2)
TOTAL

(34)
4.1.1.2 Determining Model Year Groups
Model year groups are intended capture differences in vehicles over time while still being
manageable from a computational viewpoint. Model year groups are defined based on availability
of measurement data (see Table 4-1), emissions standards, and/or new vehicle technologies that
affect real-world emissions.
4.1.1.2.1 1994-2001 Model Years
We evaluated the measured NOx, CO, PM2.5, and THC emission rates to establish model year
groups and chose to group all the CBD measurements from the literature into one model year
group, spanning from MY 1994 to MY 2001. Note that we decided to exclude one of the studies
that had four MY 2001 buses tested on the WMATA cycle from this part of the analysis. This was
done because inclusion increased the complexity of analysis by having to deal with two driving
cycles within a model year group while providing only an incremental increase in sample size.
211

-------
4.1.1.2.2
2002-2006Model Years
Of the surveyed data, only one study had vehicles newer than MY 2001.qq'130 This paper, a joint
study between NREL and WMATA, had three MY 2004 vehicles. The MY 2004 vehicles have a
visibly different emissions profile than the other vehicles. While these buses were only tested on
the WMATA cycle, they were all equipped with oxidation catalysts and had substantially lower
emissions, particularly for PM2.5, compared to the 1994-2001 buses tested on the CBD cycle. As a
result, we created a model year group from MY 2002 to MY 2006 based on the MY 2004 buses
tested on the WMATA cycle. This MY group ends before MY 2007 when a new series of stringent
emission standards went into effect, as described below.131
4.1.1.2.3 2007-2009 Model Years
MOVES2014 had a single set of emissions for 2007-and later buses. In MOVES3, we created two
groups, MY 2007-2009 and MY 2010+ (noted as MY 2010-2017 when comparing certification
data). We decided to split the groups in this way because: (a) changes to fscale values starting MY
2010 (see Appendix G) requires rates to be re-analyzed using 1 hz data; (b) the HDIUT data set
includes real world data on MY 2010+ CNG vehicles; (c) certification data showed a significant
difference between the average emissions rates for NOx and CO between these two model year
groups (but note that certification data is not used in developing the rates for MY2010+), and (d)
this allows for better representation of differences in combustion and aftertreatment technology,
such as stoichiometric-combustion with three-way catalysts (TWC) that became more prevalent
starting year 2010.
Certification emission data for natural gas heavy-duty vehicles are publicly available by model year
on the EPA's Office of Transportation and Air Quality website.132 Analysis of these data showed
that from MY 2002 to MY 2017, there have been changes in average certification levels for all the
pollutants considered in this report. In particular, NOx and PM2.5 levels have dropped dramatically.
This effect is largely attributable to increasingly stringent emission standards, which have affected
both diesel and CNG engines.
Emission rates from analysis of certification data and number of CNG engine families in the
certification data are shown in Table 4-2 below. The current, and historically most stringent, heavy-
duty compression-ignition NOx standard of 0.20 g/bhp-hr was fully phased in by 2010 and MY
2010+ heavy-duty CNG engines are required to meet this standard (even if they are not
compression-ignition). Thus, the average NOx certification value for the MY 2010-2017 group is
considerably lower compared to the MY 2007-2009 group. At the same time, and mostly to meet
the new NOx standard, heavy-duty CNG engines transitioned from lean-burn to stoichiometric-
combustion with TWC. This technology transition is the likely reason for the increase in THC and
CO certification emissions rates from MY 2007-2009 to MY 2010-2017.
qq Several papers have discussed more recent vehicles. Examples include Clark et al. (2007).129 Data from these newer
studies would provide further validation and refinement to the rates discussed in this report, however, time and
resources were not available to complete a re-analysis for MOVES3.
212

-------
The differences in MOVES emission rates across all model years are discussed below in Section
4.1.3.
Table 4-2 Model Year Group Based Certification Emission Rate for Heavy-Duty CNG Engine Families
Model Year
Number of Engine
Certification Emission Rate (g/bhp-hr)c
Group
Families3'13
NOx
CO
pm25
NMHCd
2002-2006
22
1.208
1.355
0.0078
0.147
2007-2009
30 (24 for PM2 5)
0.6123
1.940
0.0042
0.063

159 for NOx and CO, 153




2010-2017e
for THC, and 120 for
pm25
0.1051
4.413
0.0028
0.044
Notes:
a For MY 2002-2006, the number of engine families is based on HD CNG urban bus regulatory class. For MY
2007-2009 and MY 2010-2017, the number of engine families is based on all HD CNG engine families.
b Some engine families did not report emission data for THC and/or PM2 5.
0 MY 2002-2006 group emission rates are projected sales weighted average of HD CNG urban bus certification
emission rates. MY 2007-2009 and 2010-2017 group emission rates are simple average of all HD CNG
certification emission rates (no weighting for projected sales).
d Certification data has measurements of organic material non-methane hydrocarbon equivalent (OMNMHCE).
For this analysis they were treated as NMHC values.133
e Only shown for comparison. Certification data for MY 2010-2017 is not used in developing MY 2010+ rates,
which are based on MY 2010+ CNG vehicles in the HDIUT data set.
4.1.1.3 Creating Comparable MO VES Gasoline Emissions
Section 1.6 explains how MOVES operating modes relate to scaled tractive power (STP). Because
we lacked data on age and STP trends for pre-2010 CNG vehicles, we applied adjustments based
on the rates for gasoline MHD vehicles. To do this, we compared CNG emission data collected on
the CBD and WMATA cycles to what MOVES estimated for MHD gasoline vehicles on those
same cycles.
Because the pre-2009 CNG vehicles form a small and diminishing portion of the MOVES3 fleet,
we have not updated this analysis which relied on emissions, vehicle and activity information from
MOVES2010b.
This approach requires converting activity on the CBD and WMATA bus driving cycles to
MOVES operating mode distributions, and then simulating MHD gasoline emissions on those same
operating mode distributions.
4.1.1.3.1 Operating Mode Distributions for Transit Bus Drive Cycles
The MOVES2010b project level importer was used to input the second-by-second drive cycle for
the CBD and WMATA drive cycles. For each, a single link was created, with the test cycle entered
as a drive trace. Running MOVES2010b generated the operating mode distribution, which is
created by allocating the time spent in each operating mode according to the cycle speed and
acceleration, as shown in Figure 4-4 and Figure 4-5. The derivation of scaled tractive power (STP)
and operating mode attribution for heavy-duty vehicles are discussed earlier in this report, in
Section 1.5. Road grade is set to zero because these are chassis dynamometer runs.
213

-------
Since STP is dependent on mass (among other factors), the average vehicle inertial test mass for
each cycle was inserted into the MOVES2010b sourceUseType table in place of the default transit
bus mass to ensure a more accurate simulation— 14.957 metric tons for the CBD and 16.308 metric
tons for the WMATA, compared to the MOVES2010b default of 16.556 metric tons. The STP
calculations used the road-load coefficients from MOVES2010b for transit buses, assuming the
coefficients (A, B, and C) were similar to those of the tested buses.

350

300

250
u


200



-------
•	fuel type,
•	driving cycle,
•	age group,
•	regulatory class,
•	model year, and
•	pollutant and process.
„	ZioM Rp,OM * T0M,cycle
Ep,simcycie	IT	Equation 4-1
D
cycle
Where:
D cycle distance of the cycle, in miles
Rp,om = emission rate of pollutant p in operating mode OM, in g/hr
Tom,cyck = time spent in operating mode OM for given cycle, in hr
We compared the MOVES2010b simulated MHD gasoline rates with the published chassis
dynamometer measurements. We also specified the age group and model year to match individual
vehicles in the testing programs from the literature on CNG transit buses.
4.1.1.4 Emissions Rates by Model Year Group
To estimate emission rates for pre-2010 CNG vehicles, we applied STP and age trends from MHD
gasoline vehicles to cycle-based CNG certification results. Mathematically, this is the same as
applying a CNG adjustment to MHD gasoline emission rates, which is how the analysis is
described in the sections below.
Due to limited data on older vehicles in the literature, the ratios (shown in Table 4-3) developed
using vehicles in the 0-3 age group have been applied to all other age groups. In addition, we
assumed that CNG vehicles exhibit the same deterioration trend as medium heavy-duty gasoline
trucks (Table 3-3 in Section 3.1.1.1.2 for THC, CO and NOx, and Section 3.1.2.1.3 for PM2.5 ).
4.1.1.4.1 1960-2006Model Years
The operating mode based emissions rates for MOVES2010brr MHD gasoline vehicles were
adjusted by the ratio of cycle-average emissions rates from chassis dynamometer measurements to
simulated cycle modeling (see Section 4.1.3). For MY 1994-2001 and MY 2002-2006, the
adjustment ratios were based on the CBD cycle and WMATA cycle, respectively.
For each model year group, a central model year was selected as the source for the MHD gasoline
operating mode based rates. For MY group 1994-2001, we used MHD gasoline rates from MY
1997 because it is one of the median years in the group. Alternatively, we could have used the other
median year, MY 1998. Even though the average rate for MY 1998 was significantly lower (44
percent of that of MY 1997), based on Equation 4-2, we expect minimal differences in the final
estimated CNG rates (Rcng term) whether we use MY 1997 or MY 1998 as the median year since
the lower operating mode rates (Rmdg term) will lead to lower simulated cycle-average rate (Emdg
term), which in turn will lead to larger adjustment ratio (Ecng/Emdg). For MY group 2002-2006,
" The PM exhaust emission rates were subsequently updated from MOVES2010b for MHD gasoline vehicles for pre-
2010 model years.
215

-------
we used MHD gasoline rates from MY 2004 because that was the model year of the engine in each
of the CNG vehicles measured on the chassis dynamometer (the MY 2001 vehicles were not
included in this group). See Equation 4-2 and Equation 4-3 for MY groups 1994-2001 and 2002-
2006, respectively.
We assumed that the MY 1993 and earlier CNG vehicles have the same emission rates as MY
group 1994-2001.
n	n	ECNGjCBDj1994_2ooi	Equation 4-2
kCNG,OM, 1994-2001 — KMDG,OM,1997 * ~B	
^MDG,simCBD,1997
D	_ D	ECNGiWMATAi2004	Equation 4-3
KCNG,OM,2002-2006 — KMDG,OM,2004 * 	
^MDG,simWMATA,2004
Where:
Rcng,om,myg = operating mode based emissions rate for CNG vehicles for model year group (MYG) 1994-
2001 or MY 2002-2006, in g/hr
Rmdg,om,my = operating mode based emissions rate for MHD gasoline vehicles for model year 1997 or 2004
(corresponding to MYG), in g/hr
Ecng,cycle,myg = Chassis dynamometer cycle-average emissions rate for MY 1994-2001 or 2004 CNG buses
tested on a CBD or WMATA cycle, respectively, in g/mile. See Table 4-3.
EMDG,simCycie,MY = Simulated cycle-average emissions rate for MY 1997 or 2004 MHD gasoline vehicles for
CBD or WMATA cycle, respectively, in g/mile. This cycle-average rate is calculated using
the Rmdg,om,my operating mode rates. See Table 4-3.
4.1.1.4.2 2007-2009 Model Years
Due to lack of published data on MY 2007-2009 in-use vehicles, we used certification emissions
rates, shown in Table 4-2, to scale the operating mode based emissions rates. Certification
emissions rates are reported in grams per brake horsepower-hour (g/bhp-hr) and are not directly
used in formulating MOVES emission rates because they do not include real-world effects such as
deterioration134 which were present in the chassis dynamometer measurements used to estimate
emissions rates for MY 1994-2001 and MY 2002-2006. So, we created scaling factors that we
could apply to the MY 2002-2006 emissions rates to estimate rates for MY 2007-2009. This scaling
factor is the right-most term in Equation 4-4 shown below.
n	_ n	ECNGiWMATAi2004 CCNG,2007-2009	Equation 4-4
kCNG,OM,2007-2009 — KMDG,OM,2004 * "jj	* r	
^MDG,simWMATA,2004 LCNG,2002-2006
Where:
Ccng,2007-2009 = Average certification emission rate of all heavy-duty CNG engine families of model year MY
2007-2009 in g/bhp-hr
Ccng,2002-2006 = Projected sales weighted average certification emission rate for CNG urban bus engine families
in MY 2002-2006, in g/bhp-hr
216

-------
The adjustment ratio for energy consumption for MY 2002-2006 (Equation 4-3) is applied to all
model years in 2007-2009. For MY 2007+, we did not scale the energy consumption rates like we
did for other pollutants (Equation 4-4) because even though we have certification data on CO2
emission rates for 2007-2009 model years, we do not have certification data on CO2 emission rates
for MY 2002-2006. As a result, MY 2007-2009 energy consumption rates are identical to the MY
2002-2006 rates.
217

-------
4.1.1.4.3 Ratio Summary
Table 4-3 Ratios Applied to MHD Gasoline Rates to Compute CNG Rates
Ecng, Cycle-Average Chassis Dynamometer Measurement Rates (g/mile)
MY
Age
Group
Cycle
NOx
CO
PMNonEC
PM EC
THC
TOTAL
ENERGY
(BTU/mi)
1994-2001
0-3
CBD
20.8
9.97
0.037
0.0038
13.2
42782
2002-2006
0-3
WMATA
9.08
2.17a
0.0039
0.0005
11.2
40900
Emdg , Simulated Cycle-Average Medium Heavy-Duty Gasoline Rates (g/mile)
MY
Age
Group
Simulated
Cycle
NOx
CO
PMNonEC
PM EC
THC
TOTAL
ENERGY
(BTU/mi)
1997
0-3
CBD
9.63
62.4
0.0024
0.0002
1.84
31137
2004
0-3
WMATA
5.45
18.9
0.0035
0.0003
1.43
35489
Ratios Applied to the Medium Heavy-Duty Gasoline Rates to Create CNG Rates
MY
Age
Group
MHD
Gasoline
MYb
NOx
CO
PM_NonECf
PM EC
THC
TOTAL
ENERGY
1994-2001c
all
1997
2.16
0.160
15.5
21.6
7.17
1.37
2002-2006
C
all
2004
1.67
0.115
1.09
1.87
7.79
1.15
2007-2009
d
all
2004
0.842
0.157
0.587
1.01
3.34
1.15
2010+®
Age 0- 3 (and 4-5) rates are based on analysis of 1 hz data from MY 20
vehicles in the HDIUT data set. Ages 6+ apply deterioration factors to a
as described in main text.
0+ CNG
ge 0-3 rates
Notes:
a The measured CO rate (0.14 g/mi) was uncharacteristically low and thus determined to be an outlier and not used.
Each of the three post-2001 vehicles in this study had the same MY 2004 engine (John Deere 6081H). This
engine's CO certification rate was a full order of magnitude lower than certification rate of other MY 2004 engine
models, and was not supported by additional test results. We adjusted the WMATA chassis dynamometer CO rate
by the ratio between the sales-weighted average CO certification level of all MY 2004 CNG engine models and the
CO certification level for the MY 2004 John Deere 6081H engine.
b Model year of the medium heavy-duty gasoline operating mode rates to which the pollutant-specific ratios are
applied
0 The ratios are calculated using Equation 4.2 or Equation 4.3 and the ECNG and EMDG values in this table
dThe ratios are calculated using Equation 4.4, the Ecng and Emdg values in this table, and the Ccng values in Table
4.2
"Energy consumption rates for MY 2014-2017 and MY 2018+are reduced as per heavy-duty GHG Phase 1 and
Phase 2 rules, respectively. See main text for details.
f The PM2.5 exhaust emission rates have been subsequently updated in MOVES3 from MOVES2010b for MHD
gasoline vehicles for MY 1997. The ratios presented here are applicable to the MOVES2010b MHD gasoline rates
and the MOVES3 CNG emission rates.
218

-------
4.1.2 2010-2060 Model Years
Running emission rates for MY 2010 and later CNG vehicles were based on information from in-
use trucks, and thus, unlike the calculations for earlier model years, the rates by operating mode
could be calculated directly. We then applied factors to account for deterioration with vehicle age,
and adjustments to energy rates to account for the phase-in of heavy-duty greenhouse gas
standards.
4.1.2.1	Base Emission Rates
Running emission rates for MY 2010 and later CNG vehicles were based on information from in-
use trucks. To develop MY 2010+ emissions rates (for THC, CO, NOx, and PM2.5) and energy
consumption rates, we used the MY 2010+ CNG vehicles in the HDIUT data set. At the time of
analysis, there were five MY 2011 CNG vehicles and six MY 2014 vehicles. These 11 vehicles are
all stoichiometric-combustion with TWC and are certified at or below the 0.20 g/bhp-hr standard.
After quality assurance, the 1 Hz data set included about 310,000 seconds of operation. Operating
modes (Table 1-4) were assigned to the 1 Hz data using the method to calculate STP described in
section 2.1.1.3. The analysis used updated fscale values described in section 2.1.1.4.2 and Appendix
G and thus, there was no need for hole-filling of missing operating modes. The operating mode-
based rates were calculated using fscaie = 10. The rates for regClass 47 and 48 are identical.
Unlike the analysis method for HD diesel (described in section 2.1.1.5), the method for HD CNG
did not use the NOx FEL based grouping since all 11 vehicles are in the same NOx FEL group. As a
result, the zero-mile (age 0) THC, CO, NOx, and PM2.5 rates for CNG are identical for all model
years starting 2010 (unlike HD diesel where they change for each model year in 2010-2015 based
on production volume differences between the NOx FEL groups).
4.1.2.2	Age-based deterioration factors
THC, CO, and NOx age-based deterioration factors for MY 2010+ CNG vehicles in MOVES3 are
unchanged from MOVES2014. In MOVES2014, these factors were set as equal to the factors for
MY 2010+ HD gasoline vehicles, which in turn are identical to and based on MY 1960-2007 HD
gasoline vehicles. There is no deterioration for age groups 0-3 and 4-5 and the deterioration factor
(per operating mode) is same across age groups for ages 6+ but varies between operating modes
within an age group. These deterioration factors are described in Table 3-3 in Section 3.1.1.1.2.
For PM2.5, in MOVES3, ages 0-3 and 4-5 have no deterioration and the MOVES2014 light-duty
PM2.5 deterioration factor for age 6-7 is applied to all CNG PM2.5 emission rates for ages 6+, thus
making the PM2.5 and gaseous pollutant methods more (but not fully) aligned. Note that, unlike the
factor for gaseous pollutants, the MOVES PM2.5 deterioration factor does not vary between
operating modes for a given age group. See Section 3.1.2.1.3 for more details and Table 4-4 for a
comparison between MOVES3 and MOVES2014.
219

-------
Table 4-4 Age-based Deterioration Factor for PM2.5 Emission Rates for HHD and Urban Bus CNG Vehicles in
Model Year 2010+
Age
MOVES2014
MOVES31
0-3 (Baseline)
1.00
1.00
4-5
1.57
1.00
6-7
1.75
1.75
8-9
1.96
1.75
10-14
2.38
1.75
15-19
3.14
1.75
20+
4.15
1.75
Note:
1 When recreating the deterioration factor, for age 6+, from the age-group based default emissions rates in the MOVES
database, the ratios will not be exactly 1.75 because the final rates (with deterioration factors already applied) are
rounded to a set precision before submission to the database.
4.1.2.3 Application of Heavy-Duty Greenhouse Gas Phase 1 and Phase 2 Rules
To model energy consumption in MOVES3, we split the CNG MY 2010+ group into MY 2010-
2013, MY 2014-2017, and MY 2018+ groups. The MY 2010-2013 energy consumption rates are
identical across these model years and based solely on the HDIUT data set analysis. For MY 2014-
2017, the CNG energy consumption rates of MY 2013 are reduced by the percentage reduction
assigned to HHD vehicles in the Greenhouse Gas Emissions Standards and Fuel Efficiency
Standards for Medium and Heavy-Duty Engines and Vehicles Phase 1 rule59 (see Table 2-30).
Similarly, for MY 2018 and later, using MY 2017 rates as base year, the energy consumption rates
of CNG vehicles are further reduced as per the Greenhouse Gas Emissions and Fuel Efficiency
Standards for Medium and Heavy-Duty Engines and Vehicles — Phase 2 rule61 (see Table 2-31).
Note that the Phase 1 reduction for CNG vehicles is identical across all allowed source type and
regulatory class combinations. However, for the Phase 2, different reductions for CNG vehicles are
applied by source type and regulatory class (see Table 2-31). The anticipated improvements in fuel
efficiency from the Phase 2 rules are stored in the EmissionRateAdjustment table.
The GHG Phase 1 and 2 reductions reflect the percent improvements projected from the rules
based on engine technology improvements to diesel engines. In making these projections, we
assumed the HD GHG rules lead to the same reductions in the energy rates for CNG vehicles as for
heavy-duty diesel. In reality, manufacturers of CNG vehicles can meet the standards by lowering
both CH4 and CO2 emissions, and the reductions in fuel consumption (and CO2 emission rates)
between CNG and diesel vehicles will likely differ. Future MOVES versions may update the
energy consumption rates and CH4 emission rates with data from MY 2015 and later CNG vehicles
that comply with the GHG standards.
4.1.3 Model Year Trends
Figure 4-6 through Figure 4-10 display the THC, CO, NOx, PM2.5, and CO2 running exhaust
emission rates by model year and regulatory class (HHD and Urban Bus). The emission rates are
estimated in grams per mile (g/mile) using nationally representative operating mode distributions
and average speeds. The change in emissions at MY 2010 coincides with both a change in our
220

-------
analysis methodology and a shift in the CNG vehicle fleet from lean-burn combustion to
stoichiometric combustion with a three-way catalyst.
Figure 4-6 shows a significant increase in the THC emissions between the MY 2007-2009 and MY
2010 and later vehicles. Because MOVES uses the same methane fraction for 2002 and later CNG
vehicles, the CHtand non-methane hydrocarbons (NMHC) emissions follow the same trend. The
significant increase in THC, CH4. and NMHC starting in MY 2010 with the increased penetration
of CNG vehicles with stoichiometric-combustion engines and three-way catalyst (TWC) is not
supported by the certification data presented in Table 4-2 nor in recent studies comparing
stochiometric and lean-burn combustion CNG engines.135 136 The differences in the methodologies
and limitations in the pre-2010 data likely contributed to apparent increase in THC emissions
starting in 2010 model year in MOVES.
10.0'
7.5
0)
1
o>
r 5.0
+-»
ra
U.
0
1
2.5
0.0
Reg Class
¦ 47-HHD8
48-Urban Bus
1980
2000
2020
2040
Model Year
Figure 4-6. Base running emission rates for THC from age 0-3 CNG heavy-duty vehicles averaged over a
nationally representative operating mode distribution
Like THC, CO also shows a significant increase in emission rates with the updated analysis of the
MY 2010+ vehicles (See Figure 4-7). However, the increase in CO is supported by certification
data (Table 4-2) and in more recent testing comparing stoichiometric-combustion with TWC based
CNG buses.135-136
221

-------
»»»ns
Reg Class
¦ 47-HHD8
48-Urban Bus
1980
2000
2020
2040
Model Year
Figure 4-7. Base running emission rates for CO from age 0-3 CNG heavy-duty vehicles averaged over a
nationally representative operating mode distribution
Figure 4-8 shows consistent decreases in NOx emission rates from older to new model years. The
trends in NOx emissions are consistent with the certification data (Table 4-2) and recent
studies.135136
222

-------
15
10
E
3
m
rs
DC
X
O
Reg Class
¦ 47-HHD8
48-Urban Bus
1980
2000
2020
2040
Model Year
Figure 4-8. Base running emission rates for NOx from age 0-3 CNG heavy-duty vehicles averaged over a
nationally representative operating mode distribution
Figure 4-9 shows significant decreases in the PM2.5 emission rates between the 2001 and 2002
model year emission rates based on the chassis-cycle average PM2.5 emission rates. An increase in
PM2.5 emission rates is shown from model year 2009 and 2010, which is inconsistent with the
certification data (Table 4-2). The two studies evaluating stoichiometric and lean-burn showed
mixed results, with one study showing stoichiometric engines with TWC emitting lower PM2.5
emissions136 and the other study showing stoichiometric engines emitting higher PM2.5 rates.135
223

-------
0.08'
"ST 0.06'
0.04'
0.02'
0.00'
Reg Class
¦ 47-HHD8
48-Urban Bus
1980
2000
2020
2040
Model Year
Figure 4-9. Base running emission rates for PM2.5 from age 0-3 CNG heavy-duty vehicles averaged over a
nationally representative operating mode distribution
Figure 4-10 shows general decreases in the CO2 emission rates across model years, including the
impact of the HD GHG Phase 1 and Phase 2 rules discussed in 4.1.2.3.
224

-------
1500
0)
£

-------
5 Heavy-Duty Ammonia Emissions
The ammonia (NH3) running exhaust emission rates for heavy-duty diesel, heavy-duty gasoline,
and heavy-duty compressed natural gas fueled vehicles have been updated in MOVES4. We do not
estimate NH3 emissions from starts. Extended idle emission rates are only estimated for gasoline
and CNG as discussed in the following sections.
Ammonia is not formed during typical combustion processes in diesel engines but is emitted as a
undesirable byproduct of selective catalytic reduction (SCR) aftertreatment systems present on
model year 2010 and later heavy-duty diesel vehicles to meet the 2010 NOx emission standards.
The SCR system functions by injecting aqueous urea into the exhaust flow, which then thermally
decomposes to NH3 and CO2 (Equation 5-1). NOx is then reduced to N2 by reactions shown in
Equation 5-2 through Equation 5-4.137
Excessive urea injected into the exhaust stream can lead to ammonia passing through the
aftertreatment system into the atmosphere, referred to as ammonia slip. NH3 oxidation (AMOX)
catalysts can be used to reduce ammonia slip but may not eliminate the emissions entirely.138'93
Consistent with previous versions of MOVES, we estimate zero heavy-duty diesel NH3 start
emissions. We also estimate zero NH3 extended idle emissions. NH3 emissions during extended
idling should be minimal, due to little or no dosing of urea in the SCR system because current
technology diesels cannot maintain the required operational exhaust temperature at extended low
loads.51 We do not estimate NH3 emissions from auxiliary power unit exhaust, which are not
expected to have SCR systems. The remainder of this section discusses ammonia running
emissions.
Ammonia measurements were not part of the HDIUT program we used to estimate the heavy-duty
diesel and CNG THC, CO, NOx, and PM2.5 emission rates as described in Section 2.1.1.1 As such,
we relied on other data sources for developing heavy-duty NH3 emission rates for diesel vehicles.
We compared fleet-average heavy-duty vehicle NH3 emission rates reported in the literature to in-
use studies as summarized in Table 5-1.
5.1 Heavy-Duty Diesel
CO(NH2)2 + H20 -> 2NH3 + co2
4NH3 + 4N0 + 02 -> 4N2 + 6H20
Equation 5-1
Equation 5-2
Equation 5-3
Equation 5-4
4NH3 + 6NO -> 5N2 + 6H20
2NH3 + NO + N02 -> 2N2 + 3H20
226

-------
Table 5-1. Fleet-average fuel-based NH3 emission rates (± 95% Confidence Intervals) from heavy-duty vehicles
			reported from recent studies		
Study
Study Year
Location
Number of vehicles
Heavy-duty vehicle fleet
average NH3 emission rate
(g/kg-fuel)
Preble et al.
(2019)139
2018
Caldecott Tunnel near
Oakland, CA
1,186
0.10+0.03
Haugen et al.
(2018)142
2017
Peralta Weigh Station near
Anaheim, CA
1,844 (HDV)
471 (MDV)
1,408 (high)
907 (low)
0.09 + 0.02 (HDV)
0.06 + 0.05 (MDV
0.08 ± 0.02 (high)
0.06 + 0.05 (low)
Bishop et al.
(2022)143
2020
Perry Weight Station Salt
Lake City, UT
1,591 (HDV)
103 (MDV)
1,053 (high HDV)
538 (low HDV)
0.08 + 0.06 (HDV)
0.22 + 0.23 (MDV)
0.009 + 0.009 (high HDV)
0.23 + 0.02 (low HDV)
Wang et al.
(2019)144
2015
Fort McHenry Tunnel,
Baltimore, Maryland
NA
0.10 + 0.07 (winter)
0.03 + 0.08 (summer)
Preble et al. (2019)139 measured NH3 emissions rates from heavy-duty vehicles at the Caldecott
Tunnel near Oakland, California in 2018. They sampled the concentrations of NH3 and CChfrom
the exhaust plumes of individual heavy-duty vehicles as they entered the tunnel at a 4% grade
traveling between 30 and 75 mph. From the NH3 and CO2 concentrations, they estimated NH3 fuel-
based emission rates using the carbon content of diesel fuel. By matching license plate images to
state truck registration databases, they were able to obtain vehicle information, including engine
model year and aftertreatment system. The average emission rates by different mode year and
aftertreatment groups are shown in Table 5-2.
Preble et al. (2019) measured NH3 emissions from 2010 and later trucks equipped with SCR
systems. They were able to collect over 900 diesel truck NH3 emissions measurements identified
by engine model year and aftertreatment system. Collecting a large sample is important for
capturing the fleet-average emision rates, because 10% of trucks contributed 95% of the total fleet
NH3 emissions. The Preble et al. (2019) study measured a large number of model year 1994-2006
retrofit DPF trucks due to the large number of drayage trucks servicing the nearby Port of Oakland.
Between 2010 and 2012, all California drayage trucks were required to be equipped with diesel
particulate filters.140 The average ammonia emission rates for the pre-2010 model year groups are
low and uncertain, which is consistent with measurements of heavy-duty ammonia emissions made
from the Caldecott Tunnel in 2006.141 The fleet-average heavy-duty diesel emission rates for Preble
et al. (2019) are shown in Table 5-1. The sample size is larger than that listed in Table 5-2 because
the fleet-average includes all heavy-duty diesel vehicles measured, including trucks that were not
matched to the vehicle registration database.
227

-------
Table 5-2. Fuel-based NH3 emission rates (± 95% Confidence Intervals) from heavy-duty vehicles by
aftertreatment and engine model year measured at the Caldecott Tunnel by Preble et al. (2019)139
Aftertreatment
Engine Model Year
NH3 (g/kg) fuel-based
emission rate
Number of
vehicles
Model year
ranges used
in MOVES
DPF + SCR
2010-2018
0.18 + 0.07
547
2010-2060
DPF
2007-2009
0.00 + 0.01
181
2007-2009
Retrofit DPF
1994-2006
0.01+0.01
114
Not used
No DPF
2004-2006
0.00 + 0.01
24
2004-2006
No DPF
1965-2003
0.02 + 0.02
62
1960-2003
The University of Denver's research group has conducted two studies of heavy-duty diesel
emissions measured using their remote sensing device called the Fuel Efficiency Automobile Test
(FEAT) system that measures pollution concentrations across the roadway. Haugen et al. (2018)142
measured NH3 emissions from heavy-duty vehicles at the exit ramp of the Peralta Weigh Station
near Anaheim, California. They separately sampled emissions from heavy-duty vehicles with the
elevated exhaust pipes ("high" in Table 5-1), and ground-level exhaust pipes ("low" in Table 5-1).
Additionally, they classified samples into medium-duty (defined as vehicles with GVWR < 26,000
lbs, or class 2 through 6 vehicles), and heavy-duty vehicles (GVWR > 26,000 lbs, or Class 7 and 8
vehicles). The "low" sample has a high percentage of medium-duty vehicles which had lower NH3
emission rates, and newer heavy-duty vehicles which have higher than average NH3 emission rates.
The heavy-duty and medium-duty vehicles measured in Haugen et al. (2018), are over 99% and
92% diesel vehicles, respectively, with the remainder being compressed natural gas vehicles. The
research measured a large increase in fleet-average NH3 emissions in the 2017 campaign compared
to previous measurements made at the Peralta Weight Station in 2008, 2009, 2010 and 2012, due
the penetration of SCR-equipped vehicles into the in-use fleet. The average for the 2010 and later
chassis model years from this study was 0.14 g/kg/fuel, while the older model year vehicles had
NH3 rates near zero.
Bishop et al. (2022)143 measured emissions from heavy-duty vehicles at the Perry, Utah Port of
Entry (-50 miles north of Salt Lake City) in December 2020. Bishop et al. (2022) separately
measured heavy-duty vehicles with "high" and "low" exhaust pipe positions. In this study, the
"low" exhaust tailpipe trucks were almost exclusively 2011 and later heavy-duty vehicles trucks,
which had significantly higher NH3 than the older trucks included in "high" exhaust tailpipe group.
The "low HDV" group had a mean emission rate similar to that estimated from Preble et al. 2019
for DPF+SCR equipped trucks. Bishop et al. (2020) measured a smaller number of Class 4, 5, and
6 vehicles, for which the mean estimate was highly uncertain.
Wang et al. (2019)144 measured fleet average NH3 emissions for a week in both February and
July/August, 2015 from the Fort McHenry Tunnel, which is along the 1-95 corridor with a traffic
volume of-55,000 vehicles per day. They measured emission concentrations from two of the four
bores that contain heavy-duty vehicle traffic. The tunnel includes a -1.8% down grade, followed by
a 3.3%) positive grade to the exit of the tunnel. Using the measured concentration and the fraction
of heavy-duty vehicles, they estimated fuel-based NH3 emission rates for both light-duty and
heavy-duty vehicles. Because this was the earliest study conducted, we would expect the NH3
emission rates to be lowest from this study due to a smaller fraction of SCR-diesel vehicles present
228

-------
in the fleet. The summer measurements are lower than the other studies, however, the winter
measurements are similar. In addition, the difference is not statistically significant due to the large
confidence intervals of the mean.
Despite the different measuring systems, locations, and sampling years, the fleet-average emission
rates are statistically similar among the different studies. This provides confidence that the fuel-
based emission rates reported from the studies, are not strongly impacted by measurement methods,
or the sampling conditions of the location.
We developed heavy-duty NH3 emission rates in MOVES using the reported fuel-based emission
factors by model year and aftertreatment class from Preble et al. (2019) reported in Table 5-2. We
chose to use Preble et al. (2019) because they reported the emission rates exclusively for heavy-
duty vehicles by model year ranges. In addition, we also used the Preble et al. (2019) study to
update the MOVES N0/N02 fractions (Section 7.1) and N2O emission rates.3
To develop MOVES heavy-duty diesel emission rates by regulatory class, model year, and
operating mode, we multiplied the MOVES3 heavy-duty diesel vehicle fuel-consumption rates by
regulatory class, model year, operating mode (Fuel RatesRegiMYi0p) by the Preble et al. (2019)
fuel-based NH3 emission rates (FERModei Year Group) from Table 5-1 shown below in Equation
5-5.
ERReg,MY,age,op Fuel Rcit6SRegjyiy^op ^ FER]y[odel Year Group	Equation 5-5
Figure 5-1 shows example NH3 emission rates for the LHD2b3 and HHD regulatory class for
model year 2017. Even though the fuel-based emission rate is the same, the gram per hour rate is
larger for the HHD regulatory class due to higher fuel consumption rates.
We replicated the NH3 emission rates for each heavy-duty regulatory class and model year across
all vehicle age (ages 0 to 30). This differs from the ammonia rates for light-duty gasoline NH3
where we had a much larger data sample and were able to estimate age effects. Preble et al. (2019)
collected measurements of NH3 in only their most recent campaign. Haugen et al. (2018) collected
several measurement campaigns at the Peralta Weight station, but had a limited number of vehicles
from which to estimate both model year and age specific emission rates. We recommend future
studies to evaluate the impact of aging, deterioration and mal-maintenance on NH3 from heavy-
duty diesel vehicles.
229

-------
12-
I 9"
JZ
d)
d)
ro
i_
c 6
0
If)

-------
5.2.1 1960-1980 Model Years
The model year 1960-1981 heavy-duty gasoline ammonia emission rates are unchanged from
MOVES2010 and documented in a MOVES2010 technical report.145 These rates were estimated by
scaling the light-duty gasoline ammonia emission rates by the ratio of light-duty gasoline emission
rates from model year 1981-1991 and heavy-duty gasoline vehicles measured in a 1983 EPA
study.146
5.2.2 1981-2060 Model Years
In MOVES4, the heavy-duty gasoline vehicle emission rates were updated based on the new fuel-
based emission rates for light-duty trucks. We assume that the fuel-based ammonia emission rates
are similar between light-duty and heavy-duty gasoline vehicles because the same ammonia
formation pathway is present for both vehicles types. In modern gasoline vehicles, ammonia is
formed from the catalytic reduction of NO in the three-way catalytic converter during fuel-rich
conditions.
Limited data is available to evaluate this assumption. Livingston et al. (2009) 147 measured
ammonia emissions from vehicles recruited in southern California. On average, they measured
higher gram per mile ammonia rates from the six medium-duty gasoline vehicles than the 35 light-
duty vehicles. They attributed the higher ammonia emission rates from the medium-duty vehicles
due to:
1)	Larger exhaust volumes produced by the medium-duty vehicles
2)	Less stringent emission standards, which would lead to higher precursor exhaust emissions
of NOx and CO
3)	Potentially different catalyst activity
By estimating heavy-duty gasoline emission rates from light-duty gasoline vehicle fuel-based
emission rates, and heavy-duty gasoline fuel rates, we account for the larger exhaust volumes
produced by heavy-duty gasoline vehicles. While, we do not account for items 2) and 3), which
could lead to higher fuel-based emission rates from heavy-duty gasoline vehicles, we believe our
approach is reasonable given that heavy-duty gasoline vehicles are anticipated to be a minor
contributor of ammonia emissions in comparison to light-duty gasoline vehicles.
The light-duty truck fuel-based ammonia emission rates are documented in the light-duty exhaust
emission rate report.9 The fuel-based emission rates are estimated for light-duty vehicles and light-
duty trucks by model year and vehicle age.
We estimated the ammonia emission rate for heavy-duty gasoline vehicles by multiplying the light-
duty truck fuel-specific emission rates by regulatory class, model year group, and age
(¦LDT FERMYiage) by the MOVES3 heavy-duty gasoline vehicle fuel-consumption rates by
regulatory class, model year, operating mode (HDG Fuel RatesReg MYi0p) as shown in Equation
5-6.
231

-------
HDG ERReg MY,age,op = HDG Fuel RdtesReg MY,op x LDT FERMY age Equation 5-6
Example NH3 emission rates in MOVES for LHD2b3 and LHD45 gasoline vehicles for model year
2017 and ages 0-3 are shown in Figure 5-2
12.5
10.0-
=j
o
O)
d) 7.5
03
C
o
(/)
(/)
'E
a)
CO
X
5.0-
2.5
0.0-
LHD2b3

LHD45
—1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1— —1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1—
0 1 111213141516212223242527282930333537383940 0 1 111213141516212223242527282930 3335 3738 3940
Operating Mode
Figure 5-2. NHj emission rates (g/hour) by operating mode for regulatory class LHD2b3 and LHD45 and Model
Years 2017 for ages 0-3.
5.3 Heavy-Duty Compressed Natural Gas
CNG vehicles with stoichiometric three-way catalysts have been shown to emit higher ammonia
emissions than CNG vehicles with lean-burn combustion.135 136 Since ammonia measurements were
not part of the HDIUT program used to update CNG emission rates described in Section 4.1.2, the
ammonia emission rates for CNG vehicles were set equal to the ammonia emission rates from
heavy-duty gasoline vehicles by regulatory class, model year, operating mode, and vehicle age. The
extended idle emission rates for CNG vehicles are set equal to the running idle rates.
Two studies have demonstrated that stoichiometric emission rates from CNG vehicles can be
significantly higher than those from heavy-duty gasoline vehicles.148,149 However, one recent study
232

-------
by CE-CERT has shown more moderate NH3 emissions from CNG vehicles.150 These data have yet
to be incorporated into MOVES.
5.4 Summary
Figure 5-1 displays the age 0-3 ammonia emission rate for heavy-duty vehicles by regulatory class
and fuel type.
1970	1980	1990	2000	2010
Model Year
2020
2030
2040
Figure 5-3: Base running emission rates for NH3 from age 0-3 gasoline heavy-duty vehicles averaged over a
nationally representative operating mode distribution
Reg Class
41-LHD2b3
42-LHD45
46-MHD67
47-HHD8
233

-------
0.100
0.075
4)
|
s
£ 0.050
Q£
0
1
0.025'
0.000'






, t ^





N.





1980
2000
2020
2040
Model Year
Reg Class
41-LHD2b3
¦	42-LHD45
46-MHD67
¦	47-HHD8
48-Urban	Bus
49-Gliders
Figure 5-4: Base running emission rates for NH3 from age 0-3 diesel heavy-duty vehicles averaged over a
nationally representative operating mode distribution
Reg Class
¦ 47-HHD8
48-Urban Bus
1980
2000
2020
2040
Model Year
Figure 5-5: Base running emission rates for NHj from age 0-3 CNG heavy-duty vehicles averaged over a
nationally representative operating mode distribution
234

-------
6 Heavy-Duty Crankcase Exhaust Emissions
Crankcase exhaust emissions, also referred to as crankcase blowby, are combustion gases that pass
the piston rings into the crankcase and are subsequently vented to the atmosphere. Crankcase
blowby includes oil-enriched air from the turbocharger shaft, air compressors, and valve stems that
enters the crankcase. The crankcase blowby contains combustion generated pollutants, as well as
oil droplets from the engine components and engine crankcase.151
6.1 Modeling Crankcase Emissions in MOVES
MOVES calculates crankcase emissions using two code modules: a gaseous and a particulate
matter crankcase emission calculator. Within these calculators, crankcase emissions are calculated
in relationship to tailpipe exhaust emissions. In MOVES, the tailpipe exhaust processes are running
exhaust, start exhaust, and extended idle exhaust (processID 1, 2, 90). The corresponding crankcase
emission processes are crankcase running exhaust, crankcase start exhaust, and crankcase extended
idle exhaust (processID 15, 16, 17).
The gaseous crankcase calculator chains calculation of the crankcase emission rates to the tailpipe
exhaust emission rates for gaseous pollutants, but it does not change the tailpipe exhaust emission
rates. On the other hand, the particulate matter calculator has the ability to divide the particulate
matter exhaust emission rates stored in the emissionratebyage and emissionrate table into
components representing the contributions from tailpipe exhaust and crankcase emissions. Thus,
the particulate ratios may be used to adjust the particulate matter tailpipe exhaust emission rates to
account for the crankcase contribution, as was done in previous versions of MOVES. In MOVES4,
however, the particulate matter tailpipe exhaust emission rates only include tailpipe emissions and
the exhaust ratios are set equal to one as shown in the subsequent sections for all model years, fuel
types, regulatory classes and source types. More details on the particulate matter crankcase
calculator are provided in the MOVES Speciation Report.1
The crankcase ratios for non-methane hydrocarbons (NMHC), NOx, and PM2.5 are used to estimate
the crankcase emissions for each subspecies that is chained to their respective primary pollutant as
shown in Table 6-1. The crankcase emission ratios for both gaseous and particulate matter
pollutants are stored in the crankcaseEmissionRatio table. The table stores the crankcase emission
rates by pollutant, process, model year, source type, regulatory class, and fuel type. Regulatory
class was added as a primary field in MOVES4, and the crankcase emission ratios were updated
according to regulatory class. The table structure and code retain the ability to model separate
crankcase emission ratios by source type, but we use the same crankcase emission ratios across
different source types within the same fuel type, regulatory class and model year.
The PM10 crankcase emission rates are subsequently estimated from the PM2.5 exhaust and
crankcase emission rates using PM10/PM2.5 emission ratios as documented in the MOVES
Speciation Report.1
235

-------
Table 6-1 Speciation of Chained Crankcase Pollutants from the Primary Pollutants
Primary Pollutant
Chained Crankcase Pollutant
Non-methane
hydrocarbons
(NMHC)
Benzene, Dibenzo(a,h)anthracene gas, 1,3-Butadiene, Fluoranthene gas,
Formaldehyde, Acenaphthene gas, Acetaldehyde, Acenaphthylene gas, Acrolein,
Anthracene gas, 2,2,4-Trimethylpentane, Benz(a)anthracene gas, Ethyl Benzene,
Benzo(a)pyrene gas, Hexane, Benzo(b)fluoranthene gas, Propionaldehyde,
Benzo(g,h,i)perylene gas, Styrene, Benzo(k)fluoranthene gas, Toluene, Chrysene
gas, Xylene, Fluorene gas, non-methane organic gas (NMOG),
Indeno(l,2,3,c,d)pyrene gas, volatile organic carbon (VOC), Phenanthrene gas,
Naphthalene gas, Pyrene gas
NOx
NO, N02, HONO
PM2.5
EC, S04, FhO(aerosol), NonECNonS04PM
By using crankcase to tailpipe emission ratios to estimate crankcase emission rates, MOVES
implicitly assumes that any increase in emissions due to aging and deterioration also occurs for
crankcase emissions. The data sets used to derive the crankcase emission rates for 2007 and later
heavy-duty diesel engines are based on engines and vehicles with relatively low miles and no
expected deterioration in the emission control system. If crankcase emissions do not exhibit the
same increase in deterioration as tailpipe emissions, this method would lead to an overestimation of
crankcase emissions in older vehicles.
6.2 Heavy-Duty Diesel Crankcase Emissions
Crankcase emissions from pre-2007 diesel engines were typically vented to the atmosphere using
an open unfiltered crankcase system, referred to as a 'road draft tube'.151 Researchers have found
that crankcase emissions vented to the atmosphere can be the dominant source of diesel particulate
matter concentrations measured within the vehicle cabin.152'153'154
Starting in model year 2001, federal regulations require closed crankcase systems for chassis-
certified diesel vehicles.155 Federal regulations permit 2006-and-earlier engine-certified diesel
vehicles equipped with "turbochargers, pumps, blowers, or superchargers" to vent crankcase
emissions to the atmosphere.156 Starting in model year 2007, federal regulations no longer permit
crankcase emissions to be vented directly to the atmosphere, unless they are included in the
certification exhaust measurements.156 Many heavy-duty diesel manufacturers have adopted open
crankcase filtration systems in model year 2007 and later engines.151 These systems vent the
exhaust gases to the atmosphere after the gases have passed a coalescing filter which removes oil
and a substantial fraction of the particles in the crankcase blowby.151 In the ACES Phase 1
program, four MY 2007 diesel engines from major diesel engine manufactures (Caterpillar,
Cummins, Detroit Diesel, and Volvo) all employed filtered crankcase ventilation systems.157
A summary of published estimates of diesel crankcase emissions as percentages of the total
emissions (tailpipe + crankcase) are provided in
Table 6-2. For the pre-2007 diesel technologies, hydrocarbon and particulate matter emissions have
the largest contributions from crankcase emissions. There is a substantial decrease in PM2.5
emissions beginning with the 2007 model year diesel engines. The aftertreatment technologies
required for 2007-and-later engines reduces the tailpipe emissions more than the crankcase
236

-------
emissions, resulting in an increase in the relative crankcase contribution for THC, CO, and PM2.5
emissions.
Table 6-2 Literature Review on the Contribution of Crankcase Emissions to Diesel Exhaust (Tailpipe +
			Crankcase) 				


Tailpipe
Exhaust
#
Engines





Model
Aftertreat
or




Study
Year
ment
Vehicles
THC
CO
NOx
PM

1966,


0.2%-
0.01%-
0.01%-
0.9%-
Hare and Baines, 1977160
1973
None
2
3.9%
0.4%
0.1%
2.8%
Zielinska et al. 2008152
2000,





13.5% -
Ireson etal. 2011153
2003
None
2



41.4%
Clark et al. 2006159







Clark et al. 2006158
2006
None
1
3.6%
1.3%
0.1%
5.9%
Khalek et al. 20 0947

DPF-





("ACES Phase 1")
2007
equipped
4
84.6%
33.5%
0.007%
44.4%

2015,


19.0%-
14.2%-
2.3%-
Not
NVFEL Testing
2018
SCR-DPF
2
57.8%
76.7%
7.5%
measured
Note:
The crankcase ratios shown here are a fraction of the total tailpipe and crankcase exhaust. The crankcase ratios used in
MOVES are a ratio of crankcase to tailpipe exhaust.
As discussed in the following subsections, we developed crankcase emission ratios by regulatory
class and model years groups using the available studies in
Table 6-2, and additional information on requirements for closed crankcase systems.
The data on crankcase emissions are limited. The gaseous crankcase emission rates for heavy-duty
diesel vehicles are based on three studies in
Table 6-2, totaling only seven vehicles. As such, the rates have considerable, but unquantified,
uncertainties.
6.2.1 LHD2b3 Crankcase Emissions
After 2001, all chassis-certified vehicles, including diesel vehicles, are required to avoid venting
crankcase emissions into the atmosphere.155 All LHD2b3 vehicles in MOVES4 are chassis-certified
vehicles because the small number of engine-certified LHD2b3 vehicles are re-classified as LHD45
vehicles as discussed Section 1.4.
MOVES uses two model year groups for crankcase emissions from LHD2b3 diesel vehicles. Model
year 1960 to 2000 vehicles use the open crankcase ratios estimated for engine-certified vehicles as
detailed in Section 6.2.2.1. For model year 2001 to 1960 we estimate zero crankcase ratios because
all chassis-certified diesel vehicles are required to have closed crankcase systems.
237

-------
6.2.2 LHD45, MHD andHHD Crankcase Emissions
Diesel vehicles within the LHD45, MHD, and HHD regulatory classes are composed of 100%
engine-certified vehicles, which are permitted to emit crankcase emissions after model year 2007 if
they are accounted for in the engine-certification results. The crankcase emission rates for the
engine-certified vehicles are estimated by regulatory class and model year groups, to capture
differences in crankcase emission ratios reported from different studies, account for differences in
the crankcase control in 2007 and later model years, and to account for changes in tailpipe exhaust
emissions which impact the crankcase to tailpipe emission ratios.
6.2.2.1 1960-2006Model Years
Table 6-3 displays the crankcase/tailpipe emission ratios used for pre-2007 diesel exhaust. For
THC, CO, and NOx, we selected the values measured on the MY 2006 diesel engine reported by
Clark et al. 2006.159 These values compare well with the previous HC, CO, NOx values reported
much earlier by Hare and Baines (1977),160 which represent much older diesel technology. The
similarity of the crankcase emission ratios across several decades of diesel engines suggests that for
pre-2007 diesel engines, crankcase emissions can be reasonably well represented as a fraction of
the exhaust emissions. The THC crankcase ratios presented in Table 6-3 are also used for methane,
total organic gases (TOG), non-methane hydrocarbons (NMHC) and all the pollutants chained to
NMHC listed in Table 6-1.
For PM2.5 emissions, we use a crankcase/tailpipe ratio of 20 percent. The 20 percent ratio falls
within the range of observations from the literature on diesel PM emissions. Zielinska et al. 2008152
and Ireson et al. 2011153 reported crankcase contributions to total PM2.5 emissions as high as 40
percent. Jaaskelainen (2012)151 reported that crankcase can contribute as much as 20 percent of the
total emissions from a review of six diesel crankcase studies. Similarly, an industry report
estimated that crankcase emissions contributed 20 percent of total particulate emissions from 1994-
2006 diesel engines.161 The crankcase emission ratios shown Table 6-3 are applied to running, start
and extended idle exhaust to estimate the corresponding crankcase exhaust emissions.
Table 6-3 LHD45, MHD, and HHD 1960-2006 Diesel Crankcase Ratios for HC, CO, NOx, and PM2.s
Pollutant
Crankcase/Tailpipe
Ratio (MOVES inputs)
Crankcase/(Crankcase +
Tailpipe) Ratio
THC
0.037
0.036
CO
0.013
0.013
NOx
0.001
0.001
PM25
0.200
0.167
Note:
MOVES uses a crankcase/tailpipe ratios. We also calculated the crankcase to total exhaust ratio (crankcase + tailpipe)
to compare the MOVES inputs to the values reported in the literature
As outlined in the MOVES Speciation Report, MOVES does not apply the crankcase/tailpipe
emission ratio in Table 6-3 to the total exhaust PM2.5 emissions. MOVES applies
crankcase/tailpipe emission ratios to PM2.5 subspecies: elemental carbon PM2.5, sulfate PM2.5,
aerosol water PM2.5, and the remaining PM (nonECnonS04PM). This allows MOVES to account
238

-------
for important differences in the PM speciation between tailpipe and crankcase emissions. Tailpipe
exhaust from pre-2007 diesel engines is dominated by elemental carbon emissions from
combustion of the diesel fuel, while crankcase emissions are dominated by organic carbon
emissions largely contributed by the lubricating oil.152'153 Zielinska et al. 2008152 reported that the
EC/PM fraction of crankcase emissions from two pre-2007 diesel buses is 1.57 percent.
To account for the different speciation of exhaust and crankcase emissions, the crankcase emission
factors for PM species shown in Table 6-4 have been back-calculated such that the total crankcase
PM2.5 emissions are 20 percent of the PM2.5 exhaust measurements (consistent with Table 6-3) and
have an EC/PM split of 1.57 percent. The start and extended idle crankcase ratios are the same,
because the pre-2007 start and extended idle exhaust EC/PM are the same (both 46.4% as
documented in Section 2.2.2 and Section 2.3.1). The running exhaust EC/PM ratio is 79% (Section
2.1.2.1.8).
The tailpipe exhaust fractions are set equal to 1 because the tailpipe emission rates are not assumed
to include any crankcase emissions. In other words, the crankcase emissions are estimated in
addition to the tailpipe emissions.
Table 6-4. LHD45, MHD, and HHD Exhaust and Crankcase Ratios for 1960-2006 Diesel by Pollutant, Process,
	and Model Year Group for PM2.5 Species 	
Pollutant
Process
Start
Running
Extended Idle
EC

1
1
1
nonECnonSC^PM
Tailpipe
1
1
1
S04
Exhaust
1
1
1
h2o

1
1
1
EC

0.007
0.004
0.007
nonECnon SO4PM
Crankcase
0.367
0.937
0.367
S04
0.367
0.937
0.367
H20

0.367
0.937
0.367
6.2.2.2 2007-2009 Model Years
As discussed in the background section above, the 2007 heavy-duty diesel emission regulations
impacted the technologies used to control exhaust and crankcase emissions. The regulations also
expanded the types of emissions data included in certification tests by including crankcase
emissions in the regulatory standards which previously included only tailpipe emissions. Because
heavy-duty diesel engine manufacturers are using open-filtration crankcase systems, the crankcase
emissions are included in the emission certification results. In MOVES, the base exhaust rates for
2007 to 2009 diesel engines are based on certification test results for PM2.5 as discussed in this
section.
The crankcase ratios for 2007-2009 HDD emissions are based on the ACES Phase 1 study157,
which tested four MY 2007 engines from different manufactures (Caterpillar, Cummins, Detroit
Diesel and Volvo). The ACES Phase 1 engines and exhaust control systems were new, and
underwent 125 hours of "degreening" before the test program. Thus, they represent low-mileage,
239

-------
properly functioning heavy-duty diesel engine emissions. In reporting the emission rates, the
engines were anonymized as A, B, C, D and a backup engine B' which was tested at a secondary
site. The B' backup engine is the same make, model as engine B. The ACES study conducted hot
FTP cycles that sampled tailpipe exhaust emissions and repeat tests that sampled combined tailpipe
and crankcase exhaust. The crankcase emission rates for each engine were estimated by calculating
the difference between the average emissions measured with and without crankcase emissions
routed into the sampling system as shown in Equation 6-1, where average emissions are the total
mass divided by the testing time.
Remissions,i = ExhaustwithCQi - Exhaustw/oCQi	Equation 6-1
Where:
CCemissions,i= crankcase emissions (grams per hour) for engine i
Exhaustwithcc j=hot-FTP cycle average emission emissions (grams per hour) for each engine i with the
crankcase routed into the sampling system
Exhaustw/oCC j= hot-FTP cycle average emission emissions (grams per hour) for each engine i without the
crankcase
Then the crankcase emission rates were averaged together in Equation 6-2 with the backup engine
being treated as additional tests of engine B.
fr^r*	, CCengineB + CCengineB'.sitel + CCengineB',site2 ,
_ (LLengineA "i	3	LLengineC + LLengineDj Equation
ACESavg -	^	6_2
Since ACES Phase 1 hot-FTP contained a hot-start we assume starts are accounted for in the
running values, and thus for these model years, the crankcase ratio for starts is zero. Note that for
similar 2010+ vehicles, where cold start crankcase emissions were measured, positive crankcase
starts were measured for only for CO as discussed in Section 6.2.2.3. And, for extended idle, based
upon data from 2010+ HDD vehicles using the same crankcase technology that show the crankcase
emission rates for extended idling are similar to the running rates, we use the hot-FTP rates from
ACES Phase 1 for the crankcase extending idling rates. The crankcase rates for running, starts and
extended idle are listed in Table 6-5.
Equation 6-3 is used to calculate TOG crankcase emissions from the methane and NMHC
crankcase values, using the MOVES NMOG/NMHC value for 2007-2009 MY diesel exhaust.1
g	g	g NMOG
TOG(-f) = CH4(-f ) + NMHC(-—) * ——	Equation 6-3
hr	hr	hr NMHC
Where each pollutant rate is in g/hr and the ratio of NMOG/NMHC is 1.343 from the hcspeciation
table in MOVES.1
240

-------
Table 6-5 The MY 2007 crankcase exhaust rates (CCACESavg) from ACES Phase 1 FTP cycle used for running and
			extended idling (g/hr)	
CO
NOx
THC
ch4
NMHC
TOG
Total PM2 5
8.15
0.79
1.76
0.38
1.37
2.23
0.04
The crankcase ratios for the ACES program in
Table 6-2 are calculated using Equation 6-4 where the ExhaustACES is calculated using Equation
6-2, but using Exhaustw/oCC j in place of the crankcase values.
CCAGESavg	Equation 6-4
^ACES ratio — rr	,
CCACESavg + ExhaustACES
We then calculated the crankcase ratios for use in MOVES using the MOVES MY 2007-2009
exhaust base rates for all diesel HHD vehicles (regClass 47) weighted by the activity of short and
long-haul single-unit and combination trucks (sourcetypes 52,53,61,62) in each operating mode
estimated from a preliminary MOVES3 national scale run. The values for ExhaustMOvES are listed
in Table 6-6. The extended idle rates are based on a single operating mode in MOVES. We used
the HHD diesel emission rates to match the HHD engines tested in the ACES Phase 1 program.
Table 6-6: The MY 2007-2009 MOVES exhaust base rates (g/hr) for running (weighted by operating mode
		 activity) and extended idling 			

CO
NOx
THC
ch4
NMHC
TOG
Total PM2 5
Running Exhaust
11.46
288.51
2.93
1.73
1.20
3.34
0.88
Extended Idle Exhaust
39.26
100.45
8.49
5.00
3.49
9.69
0.087
We then used Equation 6-5 to estimate the base crankcase to tailpipe exhaust ratio for 2007-2009
HD vehicles.
CCAGESavg	Equation 6-5
'"'"•base " ExhaUStM0VES
We assume that crankcase emissions are proportional to the tailpipe exhaust emissions across
regulatory classes and source types. As such, we use the CCratio base to derive the crankcase ratio
for all heavy-duty diesel regulatory classes and source types. The heavy-duty base crankcase ratios
are shown in Table 6-8.
As mentioned in the background section, many manufactures employ an open crankcase ventilation
system with a coalescing filter, but a substantial fraction opt for a closed crankcase system where
the crankcase vapors are either routed into the engine with the fuel injection or into the exhaust
stream upstream of the aftertreatment. For developing crankcase emission ratio estimates, we
model heavy-duty diesel closed crankcase systems as having zero crankcase emissions, and reduce
the base open crankcase emission ratios to account for the fraction of open crankcase systems in the
vehicle fleet.
241

-------
Available certification data on the prevalence of open and closed crankcase systems are
incomplete, but the available data suggests that most manufacturers use either open or closed
systems for all their engines in a given model year. We assumed that any engine family listed in the
certification data with missing crankcase information was of the same type as all other engine
families in that model year for that manufacturer. This was our best assumption based on the
available data, despite finding one conflict.ss We then used manufacturer production volume data
from 2016-2018 (complete data for earlier years was not available) to weight the number of open
and closed systems within each regulatory class as shown in Table 6-7. In these model years, there
were between 9 and 10 engine manufacturers which produced heavy-duty engines, with between 34
and 41 certified engine families. Using this method, we estimated that 0% of LHD engines, 90.5%
of medium heavy-duty and 67% of HHD engines have open crankcase systems. The average value
across the 2016-2017 model years and by regulatory classes was used for all 2007 and later
engines.tt
Table 6-7 Fraction of Engines with Open Crankcase Systems by Vehicle Regulatory Class and Model Year
Model Year
LHD
MHD
HHD
2016
0
0.912
0.725
2017
0
0.919
0.635
2018
0
0.884
0.640
average
0
0.905
0.666
Finally, to estimate the MY 2007-2009 crankcase emission ratios for MOVES, we assume that
crankcase emissions are proportional to the exhaust emissions across each of the engine-certified
regulatory classes and source types. To estimate the crankcase emission ratio for each regulatory
class, we multiplied the open crankcase ratio, CCratio base, by the open crankcase fraction of each
heavy-duty regulatory class, OpenCCfrac,regciass as shown in Equation 6-6.
CCratio,regClass,modelyear — CCratj0 base * OpenCCfrac	Equation 6-6
Where
CCratio,regciass,modei year= the crankcase ratio used in MOVES by regulatory class and model year
CCratio,base= the heavy-duty open crankcase ratio calculated from Equation 6-5
OpenCCfrac regClass= the fraction of open crankcase systems by regulatory class determined from certification
data and manufacture production volume as shown in Table 6-7
The MOVES 2007-2009 crankcase ratios for THC, CH4, NMHC, TOG, NOx, CO, and PM2.5 for
each crankcase process for LHD45, MHD, HHD and Urban Bus regulatory classes are shown in
Table 6-8.
ss One of the engine manufacturers that we assumed produced closed crankcase systems for all of its 2007 and later
model year engines based on certification data produced one of the open crankcase MY 2007 engine tested in the
ACES Phase 1 program.
tt We recognize the uncertainties in applying open crankcase values from model year 2016-2018 engines to model year
2007-2009 engines. However, the larger uncertainty in the crankcase emission rates is driven by the measured
crankcase emission rates. Assuming an HHD open crankcase percentage of 67% decreased the crankcase tailpipe
emission rates by roughly one third, which is well within the range of variability of emissions observed in the ACES
Phase 1 crankcase emission results.
242

-------
Table 6-8 Crankcase/Tailpipe Ratios for Model Year 2007-2009 Engine-certified Vehicles by Heavy-duty Diesel
Process
Pollutant
HD
baseline
(Equation
6-5)
Crankcase/Tailpipe ratio by regulatory class
(Equation 6-6)
LHD45
MHD
HHD
Urban Bus
Running
THC
0.600
0
0.543
0.400
0.400
ch4
0.223
0
0.201
0.148
0.148
NMHC
1.141
0
1.032
0.760
0.760
TOG
0.667
0
0.603
0.444
0.444
C02
0.005
0
0.004
0.003
0.003
CO
0.711
0
0.644
0.474
0.474
NOx
0.003
0
0.002
0.002
0.002
PM
0.043
0
0.039
0.029
0.029
Starts
THC
0
0
0
0
0
ch4
0
0
0
0
0
NMHC
0
0
0
0
0
TOG
0
0
0
0
0
C02
0
0
0
0
0
CO
0
0
0
0
0
NOx
0
0
0
0
0
PM
0
0
0
0
0
Extended
Idle
THC
0.207
0
0.187
0.138
0.138
ch4
0.077
0
0.070
0.051
0.051
NMHC
0.394
0
0.356
0.262
0.262
TOG
0.230
0
0.208
0.153
0.153
C02
0.054
0
0.049
0.036
0.036
CO
0.208
0
0.188
0.138
0.138
NOx
0.008
0
0.007
0.005
0.005
PM
0.436
0
0.394
0.290
0.290
A C02 is not included in the MOVES crankcaseEmissionRatio table, in part due to its small fraction compared to running
exhaust CO2. However, it is included here for comparison with other pollutants.
For PM2.5 emissions, MOVES applies crankcase ratios to each of the intermediate PM2.5 species
(EC, nonECnonS04PM, S04, and H20). The MOVES PM2.5 speciation profile developed from the
ACES Phase 1 study combined the crankcase and tailpipe emissions. As such, we model crankcase
emissions as having the same speciation as tailpipe emissions, and the crankcase fractions for the
intermediate PM2.5 species in Table 6-9 are the same as derived for total PM2.5 in Table 6-8. In
MOVES4, we set the tailpipe ratios for model year 2007-2009 equal to one because we now
assume that the tailpipe emission factors only include tailpipe exhaust. This approach was taken
because the 2007-2009 model year running PM2.5 emission rates are based on assumed reductions
from MY 1998-2006 vehicles based on certification data (See Section 2.1.2.1.7). Assuming the
resulting running PM emission rates in 2007-2008 only include tailpipe exhaust was deemed
equally valid as our previous assumption, and simplifies the calculation of crankcase emissions by
keeping all tailpipe ratios in the crankcase calculator equal to one. The extended idle emissions for
model year 2007-2009 are based on tailpipe measurements only, and using a extended idle ratio
243

-------
equal to one is consistent with the measured data. The peer-reviewers comments from the
MOVES3 crankcase review agreed with this update.162
The crankcase emission ratios for MHD regulatory class for the intermediate PM species are the
same as total PM2.5 (0.039 for running and 0.394 for extended idle) as shown in Table 6-8.
Table 6-9. MOVES Exhaust and Crankcase Ratios for Model Year 2007-2009 HHD Diesel by Pollutant and
Process for PM2.5 Species
Process
Model Year Group
Pollutant
Start
Running
Extended Idle


EC
1
1
1
Tailpipe
2007-2009
nonECnonS04PM
1
1
1
Exhaust
S04
1
1
1


H20
1
1
1


EC
0
0.0290
0.2929
Crankcase
2007-2009
nonECnon SO4PM
0
0.0290
0.2929
S04
0
0.0290
0.2929


h2o
0
0.0290
0.2929
6.2.2.3 2010-2026Model Years
The HDIUT program (see Section 2.1.1.1) is used as the source of baseline exhaust emission rates
for model year 2010-2026 diesel vehicles. As the HDIUT program measures tailpipe exhaust
emissions from trucks in-use, the crankcase is not routed to the tailpipe and therefore not accounted
for in the data. To account for crankcase emissions for 2010 and later diesel vehicles, we used
direct crankcase measurements of NOx, THC, CH4, and CO emissions from the US EPA's
National Vehicle and Fuel Emissions Laboratory (NVFEL) testing of two heavy duty trucks (MY
2015 and MY 2019). Each of the trucks had less than 10,000 miles and represent properly
functioning low-mileage heavy-duty vehicles. Testing was conducted on a chassis dynamometer
over a drive cycle that consists of a hot or cold start followed by an ARB transient cycle (Phase 1),
followed by four repetitions of the same ARB transient (Phase 2), 10 minutes of idling (Phase 3),
and steady-state highway activity at 55 mph and 60 mph (Phase 4). The speed trace and a graphical
indication of the testing phases are shown in Figure 6-1.
244

-------
Figure 6-1 Speed trace of the NVFEL testing program along with each phase
There were a total of thirteen repetitions between the two trucks, with soak times between
repetitions of either between one and three hours (hot start) or greater than twelve hours (cold start)
as shown in Table 6-10.
Table 6-10: Testing information for the NVFEL test program
Truck
ID
Number of Tests
Dates of Testing
Soak Times
1
6 (3 Hot Start, 3 Cold Start)
May 30, 2017
June 1, 2017
Hot Starts 1 -3 hrs
Cold Starts 12+ hours
2
7 (3 Hot Start, 4 Cold Start)
August 20, 2019
- August 13,
2019
Hot Starts 1 -3 hrs
Cold Starts 12+ hours
Table 6-11 reports the average emission rates across the repetitions by truck and phase, and by
phase only (weighting the two trucks equally).
245

-------
Table 6-11 NVFEL Vehicle average emission rates by testing phase and truck

Crankcase CO
Crankcase NOx
Crankcase THC
Crankcase CH4

(g/hr)
(g/hr)
(g/hr)
(g/hr)
Phase 1 Average
1.0900
0.4427
0.3463
0.0324
Truck 1 average
0.5057
0.5527
0.2373
0.0053
Truck 2 average
1.6742
0.3327
0.4552
0.0595
Phase 2 Average
0.6224
0.9086
0.4537
0.0155
Truck 1 average
0.1916
1.0316
0.2835
0.0001
Truck 2 average
1.0532
0.7855
0.6238
0.0309
Phase 3 Average
0.2461
0.7097
0.4529
0.0058
Truck 1 average
0.1670
0.7633
0.2634
0.0000
Truck 2 average
0.3252
0.6560
0.6424
0.0115
Phase 4 Average
0.2679
1.8447
0.5915
0.0053
Truck 1 average
0.1521
1.7766
0.3711
0.0000
Truck 2 average
0.3836
1.9127
0.8118
0.0106
Average of all phase averages
0.5677
0.9764
0.4611
0.0148
The average running rate was calculated by calculating a weighted average between Phase 2
(36.7%) and Phase 4 (63.3%). The phase-weighting was calculated to match the national running
operating mode distribution for MY2015, HHD diesel vehicles in short- and long-haul single-unit
and combination trucks (source type 52,53,61,62) as estimated in a MOVES national run using a
draft version of MOVES3 for calendar year 2015.72 Phase 2 had transient operation below 50 mph
and was mapped to operating modes 0-30 {36.1% of total activity) and Phase 4 had constant high
speed data and was mapped to operating-modes 33-40 (63.3% of total activity) as shown in Table
6-12. HHD diesel vehicles in short- and long-haul single-unit and combination trucks vehicles were
chosen as to match the NVFEL testing vehicles.
246

-------
Table 6-12. Assignment of MOVES Operating Modes to Test Phase from the NVFEL data, and Phase weighting
determined from the Operating Mode activity from a National draft MOVES run for MY 2015 HHD Vehicles in
	 the Short- and Long-haul Single-unit and Combination Trucks Source Types 	
OpModelD
Operating Mode
Description
Scaled Tractive
Power (STPt
skW)
Vehicle
Speed
(vu mph)
% of Total
Activity from
MOVES run
NVEL
Test Phase
0
Deceleration/Braking


2.69

1
Idle

vt< 1.0
5.40

11
Coast
STP;< 0
I 
-------
upon dilution. For NOx emissions, the engine-out emissions during the start period can be lower
than running period, due to lower air-fuel ratios and lower in-cylinder temperatures. In fact,
MOVES models zero NOx tailpipe starts for pre-2010 heavy-duty diesel trucks based on
observations of negative start emission rates as discussed in Section 2.2.1.2. Thus, we set the start
crankcase emission rate to zero for THC and NOx. The average CH4 start emission rate was
positive, but because methane is estimated in MOVES as a fraction of THC, we also set it equal to
zero as shown in Table 6-13.
For the crankcase extended idling rate we used the average of emissions in the idling phase (Phase
3).
These rates also informed the base crankcase rates for NMHC and TOG. We used Equation 6-3 to
calculate TOG crankcase emissions, with the MOVES NMOG/NMHC ratio for 2010+ diesel
running exhaust (1.085) from the hcspeciation table in MOVES.1
The crankcase rates are ratioed to the tailpipe emissions from the HDIUT vehicles from the model
year 2010-2013 and 2014+ NOx FEL 0.2 groups (See Table 2-7) for each pollutant. These vehicles
were chosen as they comprise the majority of the fleet in most model years (See Figure 2-12). The
HDIUT data are used instead of the tailpipe values from the NVFEL test trucks because 1) they are
the basis of the MOVES tailpipe emission rates, 2) we have much more confidence in the mean
HDIUT tailpipe emission rates than the mean tailpipe emission rates from the two NVEL test-
trucks to represent fleet-average rates, and 3) the variability of the tailpipe measurements from the
NVFEL vehicles is generally much greater than the variability of the crankcase measurements,
including for THC emissions.163 Thus, by using tailpipe emissions from MOVES, we yield more
stable crankcase ratios.
The ratios for NMHC, TOG, and NOx are applied to the pollutants chained to them as well (Table
6-1).
Table 6-13 Average Crankcase Emission Rates for MOVES Processes from NVFEL data
Process
CO
NOx
THC
ch4
NMHC
TOG
Crankcase Running
Exhaust (g/hr)
0.398
1.501
0.541
0.009
0.5318
0.5861
Crankcase Start
Exhaust (g/start)
0.049
0
0
0
0
0
Crankcase Extended
Idling Exhaust (g/hr)
0.246
0.710
0.453
0.006
0.4471
0.4909
No PM2.5 crankcase emission measurements were made in the NVFEL study, so the 2010 and later
model year calculations use the crankcase PM2.5 rates from ACES Phase 1 report as described for
model years 2007-2009 in Section 0. More information on the PM calculations for model year 2010
and later vehicles is provided later in this section.
We used the base exhaust emission rates from the NOx FEL 0.2 group for HHD engines in short
and long-haul single-unit and combination trucks (source types 52,53,61,62). These vehicles were
chosen from the full HDIUT data set to match the NVFEL testing vehicles.
248

-------
Table 6-14 The NOx FEL 0.2 MOVES Exhaust Base Rates for Running (Weighted by Operating Mode Activity)
		and Extended Idling (g/hr)				

Model
Year Group
CO
NOx
THC
CH4
NMHC
TOG
Total PM2 5
Running Exhaust
2010-2013
125.42
62.51
1.50
0.57
0.93
1.58
0.15
2014+
60.06
61.87
1.47
0.56
0.91
1.55
0.12
Extended Idle Exhaust
2010-2013
39.26
42.60
2.75
1.04
1.70
2.89
0.03
2014+
39.26
42.60
1.64
0.62
1.01
1.72
0.02
The HHD tailpipe rates have two model year groups 2010-2013 and 2014+, so the ratios differ for
these model year groups as shown in Table 6-15.
As was done for the development of the 2007-2009 crankcase emission ratios documented in
Section 6.2.2.2 we assume that the crankcase ratios derived for MY2010-2026 heavy heavy-duty
diesel (HHD) engines apply to the other heavy-duty regulatory classes (LHD45, MHD and Urban
Bus) and source types (other buses, transit buses, school buses, refuse trucks, and motorhomes),
and that crankcase emissions are proportional to the exhaust emissions across regulatory classes
and source use types. Additionally, we use Equation 6-6 to account for the fraction of open
crankcase systems within each regulatory class shown in Table 6-7.
249

-------
Table 6-15 Crankcase/Tailpipe Ratios for Model Year 2010-2026 Heavy-Duty Diesel Regulatory Classes
Process
Pollutant
Model Year Group
HD
baseline
Crankcase/Tailpipe ratio by regulatory class


LHD45
MHD
HHD
Urban Bus

THC
2010-2013
0.36
0
0.33
0.24
0.24

2014-2026
0.37
0
0.33
0.24
0.24

ch4
2010-2013
0.02
0
0.014
0.011
0.011

2014-2026
0.02
0
0.015
0.011
0.011

NMHC
2010-2013
0.57
0
0.52
0.38
0.38

2014-2026
0.58
0
0.53
0.39
0.39

TOG
2010-2013
0.37
0
0.34
0.25
0.25
Running
2014-2026
0.38
0
0.34
0.25
0.25
C02
2010-2013
0
0
0.002
0.002
0.002

2014-2026
0
0
0.002
0.002
0.002

CO
2010-2013
0
0
0.003
0.002
0.002

2014-2026
0.01
0
0.006
0.004
0.004

NOx
2010-2013
0.02
0
0.022
0.016
0.016

2014-2026
0.02
0
0.022
0.016
0.016

pm25
2010-2013
0.26
0
0.24
0.17
0.17

2014-2026
0.32
0
0.29
0.21
0.21

THC
2010-2013
0
0
0
0
0

2014-2026
0
0
0
0
0

ch4
2010-2013
0
0
0
0
0

2014-2026
0
0
0
0
0

NMHC
2010-2013
0
0
0
0
0

2014-2026
0
0
0
0
0

TOG
2010-2013
0
0
0
0
0
Starts
2014-2026
0
0
0
0
0
C02
2010-2013
0.03
0
0.03
0.02
0.02

2014-2026
0.03
0
0.03
0.02
0.02

CO
2010-2013
0.16
0
0.15
0.11
0.11

2014-2026
0.16
0
0.15
0.11
0.11

NOx
2010-2013
0
0
0
0
0

2014-2026
0
0
0
0
0

pm25
2010-2013
0
0
0
0
0

2014-2026
0
0
0
0
0
250

-------
Table 6-15 (Continued) Crankcase/Tailpipe Ratios for Model Year 2010-2060 Heavy-Duty Diesel Regulatory
Classes
Process
Pollutant
Model Year Group
HD
baseline
Crankcase/Tailpipe ratio by regulatory class


LHD45
MHD
HHD
Urban Bus

THC
2010-2013
0.16
0
0.15
0.11
0.11

2014-2026
0.28
0
0.25
0.18
0.18

ch4
2010-2013
0.005
0
0.005
0.004
0.004

2014-2026
0.009
0
0.008
0.006
0.006

NMHC
2010-2013
0.26
0
0.24
0.17
0.17

2014-2026
0.44
0
0.4
0.29
0.29

TOG
2010-2013
0.17
0
0.15
0.11
0.11
Extended
2014-2026
0.28
0
0.26
0.19
0.19
Idle
C02
2010-2013
0.015
0
0.013
0.01
0.01

2014-2026
0.015
0
0.013
0.01
0.01

CO
2010-2013
0.006
0
0.006
0.004
0.004

2014-2026
0.006
0
0.006
0.004
0.004

NOx
2010-2013
0.017
0
0.015
0.011
0.011

2014-2026
0.017
0
0.015
0.011
0.011

PM2.5
2010-2013
1.11
0
1.01
0.74
0.74

2014-2026
1.82
0
1.64
1.21
1.21
As noted above, the PM2.5 crankcase emission ratios are calculated using the ACES Phase 1
crankcase emission rates (MY 2007) and the PM2.5 exhaust rates from the 2010-2013 and 2014+
NOx FEL 0.2 groups. As the tailpipe PM2.5 exhaust emission rates are based upon in-use tailpipe
testing for 2010+, the tailpipe ratio for all processes is 1, and the crankcase PM2.5 ratio is a simple
fraction of the tailpipe emissions as shown in Table 6-16 for MHD and HHD diesel vehicles. The
fraction for LHD is zero, consistent with Table 6-15. Since PM2.5 uses ACES Phase 1 data, starts
emissions are assumed to be included in the running and the crankcase starts are set to zero.
251

-------
Table 6-16. MOVES Exhaust and Crankcase Ratios for 2010-2026 MHD and HHD Diesel by Pollutant, Process,
		and Model Year Group for PM2.5 Species		
Process
Model Year
Group
Pollutant
Start
Running
Extended
Idle


EC
1
1
1

2010-2013
nonECnonS04PM
1
1
1

S04
1
1
1
Tailpipe

H20
1
1
1
Exhaust

EC
1
1
1

2014-2026
nonECnonS04PM
1
1
1

S04
1
1
1


H20
1
1
1


EC
0
0.17
0.74

2010-2013
nonECnonS04PM
0
0.17
0.74

S04
0
0.17
0.74
HHD

H20
0
0.17
0.74
Crankcase

EC
0
0.21
1.21

2014-2026
nonECnonS04PM
0
0.21
1.21

S04
0
0.21
1.21


H20
0
0.21
1.21


EC
0
0.24
1

2010-2013
nonECnonS04PM
0
0.24
1

S04
0
0.24
1
MHD

H20
0
0.24
1
Crankcase

EC
0
0.29
1.64

2014-2026
nonECnonS04PM
0
0.29
1.64

S04
0
0.29
1.64


H20
0
0.29
1.64
6.2.2.4 202 7-2060 Model Years
For MY2027 and later vehicles, the HD2027 standards require manufacturers to use one of two
options for controlling crankcase emissions, either: 1) closing the crankcase, or 2) an updated
version of the current requirements for an open crankcase that includes additional requirements for
measuring and accounting for crankcase emissions.
In the emissions impact analysis of the HD2027 rule, we assumed that closing the crankcase would
be the preferred option to meet the standards. We revised the crankcase emission rates in MOVES4
accordingly by setting the crankcase emission rates for MY2027+ HHD, MHD and LHD45 diesel
vehicles to zero.
252

-------
6.2.3 Glider Crankcase Emissions
Glider vehicles in MOVES are modeled using emission rates from MY 2000 heavy-duty diesel
engines (Section 2.5); the pre-2007 crankcase rates (Table 6-8 and Table 6-9) are applied to glider
vehicles (regClassID 49) for all model years.
6.3 Heavy-Duty Gasoline and CNG Crankcase Emissions
The data on heavy-duty gasoline and CNG crankcase emissions are limited. All 1969 and later
spark ignition heavy-duty engines are required to control crankcase emissions. All gasoline engines
are assumed to use positive crankcase ventilation (PCV) systems, which route the crankcase gases
into the intake manifold. For heavy-duty gasoline engines we use the same values of crankcase
emission ratios as light-duty gasoline as shown in Table 6-17; these are documented in the MOVES
light-duty emission rates report.9 The HD2027 standards do not affect heavy-duty gasoline and
CNG fueled engines since the rule only affects the crankcase emissions from the compression-
ignition (diesel) heavy-duty engines.
For the 1969 and later vehicles, we assume 4 percent of PCV systems fail, which would cause
increased tailpipe emissions for reasons such as misfiring of the engine, lubricating oil in the intake
manifold, and increased deterioration of the three-way catalyst. Although these processes will
increase tailpipe emissions, in MOVES we model this increase of emissions due to a failed PCV as
crankcase emissions. We assume that the elevated emissions due to a failed PCV system would be
equivalent to the crankcase emissions of a pre-1969 vehicles using Equation 6-8. The resulting
fleet-wide crankcase to exhaust emission ratios for 1969 and later vehicles are shown in Table
6-17.
Crankcase Ratio1969+ = PCV failure rate (4%) x Crankcase Ratioprel969 Equation 6-8
Table 6-17 Crankcase to Tailpipe Exhaust Emission Ratio for Heavy-Duty Gasoline and CNG Vehicles for THC,
	CO, NOx, and PM2 s	
Pollutant
pre-1969
1969 and later
HC
0.33
0.013
CO
0.013
0.00052
NOx
0.001
0.00004
PM (all species)
0.20
0.008
Due to limited information, we used the gasoline heavy-duty crankcase emission factors for heavy-
duty CNG engines because the majority of these engines are spark-ignited. However, at least one
study (Clark et al., 2017)164 suggests that CNG vehicles have open crankcase systems so we may
be underestimating crankcase emissions, especially those of methane. We hope to revisit CNG
crankcase emissions in future versions of MOVES.
The crankcase and exhaust ratios used by the crankcase calculator for PM2.5 emissions from heavy-
duty gasoline and compressed natural gas vehicles are provided in Table 6-18. These values are
applied to calculate crankcase emissions associated with start exhaust as well as to running
253

-------
exhaust. No information is available to estimate separate speciation between exhaust and
crankcase, so the factors are the same for all PM subspecies.
Table 6-18 MOVES Exhaust and Crankcase Ratios for Heavy-Duty Gasoline and CNG Vehicles by Pollutant,
Pollutant
Process
1960-1968
1969-2050
EC
Exhaust
1
1
nonECnonSC^PM
1
1
S04
1
1
H20
1
1
EC
Crankcase
0.2
0.008
nonECnonSC^PM
0.2
0.008
S04
0.2
0.008
H20
0.2
0.008
254

-------
7 Nitrogen Oxide Composition
This section discusses the values used to estimate nitric oxide (NO), nitrogen dioxide (NO2) and
nitrous acid (HONO) from nitrogen oxide (NOx) emissions from heavy-duty vehicles. A similar
section on NOx composition from light-duty emissions is included in the light-duty emissions
report. NOx emissions are reported in mass-equivalent space of NO2. In other words, the molar
mass of NO2 (46 g/mole) is used to calculate grams of NOx from the molar concentration of NOx.
Nitrogen oxides (NOx) are defined as NO + NO2.165J66 NOx is considered a subset of reactive
nitrogen species (NOy) with a nitrogen oxidation state of +2 or greater which contain other nitrogen
containing species (NOz), thus NOy = NOx + NOz.165 NOz compounds are formed in the atmosphere
as oxidation products of NOx.166
Chemiluminescent analyzers used for exhaust NOx measurements directly measure NO, as NO is
oxidized by ozone to form NO2 and produces florescent light. Chemiluminescent analyzers measure
NOx (NO + NO2) by using a catalyst that reduces the NO2 to NO in the sample air stream before
measurement. NO2 is calculated as the difference between NOx and NO measurements. The NOx
converter within chemiluminescent analyzers can also reduce other reactive nitrogen species (NOz),
including HONO to NO. If the concentrations of NOz-interfering species in the sample stream are
significant relative to NO2 concentrations, then they can bias the NO2 measurements high.167
MOVES estimates NO and NO2 by applying an NO/NOx or NO2/NOX fraction to the NOx emission
rates. The NO/NO2 and NO2/NOX fractions are stored in a MOVES table called nono2ratio. The
nono2ratio enables the nitrogen oxide composition to vary according to source type, fuel type,
model year, and pollutant process. However, the current NOx fractions in MOVES vary only
according to fuel type, model year, and emission process.
MOVES also estimates one important NOz species, nitrous acid (HONO), from the NOx values.
HONO emissions are estimated as a fraction (0.8 percent) of NOx emissions from all vehicle types
in MOVES, based on HONO and NOx measurements made at a road tunnel in Europe.168 HONO
emissions are also estimated using the nono2ratio MOVES table. For each source type, fuel type,
and emission process, the NO, NO2, and HONO values in the nono2ratio sum to one. Future work
could be conducted to update MOVES to model NOx and HONO fractions according to regulatory
class.
MOVES users should be aware that the definition of NOx in MOVES (NO+NO2+HONO) is
different than the standard NOx definition of NOx (NO + NO2). In MOVES, we include HONO in
the NOx values, because the chemiluminescent analyzers are biased slightly high by HONO in the
exhaust stream, and HONO is formed almost immediately upon dilution into the roadway
environment from NO2 emissions. To avoid overcounting reactive nitrogen formation, we include
HONO in the sum of NOx in MOVES. MOVES users should consider which measure they would
like to use depending on their use-case. For example, for comparing NOx results with a vehicle
emission test program, MOVES users may want to simply use NOx (pollutantID 3), whereas
MOVES users developing air quality inputs of NOx, NO2, and HONO, may estimate NOx as the
sum of NO + NO2 (pollutantlDs 32 and 33), rather than using the direct NOx output in MOVES
(polluantID 3).
255

-------
7.1 Heavy-Duty Diesel
The heavy-duty diesel NO/NOx, NCh/NOx, and HONO/NOx fractions were updated in MOVES4
using data reported from recent emission studies, as described below.
We summarized NCh/NOx fractions from three recent studies by aftertreatment technology and
model year range shown in Table 7-1.
Preble et al. (2019)139 sampled individual heavy-duty vehicle exhaust plumes at the entrance to the
Caldecott Tunnel near Oakland, California and at the Port of Oakland for multiple years. The data
from Preble et al. (2019) are also used to update the NH3 emission rates as discussed in Section 5.1,
however NH3 was only measured in 2018 at the Caldecott TunnelwhileNO and NO2 were measured
at both locations for multiple years. Thirugengadam et al. (2015)51 conducted exhaust sampling of
five heavy-duty diesel vehicles measured on four different driving cycles used to represent goods
movement in Southern California. Quiros et al. (2016)49 sampled six heavy-duty diesel tractors
hauling a mobile emissions laboratory trailer. They sampled the vehicles along six routes intended
to represent goods movement in Southern California. The Advanced Collaborative Emissions
Study conducted by Khalek et al. (2009)47 and (2013)169, tested four model year 2007 and three
model year 2010 heavy-duty diesel engines using an engine dynamometer.
The NO2/NOX fraction measured by aftertreatment technology and model year ranges are quite
consistent across the four different studies. This suggests that the Mh/NOx obtained from the
plume capture measurements in Preble et al. (2019)139 are relevant for the wide range of operation
conditions sampled in Thirugengadam et al. (2015)51 and Quiros et al. (2016).49
Each of the studies showed that the NO2/NOX increased with the introduction of diesel particulate
filters (DPF) in model years 2007-2009. This is expected because DPF aftertreatment systems are
designed to increase the fraction of N02to facilitate passive regeneration of the DPF. A diesel
oxidation catalyst upstream of the DPF oxides NO to NO2 which then oxidize soot collected on the
filter.170 The DPF+SCR aftertreatment systems introduced with MY 2010 and later engines also
have higher NO2/NOX fractions than pre-DPF engines, but are consistently lower than the DPF only
engines.
256

-------
Table 7-1. NCh/NOx ratios (± 95% Confidence Intervals, if available) from heavy-duty diesel vehicles reported
from recent studies
Study
Study Description
Sample
Size
Aftertreatment
Engine
Model Year
NO2/NOX
Preble et al.
(2019)139
Caldecott Tunnel near
Oakland California,
Plume-Capture, Sample
Years: 2014, 2015,2018
1,471
DPF + SCR
2010-2018
0.19 ±0.03
780
DPF
2007-2009
0.24 ± 0.02
359
DPF Retrofit
1994-2006
0.11 ±0.02
190
No DPF
2004-2006
0.06 ±0.01
454
No DPF
1965-2003
0.03 ±0.01
Preble et al.
(2019)139
Port of Oakland,
California, Plume-Capture,
Sample Years: 2011, 2013,
2015
403
DPF + SCR
2010-2016
0.20 ±0.05
1,598
DPF
2007-2009
0.23 ±0.02
399
DPF Retrofit
1994-2006
0.15 ±0.02
199
No DPF
2004-2006
0.04 ± 0.02
Thiruvengadam
et al. (2015)51
Chassis dynamometer on
four duty cycles
representative of goods
movement
1
DPF + SCR
2010-2011
-0.15
1
DPF
2011
-0.30
1
DPF
2009
-.30
Quiros et al.
(2016)49
Six good movements
routes in Southern
California sampled using
mobile laboratory
4
DPF + SCR
2013-2014
0.19 ± 0.17
1
DPF (Hybrid Diesel)
2011
0.33
1
DPF
2007
0.30
Khalek et al.
(2013)169
ACES engine
dynamometer study, 16-
hour cycle
3
DPF + SCR
2011
0.52 ±0.45
Khalek et al.
(2009)47
4
DPF
2007
0.54 ±0.20
In MOVES4, we used the NO/NOx and NCh/NOx fractions from the Caldecott Tunnel (Preble et al.,
2019).139 For model years 2004-2010, the updated values are similar to the values used in previous
versions of MOVES. For model years earlier and later, the MOVES4 NCh/NOx values are lower.
We did not use the DPF retrofit values because these are representative of California drayage
vehicles starting in 2010, but not the nation-wide fleet of heavy-duty pre-2007 vehicles.
MOVES3 and earlier versions use a HONO fraction of 0.8% obtained from Kurtenbach et al.
(2001).168 Table 7-2 summarizes HONO/NOx ratios from an updated literature review. Studies that
measure HONO often don't measure individual vehicle exhaust, thus isolating the diesel specific
HONO ratio is difficult. MOVES HONO/NOx ratios was not updated using this data as the
MOVES3 value of 0.8% is well-within the range of the diesel-only HONO/NOx measurements.
257

-------
Table 7-2. Fleet-average and diesel specific HONO/NOx Ratios
Source
Study Type
HONO/NOx (%)
Diesel fleet (%)
Kramer etal. (2020)171
Road tunnel in the
United Kingdom
1.04
Isolated diesel vehicle ratio
0.85
66
Liang et al. (2017)172
Road tunnel in Hong
Kong
1.24
33
Xu et al. (2015)173
Ambient
measurements in
Hong Kong
1.20
33
Trinhetal. (2017)174
Chassis dynamometer
across four drive
cycles
0.16 to 1
Diesel vehicle equipped with DPF
tested
Rappengltick et al.
(2013)175
Road-side
measurements in
Houston, Texas
1.17
5-10
Kurtenbach et al.
(2001)168
Tunnel Study in
Germany
0.80
6% heavy-duty vehicles, 6%
commercial vans, 12.3% diesel
passenger vehicles
Single-vehicle
Tunnel Study
0.53
Diesel truck
Single-vehicle
Tunnel Study
0.66
Diesel passenger car
Table 7-3 shows the NOx and HONO fractions for heavy-duty diesel vehicles used in MOVES4.
Vehicle model years subject to the HD2027 rule use the same fractions as model years 2010-2026,
and APU exhaust fractions are the same for all model years 2024 and later. The NO/NOx and
NCh/NOx fractions reported in Preble et al. (2019) were renormalized to account for the 0.8 percent
HONO emissions. The NOx fractions are the same across all diesel source types and across all
emission processes (running, start, extended idle), except for auxiliary power units, which use the
conventional NOx fractions (1960-2003) for all 1960-2023 model years because it is assumed that
the these APUs are not fitted with diesel particulate filters. APU exhaust rates for the model year
range 2024-2060 use the same NOx fractions as model year 2007-2009 running exhaust, because
we assume they will be equipped with DPF systems but not SCR systems. Because the nono2ratio
table is classified by source type, and not regulatory class, gliders use the same NO/NO2 fractions
as the other regulatory classes by model year, even though the 1960-2003 NOx fractions are more
relevant for this regulatory class. We hope to address this design limitation in future versions of
MOVES.
Finally, while the HD2027 rule as updated in MOVES4 will reduce NOx emissions from MY
2027+ HD gasoline vehicles, we modelled no change in the NO/NO2 fractions.
258

-------
Process
Model Year
NO
N02
HONO
Running exhaust,
start exhaust,
extended idle
exhaust
1960-2003a
0.9622
0.0298
0.008
2004-2006
0.9325
0.0595
0.008
2007-2009
0.7539
0.2381
0.008
2010-2060
0.8035
0.1885
0.008
Auxiliary power unit
exhaust
1960-2023
0.9325
0.0595
0.008
2024-2060
0.7539
0.2381
0.008
7.2 Heavy-Duty Gasoline
The NOx fractions for heavy-duty gasoline are based on the MOVES values used for light-duty
gasoline estimates. Separate values are used for running and start emission processes. As stated in
the MOVES2010 report,176 the light-duty values are shifted to later model year groups to be
consistent with heavy-duty emission standards and emission control technologies. These values are
shown in Table 7-4 for both light-duty and heavy-duty gasoline vehicles. The NO2 fractions
originally developed for MOVES2010 were reduced by 0.008 to account for the HONO
emissions.176 While the HD2027 rule will reduce NOx emissions from HD gasoline vehicles, we
modelled no change in the NOx fractions for MY 2027+.
Table 7-4 NOx and HONO Fractions for Light-Duty (Source Type 21,31,32) and Heavy-Duty Gasoline Vehicles
Light-Duty
gasoline
model year
groups
Heavy-Duty
gasoline
model year
groups
Running
Start
NO
N02
HONO
NO
N02
HONO
1960-1980
1960-1987
0.975
0.017
0.008
0.975
0.017
0.008
1981-1990
1988-2004
0.932
0.06
0.008
0.932
0.031
0.008
1991-1995
2005-2007
0.954
0.038
0.008
0.987
0.005
0.008
1996-2060
2008-2060
0.836
0.156
0.008
0.951
0.041
0.008
7.3 Heavy-Duty Compressed Natural Gas
We used the average NO2/ NOx fractions reported from three CNG transit buses with DDC Series
50 G engines by Lanni et al. (2003)126 with the 0.008 HONO fraction assumed for other fuel types,
to estimate the NOx fractions of NO, NO2, and HONO. These assumptions yield the values in Table
7-5, which are used for CNG heavy-duty vehicles of all model years. In the future, we hope to
update these values with data from more recent three-way catalyst CNG vehicles.49
Table 7-5 NOx and HONO Fractions CNG Heavy-Duty Vehicles
Model Year
NO
N02
HONO
1960-2060
0.865
0.127
0.008
259

-------
8 Appendices

-------
Appendix A Calculation of Accessory Power Requirements
Table A-l. Accessory Load Estimates for HHP Trucks
VSP
Cooling Fan
Air cond
Air comp
Alternator
Engine
Accessories
Total Accessory Load (kW)
Low
Power (kw)
% time on
19.0
10%
2.3
50%
Off = 0.5 kW
3.0
60%
1.5
100%
1.5
100%

Total (kW)
1.9
1.2
2.0
1.5
1.5
8.1
Mid
Power (kw)
% time on
19.0
20%
2.3
50%
Off = 0.5 kW
2.3
20%
1.5
100%
1.5
100%

Total (kW)
3.8
1.2
0.9
1.5
1.5
8.8
High
Power (kw)
% time on
19.0
30%
2.3
50%
Off = 0.5 kW
2.3
10%
1.5
100%
1.5
100%

Total (kW)
5.7
1.2
0.7
1.5
1.5
10.5
Table A-2. Accessory Load Estimates for MHD Trucks
VSP
Cooling Fan
Air cond
Air comp
Alternator
Engine
Accessories
Total Accessory Load (kW)
Low
Power (kw)
% time on
10.0
10%
2.3
50%
Off = 0.5 kW
2.0
60%
1.5
100%
1.5
100%

Total (kW)
1.0
1.2
1.4
1.5
1.5
6.6
Mid
Power (kw)
% time on
10.0
20%
2.3
50%
Off = 0.5 kW
2.0
20%
1.5
100%
1.5
100%

Total (kW)
2.0
1.2
0.8
1.5
1.5
7.0
High
Power (kw)
% time on
10.0
30%
2.3
50%
Off = 0.5 kW
2.0
10%
1.5
100%
1.5
100%

Total (kW)
3.0
1.2
0.7
1.5
1.5
7.8
Table A-3. Accessory Load Estimates for Buses
VSP
Cooling Fan
Air cond
Air comp
Alternator
Engine
Accessories
Total Accessory Load (kW)
Low
Power (kw)
% time on
19.0
10%
18.0
80%
Off = 0.5 kW
4.0
60%
1.5
100%
1.5
100%

Total (kW)
1.9
14.4
2.6
1.5
1.5
21.9
Mid
Power (kw)
% time on
19.0
20%
18.0
80%
Off = 0.5 kW
4.0
20%
1.5
100%
1.5
100%

Total (kW)
3.8
14.4
1.2
1.5
1.5
22.4
High
Power (kw)
% time on
19.0
30%
18.0
80%
Off = 0.5 kW
4.0
10%
1.5
100%
1.5
100%

Total (kW)
5.7
14.4
0.9
1.5
1.5
24.0
261

-------
Appendix B Tampering and Mal-maintenance for Diesel Running
Exhaust
Tampering and mal-maintenance (T&M) effects represent the fleet-wide average increase in
emissions as the fleet ages. In laboratory testing, properly maintained engines often yield very
small rates of emissions deterioration through time. We assume that in real-world use, tampering
and mal-maintenance dominate emissions deterioration over time for heavy-duty diesel vehicles.
As a result, MOVES specifically models the deterioration due to tampering and mal-maintenance,
which we assume also includes any other emission increases due to vehicle aging and deterioration.
The tampering and mal-maintenance methodology was first incorporated into MOVES2010177
from studies conducted between 1988 and 2007 (See Section B.2.) The T&M methodology used in
MOVES3 is unchanged from MOVES2010 along with much of the original T&M assumptions on
T&M frequency and T&M adjustment factors. Slight updates to NOx and PM2.5 T&M adjustment
factors were made for MOVES2014.35 No changes were made to the T&M assumptions or data
between MOVES2014 and MOVES3. Minor corrections to the warranty and useful life of LHD
vehicles were made in MOVES4. In MOVES4, we also updated the warranty and useful life of the
MY2027+ heavy-duty vehicles based on the HD2027 standards and we calculated new T&M
percentage effects for these vehicles that assume aftertreatment failure in a MY2027 vehicle would
bring NOx tailpipe emissions to the same level as a MY 2010 vehicle with the same failure.
In the future, T&M adjustment factors in MOVES should be re-evaluated and updated, particularly
to incorporate data on the durability and emissions performance of advanced aftertreatment systems
on modern heavy-duty diesel vehicles and to account for recent work surveying intentional
tampering in diesel pickup trucks.178
This section describes the derivation of T&M emission rates applied to diesel running exhaust. The
estimation of heavy-duty gasoline deterioration is discussed in the derivation of the heavy-duty
gasoline rates (Section 3). The derivation of the T&M effects for diesel extended idle emissions are
described in Section 2.3.
B. 1 Modeling Tampering and Mal-maintenance
As T&M affects emissions through age, we developed a simple function of emission deterioration
with age. New vehicles and 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. 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 & mal-maintenance (T&M) adjustment factor. Figure B-l shows this relationship.
The actual emission levels were determined through data analysis detailed below.
262

-------
period
Figure B-l. Qualitative Depiction of the Implementation of Age Effects
The T&M adjustment factor is calculated as the sum of the product of the T&M frequency for each
failure and the corresponding T&M emission effect, as shown in Equation 8-1.
fr&M.p = / < (T&M frequency; X T&M emission effectpi)	Equation 8-1
Where:
fT&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 effect^ estimated emission effect for pollutant p associated with tampering & mal-
maintenance failure i.
The emission rate at the end of useful life is then calculated using Equation 8-2.
E^End of useful life,p,r,o ERzero mile,p,r,o ^ (l ^T&M.p)
Equation 8-2
Where:
EREnd of useful 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
fT&M= the tampering and mal-maintenance adjustment factor for each pollutant p (Equation 8-1)
263

-------
The useful life refers to the length of time that engines are required to meet emissions standards.
We incorporated this age relationship by averaging emissions rates across the ages in each age
group. Mileage was converted to age with VIUS179 (Vehicle Inventory and Use Survey) data,
which contains data on how quickly trucks of different regulatory classes accumulate mileage.
Table B-l shows the emissions warranty period and approximate useful life requirement period for
each of the regulatory classes for pre-MY2027 vehicles. This table and the resulting values in
MOVES have been corrected in MOVES4 to reflect the 50,000 miles warranty requirement for
LHD vehicles, and the 150,000 useful life mileage for Tier 3 LHD2b3 vehicles.
Table B-l. Warranty and Useful Life Requirements by Regulatory Class For Pre-MY2027 Vehicles
Regulatory class
Warranty
requirement
mileage/age
requirementb
Calculated warranty
age3
Useful life
mileage/age
requirement
Calculated
useful life
age3
Assumed
mileage
per year
LHD2b3 (Tier 2
and earlier)
50,000/5
2
120,000/1 lc
5
26,000
LHD2b3 (Tier 3)
50,000/5
2
150,000/10 d
6
26,000
LHD45
50,000/5
2
110,000/10e
4
26,000
MHD
100,000/5
2
185,000/10 e
5
41,000
HHD
100,000/5
1
435,000/10 e
4
105,000
BUS
100,000/5
2
435,000/10 f
10
44,000
Notes:
a The calculated warranty age and useful life age here are based on typical miles driven by vehicles in the regulatory
class. For example, HHD vehicles typically accumulate a large number of miles per year (100,000+/year). Thus, HHD
vehicles complete their warranty and useful life requirements based on mileage while the vehicle age is still much
below the requirement.
b 40 CFR 1037.120
c 40 CFR 86.096-2, 40 CFR 86.1805-12
d 40 CFR 86.1805-17
e 40 CFR 86.001-2 (4). The useful life mileage is the same for each regulatory class for all exhaust pollutants (NOx,
HC, CO, and PM). The useful life age requirement is generally 10 years for NOx, while it is 8 years for the other
pollutants (and for NOx in 1996-1997). However, we calculated that the mileage requirement is the forcing requirement
for all the heavy-duty regulatory classes.
f 40 CFR 86.098-2. The usefule life standard is 10 years for urban buses for both NOx and PM.
Starting from MY2027, the HD2027 standards require manufacturers to comply with new warranty
and useful life provisions as shown in Table B-2.
264

-------
Table B-2 Warranty and Useful Life Requirements by Regulatory Class For MY2027+ Vehicles
Regulatory class
Warranty
requirement
mileage/age
requirement
Calculated warranty
age
Useful life
mileage/age
requirement
Calculated
useful life
age
Assumed
mileage
per year
LHD2b3
Same as Pre-MY2027+ Vehicles
LHD45
210,000/10
8
270,000/15
10
26,000
MHD
280,000/10
7
350,000/12
9
41,000
HHD
450,000/10a
4
650,000/11
6
105,000
BUS
450,000/10a
10
650,000/11
15
44,000
a The HHD diesel and Urban Bus warranty year values in the tables were updated to 10 years in MOVES4 to be consistent
with HD2027 final rule.
While both age and mileage metrics are given for these periods, whichever comes first determines
the applicability of the warranty. As a result, since the mileage limit is usually reached before the
age limit, but MOVES deals with age and not mileage, we needed to convert all the mileage values
to age equivalents. The data show that on average, heavy heavy-duty trucks accumulate mileage
much more quickly than other regulatory classes and reach the end of their warranty period more
quickly. Therefore, deterioration in heavy heavy-duty truck emissions will presumably happen at
younger ages than for other regulatory classes. Buses, on average, do not accumulate mileage as
quickly. Therefore, their useful life period is governed by the age requirement, not the mileage
requirement.
We use a "scaled age effect" to calculate the age-adjusted emission rates for each age. 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 B-2. When the vehicle age is between the end of the warranty
and the useful life, sa is interpolated between 0 and 1 as summarized in Table B-3 below, and
illustrated in Figure B-l above.
Table B-3. Calculation of sa
Where:

age
< end of warranty age
0
end of warranty age
< age < useful life
{age — end of warranty age)
(Useful life age — end of warranty age)
age > useful life
1
Since MOVES deals with age groups and not individual ages (Table 1-6), the increase in emissions
by age must be calculated by age group. For simplicity, we modeled an even age distribution within
each age group (e.g. ages 0, 1,2, and 3 are equally represented in the 0-3 age group). We then
calculated average scaled age effects for each age group. This is important since, for example,
HHD trucks reach their useful life at four years, which means they will increase emissions through
the 0-3 age group. As a result, the 0-3 age group emission rate will be higher than the zero-mile
emission rate for HHD trucks. Table B-4 and Table B-5 show the average scaled age effect by age
group for pre-2027 and 2027+ MY vehicles, respectively. In these tables, a value of 0 indicates no
265

-------
deterioration, (i.e., the zero-mile emissions level (ZML)), and a value of 1 indicates a fully
deteriorated engine, or maximum emissions level, at or beyond the useful life (UL).
Table B-4. Average Scaled Age Effect, Sa For Pre-MY2027 Vehicles
Age Group
LHD2b3
(Tier 2 and
earlier)
LHD2b3
(Tier 3)
LHD45
MHD
HHD
Bus
0-3
0.0833
0.0625
0.125
0.083
0.25
0.0313
4-5
0.8333
0.6250
1
0.833
1
0.3125
6-7
1
1
1
1
1
0.5625
8-9
1
1
1
1
1
0.8125
10-14
1
1
1
1
1
1
15-19
1
1
1
1
1
1
20+
1
1
1
1
1
1
Table B-5 Average Scaled Age Effect, Sa For MY2027+ Vehicles
Age Group
LHD2b3
(Tier 2 and
earlier)
LHD2b3
(Tier 3)
LHD45
MHD
HHD
Bus
0-3


0
0
0.00
0
4-5


0
0
0.25
0
6-7
Not subject to
HD2027 rule.
Same as Pre-
0
0
1.00
0
8-9
MY2027
0.25
0.75
1.00
0
10-14
Vehicles
1
1
1.00
0.75
15-19


1
1
1.00
1
20+


1
1
1.00
1
Then, for each pollutant and age, we multiplied the zero-mile emission rate by one plus the product
of the average scaled age effect and the T&M adjustment factor.
ERp,r,a,o — ERzero mile,p,r,o ^ (1 + Sa X fj&M)	Equation 8-3
Where:
ERp r,o,a = the heavy-duty diesel emission rate for each pollutant p, regulatory class r, age a, operating mode,
o,
ERzero miie~ the zero-mile heavy-duty diesel emission rate for each pollutant p, regulatory class r, operating
mode, o
s^= average scaled age effect at age group, a
fx&M= the tampering and mal-maintenance adjustment factor (Equation 8-1)
Sections B.2 through B.9 discuss the data sources and assumptions used to determine the T&M
failure frequencies and T&M emission effects used to derive the T&M adjustment factor in
Equation 8-1 for each pollutant and model year range of vehicle.
266

-------
B.2 Data Sources
EPA used the following information to develop the tampering and mal-maintenance occurrence
rates used in MOVES:
•	California's ARB EMFAC2007 Modeling Change Technical Memo180 (2006). The
basic EMFAC occurrence rates for tampering and mal-maintenance were developed
from Radian and EFEE reports and CARB engineering judgment.
•	Radian Study (1988). The report estimated the malfunction rates based on survey and
observation. The data may be questionable for current heavy-duty trucks due to
advancements such as electronic controls, injection systems, and exhaust aftertreatment.
•	EFEE report (1998) on PM emission deterioration rates for in-use vehicles. Their work
included heavy-duty diesel vehicle chassis dynamometer testing at Southwest Research
Institute.
•	EMFAC2000 (2000) Tampering and Mal-maintenance Rates
•	EMA's comments on ARB's Tampering, Malfunction, and Mal-maintenance
Assumptions for EMFAC 2007
•	University of California -Riverside (UCR) "Incidence of Malfunctions and Tampering
in Heavy-Duty Vehicles"
•	Air Improvement Resources, Inc.'s Comments on Heavy-Duty Tampering and Mal-
maintenance Symposium
B. 3 T&M Failure Modes
EPA generally adopted the T&M failure modes developed by CARB, with a few exceptions. The
high fuel pressure category was removed. We added a failure mode for mis-fueling to represent the
use of nonroad diesel in cases when ULSD onroad diesel is required. We combined the injector
failure modes into a single group. We reorganized the EGR failure modes into "Stuck Open" and
"Disabled/Low Flow." We included the PM regeneration system, including the igniter, injector, and
combustion air system in the PM filter leak failure mode.
For model years 1994-2007, the EPA developed failure mode frequencies for model year groups
that apply to all heavy-duty diesel vehicles, including earlier model years. For model year 2007-
2012, we developed separate failure mode frequencies for heavy-duty diesel vehicles that are
equipped with Lean NOx Traps (LNT) and Selective Catalyst Reduction (SCR) systems,
respectively. Beyond model year 2012, we assume all heavy-duty vehicles are using SCR systems.
Better understanding tampering and mal-maintenance effects in contemporary vehicles is an area
where additional research would be beneficial.
267

-------
B.4 T&MModel Year Groups
EPA developed the model year groups based on regulation and technology changes.
•	Pre-1994 represents non-electronic fuel control.
•	1998-2002 represents the time period with consent decree issues.
•	2003 represents early use of EGR.
•	2007 and 2010 contain significant PM and NOx regulation changes.
•	2010-and later represent heavy-duty trucks with required OBD. This rule began in MY
2010 with complete phase-in by MY 2013. The OBD impacts are discussed in Section
B.10.
•	2027 and later represent the heavy-duty vehicles subject to the HD2027 standards
including further technology improvements of heavy-duty engines and after-treatment
system.181
B. 5 T&M Failure Frequency Rates and Differences
EPA adopted the CARB EMFAC2007 occurrence rates, except as noted below.
Clogged Air Filter: EPA reduced the frequency rate from EMFAC's 15 percent to 8 percent. EPA
reduced this value based on the UCR results, the Radian study, and EMA's comments that air
filters are a maintenance item. Many trucks contain indicators to notify the driver of dirty air filters
and the drivers have incentive to replace the filters for other performance reasons.
Other Air Problems: EPA reduced the frequency rate from 8 percent to 6 percent based on the
UCR results.
Electronics Failed: EPA continued to use the 3 percent frequency rate for all model years beyond
2010. We projected that the engine hardware would evolve through 2010, rather than be replaced
with completely new engine systems that would justify a higher rate of failure. For 2010 and later
vehicles, the occurrence of T&M on electronics associated with SCR and DPF aftertreatment
systems is counted with the aftertreatment specific failure modes (including "NOx aftertreatment
malfunction" and "PM Filter Disable"), rather than in the "Electronics Failed" mode.
EGR Stuck Open: EPA believes the failure frequency of this item is rare and therefore set the
level at 0.2 percent. This failure will lead to drivability issues that will be noticeable to the driver
and serve as an incentive to repair.
EGR Disabled/Low Flow: EPA estimates the ERG failure rate at 10 percent. All but one major
engine manufacturer had EGR previous to the 2007 model year and all have it after 2007, so a large
increase in rates seem unwarranted. However, the Illinois EPA stated that "EGR flow insufficient"
is the top OBD issue found in their LDVI/M program182 so it cannot be ignored.
NOx Aftertreatment malfunction: EPA developed a NOx aftertreatment malfunction rate that is
dependent on the type of system used. We assumed that HHDD will use primarily SCR systems
and LHDD will primarily use LNT systems. We estimated the failure rates of the various
components within each system to develop a composite malfunction rate (Table B-6).
The individual failure rates were developed considering the experience in agriculture and stationary
industries of NOx aftertreatment systems and similar component applications. Details are included
in the chart below. We assumed that tank heaters had a five percent failure rate but were only
required in one third of the country during one fifth of the year. The injector failure rate is lower
than fuel injectors, even though they have similar technology, because there is only one required in
268

-------
each system and it is operating in less severe environment of pressure and temperature. We believe
the compressed air delivery system is very mature based on a similar use in air brakes. We also
believe that manufacturers will initiate engine power de-rate as incentive to keep the urea supply
sufficient.
Table B-6. NOx Aftertreatment Failure Rates
Occurrence Rate
SCR
Urea tank
0.5%
Tank heaters
1%
In-exhaust injectors
2%
Compressed air delivery to injector
1%
Urea supply pump
1%
Control system
5%
Exhaust temperature sensor
1%
Urea supply
1%
Overall 13%
LNT
Adsorber
7%
In-exhaust injectors
2%
Control system
5%
Exhaust temperature sensor
1%
Overall	16%
NOx aftertreatment sensor: EPA will assume a 10 percent failure mode for the aftertreatment
sensor. We developed the occurrence rate based on the following assumptions:
•	Population: HHDD: vast majority of heavy-duty applications will use selective catalytic
reduction (SCR) technology with a maximum of one NOx sensor. NOx sensors are not
required for SCR - manufacturers can use models or run open loop. Several engine
manufacturers representing 30 percent of the market plan to delay the use of NOx
aftertreatment devices through the use of improved engine-out emissions and emission
credits.
•	Durability expectations: SwRI completed 6000 hours of the European Stationary Cycle
(ESC) cycling with NOx sensor. Internal testing supports longer life durability. Discussions
with OEMs in 2007 indicate longer life expected by 2010.
•	Forward looking assumptions: Manufacturers have a strong incentive to improve the
reliability and durability of the sensors because of the high cost associated with frequent
replacements.
PM Filter Leak: EPA will use 5 percent PM filter leak and system failure rate. They discounted
high failure rates currently seen in the field.
PM Filter Disable: EPA agrees with CARB's 2 percent tamper rate of the PM filter. The filter
causes a fuel economy penalty so the drivers have an incentive to remove it.
Oxidation Catalyst Malfunction/Remove: EPA believes most manufacturers will install
oxidation catalysts initially in the 2007 model year and agrees with CARB's assessment of 5
percent failure rate. This rate consists of an approximate 2 percent tampering rate and 3 percent
269

-------
malfunction rate. The catalysts are more robust than PM filters, but have the potential to experience
degradation when exposed to high temperatures.
Misfuel: EPA estimated that operators will use the wrong type of fuel, such as agricultural diesel
fuel with higher sulfur levels, approximately 0.1 percent of the time.
In the future, we hope to collect updated real-world failure frequencies for newer technologies.
270

-------
B. 6 Tampering & Mal-maintenance Failure Frequency Rate
Summary
Table B-7. T&M Failure Frequency Rate by Model Year Group
Model Year
1994-1997
1998-2002
2003-2006
2007-2009
2007-2012
2010+
NOx Afterteatment Technology:
None
None
None
None
LNT
SCR







Timing Advanced
5%
2%
2.0%
2.0%
2.0%
2.0%
Timing Retarded
3%
2%
2.0%
2.0%
2.0%
2.0%
Injector Problem (all)
28%
28%
13.0%
13.0%
13.0%
13.0%
Puff Limiter Mis-set
4%
0%
0.0%
0.0%
0.0%
0.0%
Puff Limited Disabled
4%
0%
0.0%
0.0%
0.0%
0.0%
Max Fuel High
3%
0%
0.0%
0.0%
0.0%
0.0%
Clogged Air Filter - EPA
8%
8%
8.0%
8.0%
8.0%
8.0%
Wrong/Worn Turbo
5%
5%
5.0%
5.0%
5.0%
5.0%
Intercooler Clogged
5%
5%
5.0%
5.0%
5.0%
5.0%
Other Air Problem - EPA
6%
6%
6.0%
6.0%
6.0%
6.0%
Engine Mechanical Failure
2%
2%
2.0%
2.0%
2.0%
2.0%
Excessive Oil Consumption
5%
3%
3.0%
3.0%
3.0%
3.0%
Electronics Failed - EPA
3%
3%
3.0%
3.0%
3.0%
3.0%
Electronics Tampered
10%
15%
5.0%
5.0%
5.0%
5.0%
EGR Stuck Open
0%
0%
0.2%
0.2%
0.2%
0.2%
EGR Disabled/Low-Flow - EPA
0%
0%
10.0%
10.0%
10.0%
10.0%
NOx Aftertreatment Sensor
0%
0%
0.0%
0.0%
10.0%
10.0%
Replacement NOx Aftertreatment
Sensor
0%
0%
0.0%
0.0%
1.0%
1.0%
NOx Aftertreatment Malfunction -
EPA
0%
0%
0.0%
0.0%
16.0%
13.0%
PM Filter Leak
0%
0%
0.0%
5.0%
5.0%
5.0%
PM Filter Disabled
0%
0%
0.0%
2.0%
2.0%
2.0%
Oxidation Catalyst
Malfunction/Remove - EPA
0%
0%
0.0%
5.0%
5.0%
5.0%
Mis-fuel - EPA
0.1%
0.1%
0.1%
0.1%
0.1%
0.1%
271

-------
B. 7 NOx T&M Emission Effects
B. 7.1 Model Years 1994 through 2026
For model years 1994 through 2026, EPA developed the emission effect from each tampering and
mal-maintenance incident from CARB's EMFAC, Radian's dynamometer testing with and without
the malfunction present, Engine, Fuel, and Emissions Engineering Inc. (EFEE) results, and EPA
staff testing experience.
EPA estimated that the lean NOx traps (LNT) in LHD are 80 percent efficient and the selective
catalyst reduction (SCR) systems in HHD are 90 percent efficient at reducing NOx.
EPA developed the NOx emission factors of the NOx sensors based on SCR systems' ability to run
in open-loop mode and still achieve NOx reductions. The Manufacturers of Emission Controls
Association (MECA) has stated that a 75-90 percent NOx reduction should occur with open loop
control and >95 percent reduction should occur with closed loop control.183 Visteon reports a 60-80
percent NOx reduction with open loop control.184
In testing, the failure of the NOx aftertreatment system had a different impact on the NOx emissions
depending on the type of aftertreatment. The HHD vehicles with SCR systems experienced a 1000
percent increase in NOx during a complete failure, therefore we estimated a 500 percent increase as
a midpoint between normal operation and a complete failure. The LHD vehicles with LNT systems
experienced a 500 percent increase in NOx during a complete failure. We estimated a 300 percent
increase as a value between a complete failure and normal system operation. The values with 0
percent effect in shaded cells represent areas which have no occurrence rate.
As discussed in Section 2.1.1.4.6, we estimate that 25 percent of LHD MY 2007-2009 vehicles
were equipped with LNT aftertreatment systems. For LHD2b3 MY 2010-2012 vehicles, we
modeled that 25 percent of vehicles had LNT aftertreatment systems, and 75 percent had SCR
systems. For LHD2b3 MY 2013+, we assume that all are equipped with SCR aftertreatment
systems. For LHD45, MHD, HHD, and Urban buses, we modeled the model year 2010 and later
T&M effects assuming all engines are equipped with SCR aftertreatment systems. We recognize
this is a simplification as manufacturers produced non-SCR equipped engines in the initial
implementation years of the 2010 standard due to average, banking, and trading, and the EPA
allowance of nonconformance penalty (NCP) engines in 2012.90
B. 7.2 Model Years 2027 and Later
For MY2027+ vehicles, we further adjusted the MY2010 NOx T&M emission effect to reflect the
HD2027 standards in MOVES4. As NOx emissions become more tightly controlled with the
application of advanced technologies to meet the standards, we anticipate the NOx T&M emission
effects will increase (i.e., there will be a relatively larger impact of T&M because the emission
control system is reducing a greater percentage of the NOx produced by the engine).
To estimate the NOx T&M emission effects for the HD2027 standards, we first calculated the
average zero-mile NOx emission rate ERzero mue,NOx prior to the standard based on the weighted
average of the different operating modes o, and regulatory class r, using Equation 8-4.
272

-------
2r,o(ERzero mile,NOx,r,o ^ ^r,o )
2r,o ^r,o
nri	iJ/,U v Zt?/C IIUte.lVUA.I ,U	1 ,U J	r*	. •	i
ERzero mile,NOX =			^	:			Equation 8-4
Where:
ERzero miie,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, o
tr,o= operation time by regulatory class and operating mode estimated by MOVES3.
Next, we estimated the NOx emission rate of MY 2010 vehicles with a tampering and mal-
maintenance failure i, using Equation 8-5, which was derived from Equation 8-2 using the fleet
average emission rate from Equation 8-4 assuming the T&M frequency is 100 percent.
ERt&M \,NOx — ERzero mile.WOv	_	_ _
•* ,	A	A	Equation 8-5
x (1 + T&M emission effect;iNOx)
We then derived Equation 8-6, assuming that a NOx aftertreatment equipment failure i, in the
control scenario, would cause the average of the MY 2027+ failed emission rates, ERT&M i:Nox->t0
be the same as a NOx aftertreatment failure in the baseline MY 2010 case, Baseline ERx&Mi NOx
MY2010 ERx&MjiWOx = MY2027 ERx&MjiWOx	Equation 8-6
MY2010 ERzero miieNOx
x (l + MY2010 T&M emission effect; WOx)
= MY2027 ERzero miie WOx
x (l + MY2027 T&M emission effect; WOx)
By rearranging Equation 8-6, we derived Equation 8-7 to estimate the control scenario NOx T&M
emissions effects.
MY2027 T&M emission effecti W0x
MY2010 ERzero miie,wox x (l + MY2010 T&M emission effect; WOx)
MY2027 ERzero miieWOx
Equation 8-7
- 1
The MY2027 T&M NOx emission effects for the NOx aftertreatment failures are much larger than
the MY2010 values, because the zero-mile NOx emission rate for MY2027 vehicles are lower than
the MY2010 zero-mile NOx emission rates.
273

-------
Table B-8. NOx T&M Emission Effect by Model Year Group
Model Year
1994-
1997
1998-
2002
2003-
2006
2007-
2009
2007-
2012
2010-
2026
2027-
2028
2029-
2060
NOx Afterteatment
Technology:
None
None
None
None
LNT
SCR
SCR
SCR
Federal NOx Emission
Standard (g/bhp-hr)
5.0
5.0
4.0
2.0
0.2
0.2
0.05
0.05
Timing Advanced
60%
60%
60.0%
60.0%
12.0%
6.0%
6.0%
6.0%
Timing Retarded
-20%
-20%
-20.0%
-20.0%
-20.0%
-20.0%
-20.0%
-20.0%
Injector Problem (all)
-5%
-1%
-1.0%
-1.0%
-1.0%
-1.0%
-1.0%
-1.0%
Puff Limiter Mis-set
0%
0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Puff Limited Disabled
0%
0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Max Fuel High
10%
0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Clogged Air Filter - EPA
0%
0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Wrong/Worn Turbo
0%
0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Intercooler Clogged
25%
25%
25.0%
25.0%
5.0%
3.0%
3.0%
3.0%
Other Air Problem - EPA
0%
0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Engine Mechanical Failure
-10%
-10%
-10.0%
-10.0%
-10.0%
-10.0%
-10.0%
-10.0%
Excessive Oil Consumption
0%
0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Electronics Failed - EPA
0%
0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Electronics Tampered
80%
80%
80.0%
80.0%
16.0%
8.0%
8.0%
8.0%
EGR Stuck Open
0%
0%
-20.0%
-20.0%
-20.0%
-20.0%
-20.0%
-20.0%
EGR Disabled/Low-Flow -
EPA
0%
0%
30.0%
50.0%
10.0%
5.0%
5.0%
5.0%
NOx Aftertreatment Sensor1
0%
0%
0.0%
0.0%
200.0%
200.0%
1294%;
1271%;
1620%;
1713% B
1301%;
1277%;
1643%;
1741% B
Replacement NOx
Aftertreatment SensorA
0%
0%
0.0%
0.0%
200.0%
200.0%
1294%;
1271%;
1620%;
1713%B
1301%;
1277%;
1643%;
1741 %B
NOx Aftertreatment
Malfunction - EPAA
0%
0%
0.0%
0.0%
300.0%
500.0%
2688%;
2641%;
3339%;
3527%B
2703%;
2655%;
3386%;
3582%B
PM Filter Leak
0%
0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
PM Filter Disabled
0%
0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Oxidation Catalyst
Malfunction/Remove - EPA
0%
0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Mis-fuel - EPA
0%
0%
0%
0%
0%
0%
0%
0%
A NOx aftertreatment failure modes
B The values in the cells are for LHD45, MHD, HHD, Bus categories, respectively
274

-------
B. 7.3 NOx Adjustment Factor Calculations
Combining the NOx emission effects with the frequency rates results in the initial T&M adjustment
factors shown in the Table B-9 below. This methodology estimated a small (9-14%) T&M NOx
adjustment factor for 2009 and earlier models due to NOx effects of the following failure modes:
electronics tampered, timing advances, intercooler clogged, and ERG disabled/Low Flow.
However, MOVES does not use the estimated NOx T&M emission effects initially estimated for
2009 and earlier model years, and assumes no NOx increase (with the exception of the LNT effect
for LHD explained below). This is indicated in the 3rd column of Table B-9 labeled "(Removed
2009 and earlier)." Instead, MOVES assumes NOx increases only in the vehicles with advanced
NOx exhaust aftertreatment technologies for a few reasons:
•	The Consent Decree Testing conducted by West Virginia University did not show an
increase in NOx emissions with odometer (and consequently, age) during or following
the regulatory useful life.185 Since the trucks in this program were collected from in-
use fleets, we do not believe that these trucks were necessarily biased toward cleaner
engines.
•	Heavy-duty diesel manufacturers often certify zero or low deterioration factors for
these engines.
•	Starting with MY 2010 (2007 for vehicles with LNT), we expect T&M effects to
become much more significant, because a failure in the NOx aftertreatment system will
substantially increase emissions. We decided to initiate modeling of the NOx T&M
adjustment factor with the implementation of the 2010 standards.
The assumption of no T&M NOx increases for pre-2010 heavy-duty vehicles (except LHD with
LNT), extends to glider vehicles for all model years (regClassID 49).
Table B-9. Tampering & Mai-Maintenance NOx Adjustment Factors (fr&M,NOx) (Percent) for Heavy-Duty Diesel
Model years
NOx
Aftertreatment
Technology
fr&M,NOx,nonOBD
(Initial)
fr&M,NOx,nonOBD
(Removed 2009
and earlier for non-
LNT engines)
1994-1997
None
10
0
1998-2002
None
14
0
2003-2006
None
8.7
0
2007-2009
None
10.7
0
2007-2012
LNT
71.5
71.5
2010-2026
SCR
87.4
87.4
2027-2028
SCR
492.2	(LHD45);
483.5 (MHD);
612.7 (HHD);
647.3	(Bus)
492.2	(LHD45);
483.5 (MHD);
612.7 (HHD);
647.3	(Bus)
2029+
SCR
494.9 (LHD45);
486.0 (MHD);
621.3 (HHD);
657.6 (Bus)
494.9 (LHD45);
486.0 (MHD);
621.3 (HHD);
657.6 (Bus)
275

-------
The T&M NOx emission rates for LHD2b3 vehicles equipped with LNT aftertreatment in 2007-
2009 are calculated by first adjusting Equation 2-11 to account for T&M of LNT aftertreatment, as
shown in Equation 8-8. The derivation of Equation 2-11 including the definition of normal
operation frequency and DPF regeneration frequency are discussed in Section 2.1.1.4.6
2007 — 2009 LNT NOx emissions (T&M)
Baseline Emissions
(LNT normal emissions \
= (normal op. frequency) X 				Equation 8-8
\ Baseline emissions J
X (1 + 2007 - 2009 LHD LNT T&M effect)
(Baseline emissions\
+ (DPF reg. frequency) x 			
\ Baseline emission J
= (0.90) x (0.10) x (1.715) + (0.10) x (1) x (1) = 0.2544
Where Baseline Emissions = MOVES2010 MY 2003-2006 NOx emission rates for LHD2b3
Because MOVES does not model LNT vehicles separately, we then calculated an average ratio for
all 2007-2009 LHD2b3 NOx rates (both non-LNT and LNT with T&M) over the baseline 2003-
2006 NOx rates by adjusting Equation 2-12 to account for the T&M effects of LNT, as shown in
Equation 8-9.
2007 — 2009 LHD2b3 NOx emissions (T&M)
Baseline emissions
/2007 — 2009 LNT NOx emissions (T&M)\
= (LNT market share) 				„ .. 0 „
V	Baseline emissions	)	Equation 8-9
+ (non
/2007 — 2009 emission standards\
— LNT market share) 			
\	Baseline emissions	)
= (0.25) x (0.2544) + (0.75) x (0.5) = 0.4386
Where Baseline Emissions = MOVES2010 MY 2003-2006 NOx emission rates for LHD2b3
Then, the T&M effect for 2007-2009 LHD2b3 is calculated in Equation 8-10 by dividing Equation
8-9 by Equation 2-12 and subtracting 1.
276

-------
f	2007 — 2009 LHD2b3 NOx emissions (T&M) ^
ft&m,nox,lhD2b3,2007—2009 - 2 0 0 7 - 2 0 09 LHD2b3 NOx emissions (zero mile) ~ 1 Equation
8-10
/2007 - 2009 LHD2b3 NOx (T&M)\ //2007 - 2009 LHD2b3 NOx (zero mi/e)\
\ Baseline emissions	JI V	Baseline emissions	J
= 0.4386/0.4225 — 1 = 1.038 — 1 = 3.8% increase due to T&M
Where Baseline Emissions = MOVES2010 MY 2003-2006 NOx emission rates for LHD2b3
For 2007-2009, LHD45 uses the same emission rates and T&M factors as LHD2b3.
As noted earlier, we assume no T&M NOx effects for pre-2010 MY vehicles in the other heavy-
duty regulatory classes.
The T&M adjustment factors (fr&M,NOx) for model year 2010 and later model years incorporate the
impact of onboard diagnostic (OBD) emission effect assumptions discussed in Section B.10, and
calculated with Equation 8-15. As explained in that section, for LHD2b3 vehicles, we assume
100% OBD penetration starting in 2010. This reduces the T&M adjustment factor by 0.33 for these
years.
For 2010-2012, LHD2b3, we assume that both LNT and SCR equipped vehicles will provide the
same level of control with a 90 percent reduction from 2003-2006 levels (ignoring the PM
regeneration NOx benefit for LNT aftertreatment considered for the 2007-2009 rates for simplicity).
To calculate the T&M NOx effects for 2010-2012 (Jt&m,N0x,LHD2b3,2010-2012). we weighted the
LNT-specific and SCR-specific T&M effects (from Table B-9) according to the market shares, and
applied the 33% percent reduction for OBD as shown in Equation 8-11:
fT&.M ,LHD2b3,2010—2012	Equation
= (LNT market share) x (/V&m, w ox.lnt .nonOBD*) x (Jobd)	8-11
+ (SCR market share) x ([fr&M.NOx.scR.nonOBD) x (Jobd)
= (25%) x (71.5%) x (67%) + (75%) x (87.4%) x (67%) = 55.9%
For LHD45 and heavier regulatory classes, we assume a 33 percent OBD penetration in model year
2010-2012 as shown in Equation 8-12.
fT&.M,NOx,LHD4S,2010—2012
= (fT&M,NOx,SCR,nonOBD) X (/oBd) X (POBd)
+ (fT8iM,NOx,SCR,nonOBD) X (1 — POBd)
= (87.4%) X (67%) X (33%) + (87.4%) X (67%) = 77.9%
For 2013 and later model years, the T&M adjustment factors are calculated for heavy-duty vehicles
assuming that all (with the exception of gliders) are using SCR technology and 100 percent OBD.
277

-------
fT&M,NOx,SCR,2013-2026 — yfT&M,NOx,SCR,nonOBDj X C/oBd) X (POBd)
= (87.4%) x (67%) x 100% = 58.6%
fr&M.NOx.SCR,2027-2028,HHD = (fr&M.NOx.SCR.nonOBD) X (/ofio) X (Pofifl)
= (612.7%) x (67%) x 100% = 410.5%
The NOx Tampering & Mai-maintenance adjustment factors by regulatory class and model year
groups are summarized in Table B-10.
Table B-10. NOx T&M Adjustment Factors (t'r&M,\ox) by MOVES Regulatory Classes and Model Year Groups
Model Year Group
LH2b3
LHD45
MHD, HHD, Bus
Gliders
(RegClass 41)
(RegClassID 42)
(RegClassID 46,47,48)
(RegClassID 49)
2007-2009
3.81%
3.81%
0%
0%
2010-2012
55.9%
77.9%
77.9%
0%
2013-2027
58.6%
58.6%
58.6%
0%
2027-2028
Same as 2013-
2027
329.7%
324.0% (MHD); 410.5%
(HHD); 433.7% (Bus)
0%
2029+
Same as 2013-
2027
331.6%
325.6% (MHD); 416.3%
(HHD); 440.6% (Bus)
0%
B. 8 PM T&M Emission Effects
EPA developed the PM emission effects for each tampering and mal-maintenance incident from
CARB's EMFAC, Radian's dynamometer testing with and without the malfunction present, EFEE
results, and internal testing experience.
EPA estimates that the diesel PM filter has 95 percent effectiveness. Many of the tampering and
mal-maintenance items that impact PM also have a fuel efficiency and drivability impact.
Therefore, operators will have an incentive to fix these issues.
EPA estimated that excessive oil consumption will have the same level of impact on PM as engine
mechanical failure. The failure of the oxidation catalyst is expected to cause a PM increase of 30
percent; however, this value is reduced by 95 percent due to the PM filter effectiveness. We also
considered a DOC failure will cause a secondary failure of PM filter regeneration. We accounted
for this PM increase within the PM filter disabled and leak categories.
The values with 0 percent effect in shaded cells represent areas which have no occurrence rate.
In MOVES2014, we increased the PM emission effect for PM Filter Leaks and PM Filter
Tampering for the 2007-2009 and 2010+ model year groups. The PM filter leak was increased from
600 percent to 935 percent and the PM Filter Disabled emission effect was increased from 1000
percent to 2670 percent. These in Table B-9 effects along with the OBD effects discussed in
Equation
8-13
Equation
8-14
278

-------
Section B.10 results in a fleet average PM2.5 Tampering & Mal-maintenance effect of 100 percent
in 2007-2009 and 89 percent in 2010-2012 (Table 2-25).
Table B-ll. PM2.5 T&M Emission Effect by Model Year Group

1994-
1998-
2003-
2007-


1997
2002
2006
2009
2010+
Federal Emission Standard
0.1
0.1
0.1
0.01
0.01






Timing Advanced
-10%
-10%
-10%
0%
0%
Timing Retarded
25%
25%
25%
1%
1%
Injector Problem
100%
100%
100%
5%
5%
Puff Limiter Mis-set
20%
0%
0%
0%
0%
Puff Limiter Dsabled
50%
0%
0%
0%
0%
Max Fuel High
20%
0%
0%
0%
0%
Clogged Air Filter
50%
50%
30%
2%
2%
Wrong/W orn Turbo
50%
50%
50%
3%
3%
Intercooler Clogged
50%
50%
30%
2%
2%
Other Air Problem
40%
40%
30%
2%
2%
Engine Mechanical Failure
500%
500%
500%
25%
25%
Excessive Oil Consumption
500%
500%
500%
25%
25%
Electronics Failed
60%
60%
60%
3%
3%
Electronics Tampered
50%
50%
50%
3%
3%
EGR Stuck Open
0%
0%
100%
5%
5%
EGR Disabled/Low Flow
0%
0%
-30%
-30%
-30%
NOx Aftertreatment Sensor
0%
0%
0%
0%
0%
Replacement NOx Aftertreatment
Sensor
0%
0%
0%
0%
0%
NOx Aftertreatment Malfunction
0%
0%
0%
0%
0%
PM Filter Leak
0%
0%
0%
935%
935%
PM Filter Disabled
0%
0%
0%
2670%
2670%
Oxidation Catalyst
Malfunction/Remove
0%
0%
0%
0%
0%
Mis-fuel - EPA
30%
30%
30%
100%
100%
B.9 THC and CO T&M Emission Effects
EPA estimated oxidation catalysts are 80 percent effective at reducing hydrocarbons. All
manufacturers will utilize oxidation catalysts in 2007, but only a negligible number were installed
prior to the PM regulation reduction in 2007. We assumed that with Tampering and Mal-
maintenance, the THC zero level emissions will increase by 50 percent. This still represents a 70
percent reduction in THC emissions between zero-mile 2006 emissions and fully deteriorated 2007
vehicles.
We reduced CARB's THC emission effect for timing advancement because earlier timing should
reduce THC, not increase them. The effect of injector problems was reduced to 1000 percent based
on EPA's engineering staff experience. We increased the THC emission effect of high fuel pressure
(labeled as Max Fuel High) to 10 percent in 1994-1997 years because the higher pressure will lead
to extra fuel in early model years and therefore increased THC. Lastly, we used the THC emission
effect of advanced timing for the electronics tampering (0 percent) for all model years. The values
with 0 percent effect in shaded cells represent areas which have no occurrence rate.
279

-------
Table B-12. THC T&M Emission Effect by Model Year Group
Model Year
1994-1997
1998-2002
2003-2006
2007-2009
2010+
Federal HC Emission Standard (g/bhp-hr)
1.3
1.3
1.3
0.2
0.14
Timing Advanced
0%
0%
0%
0%
0.0%
Timing Retarded
50%
50%
50%
50%
10.0%
Injector Problem (all)
1000%
1000%
1000%
1000%
200.0%
Puff Limiter Mis-set
0%
0%
0%
0%
0.0%
Puff Limited Disabled
0%
0%
0%
0%
0.0%
Max Fuel High
10%
0%
0%
0%
0.0%
Clogged Air Filter - EPA
0%
0%
0%
0%
0.0%
Wrong/Worn Turbo
0%
0%
0%
0%
0.0%
Intercooler Clogged
0%
0%
0%
0%
0.0%
Other Air Problem - EPA
0%
0%
0%
0%
0.0%
Engine Mechanical Failure
500%
500%
500%
500%
100.0%
Excessive Oil Consumption
300%
300%
300%
300%
60.0%
Electronics Failed - EPA
50%
30%
50%
50%
10.0%
Electronics Tampered
0%
0%
0%
0%
0.0%
EGR Stuck Open
0%
0%
100%
100%
20.0%
EGR Disabled/Low-Flow - EPA
0%
0%
0%
0%
0.0%
NOx Aftertreatment Sensor
0%
0%
0%
0%
0.0%
Replacement NOx Aftertreatment Sensor
0%
0%
0%
0%
0.0%
NOx Aftertreatment Malfunction - EPA
0%
0%
0%
0%
0.0%
PM Filter Leak
0%
0%
0%
0%
0.0%
PM Filter Disabled
0%
0%
0%
0%
0.0%
Oxidation Catalyst Malfunction/Remove - EPA
0%
0%
0%
50%
50.0%
Mis-fuel - EPA
0%
0%
0%
0%
0%
A separate tampering analysis was not performed for CO; rather, the THC effects were assumed to
apply for CO.
Combining all of the emissions effects and failure frequencies discussed in this section, and the
OBD effects discussed in the next section, we summarized the aggregate emissions impacts over
the useful life of the fleet in the main body of the document in Table 2-28 (THC and CO).
B.10 HD OBD impacts
With the finalization of the heavy-duty onboard diagnostics (HD OBD) rule, we made adjustments
to 2010 and later model years to reflect the rule's implementation.
280

-------
Specifically, we reduced the emissions increases for all pollutants due to an OBD tampering and
mal-maintenance factor, f0BD, which reduced the T&M adjustment factors by 33 percent. Data on
the impact of OBD were not available for heavy-duty trucks, and this number is probably a
conservative estimate. This is in addition to the substantial PM2.5 and NOx reductions for 2010 and
later vehicles due to the implementation of other standards. We assumed, since the rule phased-in
OBD implementation, that 33 percent of all LHD45, MHD, HHD and Urban Bus engines would
have OBD in the 2010, 2011, and 2012 model years, and 100 percent would have OBD by 2013
model year and later. For LHD2b3 vehicles, we assumed they would have 100% OBD penetration
starting in 2010. Equation 8-15 describes the calculation of the percent increase in emission rate
through useful life (T&M adjustment factors (fr&ivOX where poso represents the fraction of the fleet
equipped with OBD (Table B-l 1).).
^T&M.p = fT&M,nonOBD,p X (1 — Pobd)	„
K	f	f	( n	Equation 8-15
+ fT&M,nonOBD,p X *OBD X (PoBdJ
Where:
fi&M,P= the tampering and mal-maintenance adjustment factor for pollutant, p, that accounts for the phase-in
of OBD
fT&M,nonOBD,P= the tampering and mal-maintenance adjustment factor for pollutant, p, for engines without
OBD; calculated in Sections B.7 through B.9
Pobd= penetration of the fleet equipped with OBD, as shown in Table B-ll.
fobd= the effect of OBD on the T&M adjustment factor = 0.67= 33% reduction
Table B-13. Onboard Diagnostic (OBD) Assumed Phase-in (pobd) by Model
Year and Regulatory Class
Model years
Regulatory Class
Pobd (%)
Pre-2010
LHD2b3, LHD45, MHD, HHD, Urban Bus
0
2010-2012
LHD2b3
100
2010-2012
LHD45, MHD, HHD, Urban Bus
33
2013+
LHD2b3, LHD45, MHD, HHD, Urban Bus
100
281

-------
Appendix C Tampering and Mal-maintenance for MY 2007 and
Later Diesel Extended Idle
As discussed in Section 2.3.2.3 we assume the failure of diesel particulate filters (DPF) is the
primary cause of T&M effects on emission deterioration in 2007+ extended idle emissions. We
made assumptions about the failure rates of DPFs from in-use trucks based on consultation with
several references and staff at the California Air Resources Board (CARB) as summarized in Table
C-l. We adopted the assumption shared by CARB staff that 10 percent of 2007-2009 DPFs fail in
the real-world, and 5 percent of 2010+ DPFs fail in the real-world.
Table C-l. References Used to Support In-Use DPF Failure Rate Assumption for Extended Idling Emissions
Study
Relevant Information
US EPA (2015)109
7% of 2007+ trucks in MOVES are assumed to either have a PM filter leak or
have the PM filter disabled. Current assumption for running exhaust emissions in
MOVES3.
Preble et al.
(2015)186
20% of trucks produce 80% of black carbon (BC) emissions from Port of Oakland
2013 truck fleet, where 99% of the trucks are equipped with DPFs
Bishop et al.
(2014)187
3% of 2007+ trucks at Port of LA have PM emissions 3 x the standard. 9% of
2008+ trucks at Cottonwood site have PM emissions 3/ the standard
CARB (2015)188
35% to 4% of trucks submitted warranty claims related to the PM filter between
2007 and 2011
CARB (2015)188
8% of trucks were classified as high emitters (emitting over 5% opacity) from a
sample of >1,800 trucks test in the snap-idle acceleration test by CARB, about
-1/2 equipped with DPFs
CARB
correspondence
(2016)
-10% of 2007-2009 DPFs and -5% of 2010+ DPFs to fail in real-world, based on
their observations from warranty claims, snap-idle acceleration opacity tests, and
their review of the Bishop et al. (2014)187 and Preble et al. (2015)186 studies.
To account for the failure of DPF in the THC and PM2.5 emission rates, we used the 2005-2006
average extended idle emission rates to represent the 'failed' DPF emission rates. We then
calculated a 'Deteriorated' emission rate that represents a mix of failed and properly operating
systems by assigning the 'failed' DPF emission rates a weight of 10 percent in the 2007-2009
model year group, and 5 percent weight in the 2010-2012, and 2013+ model year groups, as shown
in Table C-2. The 'Deteriorated' emission rate represents the presumed emission rate of fully-aged
heavy-duty diesel trucks. Unlike the start and running MOVES emission rates, extended idle
emission rates in MOVES are not distinguished by age. Thus, these rates are constant with respect
to age.
282

-------
Table C-2. Baseline and deteriorated THC and PM2.5 emission rates to account for failure of diesel particulate
Engine
Model
Year
Baseline
Deteriorated
THC
(g/hr)
pm25
(g/hr)
EC
(g/hr)
nonEC
(g/hr)
Failure
rate
THC
(g/hr)
pm25
(g/hr)
EC
(g/hr)
nonEC
(g/hr)
2005-
2006
8.49
0.251
0.065
0.187

8.49
0.251
0.065
0.187
2007-
2009
8.49
0.075
0.007
0.067
10%
8.49
0.092
0.013
0.079
2010-
2012
2.53
0.026
0.004
0.022
5%
2.83
0.037
0.007
0.030
2013+
1.38
0.012
0.002
0.010
5%
1.74
0.024
0.005
0.019
We assume that trucks that are under warranty would have substantially fewer aftertreatment
failures than older trucks. Because extended idle rates are modelled as constant with age, to
estimate the fleet-average emission rates used in MOVES, we used the 'Baseline' emission rates to
represent trucks that are within the specified 435,000 miles useful-life of the engine in the US EPA
regulations. We use the deteriorated emission rate to represent the years between the regulated
"useful life" and the 1,530,000 miles that MOVES models as the mean life-time miles for a long-
haul combination trucks. Using the 'deterioration fraction' [(l-.435)/l.53 = 0.72] as the fraction of
the vehicle miles traveled during the deterioration phase, we calculated fleet-average emission rates
used for MOVES in Table C-3. As shown, the MOVES EC/PM emission rates for MY 2007+
trucks are slightly higher than the 'Baseline' EC/PM fractions in Table C-2, because the fleet
emissions are assumed to include some contribution of emissions from trucks with failed DPFs,
which have a higher EC/PM fraction.
Table C-3. Emission Rates Calculated from Weighting the 'Baseline' and 'Deteriorated' Emission Rates from
Engine
Model
Year
MOVES
Deteriorated
Fraction
THC
(g/hr)
PM2.5
(g/hr)
EC
(g/hr)
nonEC
(g/hr)
EC/PM
2005-2006
-
8.49
0.251
0.065
0.187
0.26
2007-2009
0.72
8.49
0.087
0.012
0.076
0.13
2010-2012
0.72
2.75
0.034
0.006
0.028
0.18
2013+
0.72
1.64
0.021
0.004
0.017
0.20
Although, 2005-2006 model year engine data was used in this analysis, the update itself is limited
to the model year 2007 and later emission rates.
283

-------
Appendix D Pre-2007 Model Year Extended Idle Data Summary
These tables provides additional information on the data used to estimate extended idle emissions
for pre-2007 MY vehicles as described in Section 2.3.1.
Table D-l. Data for Pre-2007 Extended Idle NOx Emissions
Idle NOx Rates (gram/hour) Summary
Program
Condition
#Samples
Mean NOX Emiss Rate
1991-2006 Low Speed Idle, A/C Off
McCormick, High Altitude, HDT
Low RPM, AC Off
12
85
Lim, EPA
Low RPM, No access
12
109
Irick, Clean AirTech & IdleAire

49
87
WVU - 1991-2004
Low RPM, AC Off
48
83
WVU, NCHRP

2
47
Tang, Metro NY 1984-1999

33
81
Calcagno
Low RPM, AC Off
27
120
Brodrick, UC Davis
Low RPM, AC Off
1
104
Storey
Low RPM, AC Off
4
126

Overall
188
91


1991-2006 High Speed Idle, A/C Off
Lim, EPA CCD
High RPM, No access
5
169
Calcagno
High RPM, AC Off
21
164

Overall
26
165


1991-2006 High Speed Idle, A/C On
Lim, EPA CCD
High RPM, AC On
5
212
Brodrick, UC Davis
High RPM, AC On
1
240
Calcagno
High RPM, AC On
21
223
Storey
High RPM, AC On
4
262

Overall
31
227


1975-1990 Low Speed Idle, A/C Off
WVU - 1975-1990
Low RPM, AC Off
18
48
Lim, EPA, CCD, 1985 MY
Low RPM, AC Off
1
20

Overall
19
47
1975-1990 High Speed Idle, A/C On (calculated)
Ratio of 1991-2006 "High Idle, A/C On" to "Low Idle, A/C Off"
2.5
Overall (calculated)
115.4
Calculated Extended Idle MYs 1975-1990:
69.3
Calculated Extended Idle MYs 1991-2006:
136.1
284

-------
Table D-2. Data for Pre-2007 Extended Idle HC Emissions
Idle HC Rates (gram/hour) Summary
Program
Condition
#Samples
Mean HC Emiss Rate
1991-2006 Low Speed Idle, A/C Off
McCormick, High Altitude, HDT
Low Idle, AC Off
12
10.2
WVU - 1991-2004
Low Idle, AC Off
48
9.5
Storey
Low Idle, AC Off
4
28

Overall
64
10.8

1991-2006 High Speed Idle, A/C On
Brodrick, UC Davis
High Idle, AC On
1
86
Storey
High Idle, AC On
4
48

Overall
5
55.6


1975-1990 Low Speed Idle, A/C Off
WVU - 1975-1990
Low Idle, AC Off
18
21

Overall
18
21
1975-1990 High Speed Idle, A/C On (calculated)
Ratio of 1991-2006 "High Idle, A/C On" to "Low Idle, A/C Off"
5.2
Overall (calculated)
108.2
Calculated Extended Idle MYs 1975-1990:
49.8
Calculated Extended Idle MYs 1991-2006:
25.6
285

-------
Table D-3. Data for Pre-2007 Extended Idle CO Emissions
Idle CO Rates (gram/hour) Summary
Program
Condition
#Samples
Mean CO Emiss Rate
1991-2006 Low Speed Idle, A/C Off
McCormick, High Altitude, HDT
Low Idle, AC Off
12
71
Calcagno
Low Idle, AC Off
27
37
WVU - 1991-2004
Low Idle, AC Off
48
23
Storey
Low Idle, AC Off
4
25

Overall
91
33.6




1991-2006 High Speed Idle, A/C On
Calcagno
High Idle, AC On
21
99
Brodrick, UC Davis
High Idle, AC On
1
190
Storey
High Idle, AC On
4
73

Overall
26
98.5




1975-1990 Low Speed Idle, A/C Off
WVU - 1975-1990
Low Idle, AC Off
18
31

Overall
18
31
1975-1990 High Speed Idle, A/C On (calculated)
Ratio of 1991-2006 "High Idle, A/C On" to "Low Idle, A/C Off"
2.9
Overall (calculated)
91.0
Calculated Extended Idle MYs 1975-1990:
50.8
Calculated Extended Idle MYs 1991-2006:
55.0
286

-------
Table D-4. Data for Pre-2007 Extended Idle PM Emissions
Idle PM Rates (gram/hour) Summary
Program
Condition
#Samples
Mean PM Emiss Rate
1991-2006 Low Speed Idle, A/C Off
McCormick, High Altitude, HDT
Low Idle, AC Off
12
1.8
Calcagno
Low Idle, AC Off
27
2.55
WVU - 1991-2004
Low Idle, AC Off
48
1.4
Storey
Low Idle, AC Off
4
0.3

Overall
91
1.7




1991-2006 High Speed Idle, A/C On
Calcagno
High Idle, AC On
21
4.11
Storey
High Idle, AC On
4
3.2

Overall
25
4.0




1975-1990 Low Speed Idle, A/C Off
WVU - 1975-1990
Low Idle, AC Off
18
3.8

Overall
18
3.8
1975-1990 High Speed Idle, A/C On (calculated)
Ratio of 1991-2006 "High Idle, A/C On" to "Low Idle, A/C Off"
2.3
Overall (calculated)
8.6
Calculated Extended Idle MYs 1975-1990:
5.4
Calculated Extended Idle MYs 1991-2006:
2.5
287

-------
Appendix E Developing Pre-2007 Model Year HD Diesel PM2.5
Emission Rates for Missing Operating Modes
As noted in Section 2.1.2.1 , in cases where an estimated operating mode PM2.5 rate for pre-2007
MY HD diesel trucks could not be directly calculated from data, we imputed the missing value
using a log-linear least-squares regression procedure. Regulatory class, model year group and speed
class (0-25 mph, 25-50 mph and 50+ mph) were represented by dummy variables in the regression.
The natural logarithm of emissions was regressed versus scaled tractive power (STP) to represent
the operating mode bins. The regression assumed a constant slope versus STP for each regulatory
class. Logarithmic transformation factors (mean square error of the regression squared / 2) were
used to transform the regression results from a log based form to a linear form. Due to the huge
number of individual second-by-second data points, all of the regression relationships were
statistically significant at a high level (99 percent confident level). The table below shows the
regression statistics, and the equation shows the form of the resulting regression equation.
Table E-l. Regression Coefficients for HD Diesel Pre-2007 PM2.5 Emission Factor Model
Model-year
Speed Class (mph)
Type
Medium
Heavy Heavy-
group


Heavy-Duty
Duty
1960-87
1-25
Intercept (//,)
-5.419
-5.143

25-50

-4.942
-4.564

50+

-4.765
-4.678
1988-90
1-25

-5.366
-5.847

25-50

-4.929
-5.287

50+

-4.785
-5.480
1991-93
1-25

-5.936
-5.494

25-50

-5.504
-5.269

50+

-5.574
-5.133
1994-97
1-25

-5.927
-6.242

25-50

-5.708
-5.923

50+

-5.933
-6.368
1998-2006
1-25

-6.608
-6.067

25-50

-6.369
-5.754

50+

-6.305
-6.154

STP
Slope (/?i)
0.02821
0.0968


Transformation




Coefficient
0.5864
0.84035


(0.5c2)


ln(PM) = p0+ PiSTP + 0.5a2
Where :
/?o = an intercept term for a speed class within a model year group, as shown in the table above,
[i\ = a slope term for STP, and
o2 = the mean-square error or residual error for the model fit,
STP = the midpoint value for each operating mode (kW/metric ton, see Table 1-4).
288

-------
Appendix F Heavy-Duty Gasoline Start Emissions Analysis
Figures
The figures below show heavy-duty gasoline start emissions as mentioned in Section 3.2.1,
FTP Cold-Starts (g). HD SI (hD< = UK)
CO slorls vs. Age by MYG
node I year group OOP 19601989 O ~ O |
1991199?	19992004
FTP Cold-Starts (g), HD SI (|-D< = 14K)
THC starts vs. Age by I1YG
(b) THC



II	12	13
I9B0I989 EJ-B-B 19901990 ' ' ' 1991199/ A-A-ir 19982004
FTP Cold—Starts (g). HD SI (FO< = 14K)
HOx starts vs. Age by KTG
lycaroroup S-©-© 19601989 B-e-e 19901990 "I • •" 19911997 ii-ir-c 19982004
Figure F-l Cold-Start FTP Emissions for Heavy-Duty Gasoline Vehicles, Averaged by Model-year and Age
Groups
289

-------
FTP Cold-Starts (g), HD SI (HD< = 14K)
CO GEO-meon storls vs. Age by MYG
FTP Cold-Starts (g). HD SI (HD< = 14K)
NOx GEO-meon storls vs. Age by MYG
FTP Cold-Starts (g). HD SI (HD< = 14K)
THC GEO-meon storls vs. Age by MYG
Figure F-2 Cold-Start FTP Emissions for Heavy-Duty Gasoline Vehicles, GEOMETRIC MEANS by Model-year
and Age Groups
290

-------
CO I n_SD vs. Age by MYG
node 1 yeargroup ooe 19601983 B- P'O 19901990 H—I—H 19911997 A & & 19902004
FTP Cold—Starts (g), HD SI (l-D< = I4»g
THC In_SD vs. Age by MYG
node 1 yeargroup S-©-© 19601989 B-B-B 19901990 H—I—t- 19311997	1 9982004
FTP Cold-Starts (g). HD SI (HD< = 14K)
NOx I n_SD vs. Age by WfG
mode 1 yeargroup O O O 19601983 POO 19301330 ¦ ' ¦ 13311337 a & & 13382004
Figure F-3 Cold-start FTP Emissions for Heavy-Duty Gasoline Trucks: LOGARITHMIC STANDARD
DEVIATION by Model-year and Age Groups
291

-------
CO ARITH-mean starts »s. Age by MYC


ca
(a
)CO
!\ ! ! 1


. \ :


j j V ! ! j i |
:\
1 1 : \ : : j I i
f


2	3	4	5	6	7	8	S	10	11	12	13
modelyeargroup O O O 19601389 ODD 19901990 I ¦ I 19911997 AAA 19982004
FTP Cold-Starts (g), HD SI (I-©< = 14K)
THC ARITH-mean starts vs. Age by MYG
modelyeargroup O P O 19601989 D ~ D 19901390 I ' I 19911337 AAA 13382004
FTP Cold-Starts (g). HD SI (hC< = 14K)
NOx ARITH-mean starts vs. Age by MYG
agenid
model yeargroup OOP 19601989 B—B-B 19901990 I I I 19311337 A-ts-A 19982004
Figure F-4 Cold-Start Emissions for Heavy-Duty Gasoline Trucks: RECALCULATED ARITHMETIC MEANS
by Model-year and Age Groups
292

-------
Table F-l Emission Standards for Heavy-Duty Spark-Ignition Onroad Engines
Regulatory Class
Model Year
Emissions Standards (g/hp-hr)


CO
THC
NMHC
NOx
NMHC + NOx
LHD2b3
1990
14.4
1.1

6.0


1991-1997
14.4
1.1

5.0


1998-2004
14.4
1.1

4.0


2005-2007
14.4



1.0

2008+
14.4

0.14
0.20

LHD45, MHD
1990
37.1
1.9

6.0


1991-1997
37.1
1.9

5.0


1998-2004
37.1
1.9

4.0


2005-2007
37.1



1.0

2008+
14.4

0.14
0.20

293

-------
Appendix G Selection of Fixed Mass Factor {fscale) values for MY
2010+ Heavy-Duty Vehicles
In MOVES3, for model year 2010 and newer heavy-duty diesel, gasoline, and CNG running-
exhaust emissions operating mode based rates, we discarded the fixed mass factor (fSCaie) value of
17.1 metric tons used in previous MOVES versions, and used the manufacturer-run HDIUT data to
estimate new fscale values for LHD, MHD, and HHD weight classes. New fscaie values were needed
because the 17.1 value was too large, limiting emission rate data to low and medium power
operating modes, and requiring gap-filling for high-power operating modes.
The new fscaie values for MY 2010+ vehicles are 5.00, 7.00, and 10.00 metric tons for LHD
(regClass 41 and 42), MHD (regClass 46), and HHD (regClass 47 and 48), respectively. These fscaie
values are used when analyzing the real-world emissions data that leads to the base emission rates
in the MOVES database. The fscaie values are also used by the MOVES model, at run time, to
convert vehicle activity to operating mode-based time distributions. For consistency, operating
mode-based emissions rates and time distributions, for a given regulatory class and model year,
must be based on the same fscaie value. Glider vehicles (regClass 49) continue to use emission rates
from pre-2010 vehicles, and thus their fcaie value is unchanged at 17.1. Note that it is not
meaningful to compare operating mode based rates based on different fscaie values. This appendix
describes how we arrived at the fscaie values.
The entire MY 2010+ HDIUT dataset (Section 2.1.1.1) was analyzed using a range offscaie values.
For this exercise we analyzed LHD, MHD, and HHD separately, but within those regulatory
classes, we did not divide the data set by NOx FEL or model years. We included the MOVES2014
value (fscaie = 17.1) to show how the operating mode distribution would look for a "business as
usual" case. It is expected that the fscaie for LHD should be lower than MHD, which in turn should
be lower than HHD. Our goal was to find fscaie values that allow the HDIUT data to cover all
operating modes, thus reducing the need for gap-filling while also leaving the highest power
operating modes (30 and 40) as not saturated because the HDIUT data is not expected to have very
aggressive operation. If the fscaie is too big, the high power operating modes are left vacant. On the
other hand, if the fscaie is too small, a lot of the data gets pushed to the high power operating modes,
and the high operating modes no longer capture emissions at only the most extreme and aggressive
operating conditions measured in the HDIUT data set. Both cases are sub-optimal because they
reduce the model's capability to distinguish operating modes in a meaningful way.
When analyzing the HDIUT data for various fscaie values, we estimated the number of vehicles,
time, and mass/time emission rates for each operating mode. Vehicle count and time, per operating
mode, were first cut criteria during the fscaie selection process. We used the CO2 mass/time rates as
an additional check because these rates are known to have consistent and predictable monotonically
rising trend within each speed-bin (since higher power demand requires burning more fuel which
leads to more CO2).Table G-l through Table G-3 show how the choice offscaie values would affect
the vehicle count and seconds in each operating mode, for LHD, MHD, and HHD, respectively.
The number of seconds is based on the HDIUT-based operating mode time fractions applied to a
cycle of one million seconds. Using a unique but representative fscaie for each regClass, when
combined with a cycle of the same number of total seconds, should result in similar number of
seconds in high power operating modes. In other words, we expect LHD, MHD, and HHD vehicles
294

-------
in the HDIUT data set to have somewhat similar time distribution across power modes. Finally, we
used the CO2 mass/time rate trends as an additional metric to pick a final fscale between candidate
values that look reasonably good for both vehicle count and time distribution.
Looking at Table G-l for LHD vehicles, fscale = 2.06 results in every one of the 64 vehicles having
operation in operating modes 30 and 40 and significantly more seconds of data than operating
modes 29 and 39, respectively. On the other hand, a fscale value of 9.00 or 17.1 meant the high
power operating modes had only a couple vehicles and seconds, which is a sign of under-
representation in those operating modes. Thus, a suitable fscale value, for LHD, should be between
2.06 and 9.00. Based on further analysis, the final fscale candidates for LHD were 4.00, 5.00, and
6.00. A value of 4.00 seemed too small because we did not expect over 40 (out of 64) vehicles to
have operation in operating modes 30 and 40. A value of 6.00 seemed too high because it led to
only 40 seconds and 135 seconds of data (from a cycle with a million seconds) in operating modes
30 and 40, respectively. The small sample size was deemed insufficient to determine robuts
emission rates for these operating mode bins. We picked 5.00 as the final fscale value for LHD
because it resulted in a reasonable number of vehicles and seconds in the high power operating
modes 29, 30, 39, and 40. For confirmation purpose, we also compared the CO2 mass/time rates for
all thefscale values considered during the analysis and Figure G-l shows a comparison between the
final candidates of 4.00, 5.00, and 6.00. As seen in the figure, all three values provide good
monotonically increasing trend, fscale 5.00 yields much more aggressive driving behavior in
operating modes 30 and 40 compared to fscale 4.00, with mean CO2 emissions rates approximately
30% higher in these operating modes. As stated earlier, our objective is to select the fscale that yields
the most aggressive operation in the highest operating modes while still providing sufficiently
robust estimates of the emission rates. Because fscale 5.00 still provides mean CO2 emissions rates
that are robust for the high operating mode bins, fscale 5.00 is preferable to an fscale of 4,00, /«„/e
5.00 provides strong and expected increasing trends in CO2 emission rates between operating
modes 29 and 30, whereas fscale 6.00 does not. We believe this is due to insufficient data in the
highest operating mode bins with fscaie of 6.00 to determine robust estimates at the highest operating
modes. Thus, we decided to use an fscaie of 5.00 for LHD in MOVES.
For MHD and HHD, we went through similar reasoning and steps as for LHD. Our final fscaie
values for LHD, MHD, and HHD are 5.00, 7.00, and 10.00, respectively. From Table G-l - Table
G-3, these fscaie values lead to comparable vehicle count (20-40 % of total vehicles in the regulatory
class) and seconds of data (1000-3000 seconds out of one million) in OpModes 30 and 40.
We did not try to find a precise and even more suitable fscaie value. Thus, for example, whether fcaie
of 4.80 or 5.20 is better than 5.00, for LHD, was not tested. There are diminishing returns for the
extra time and effort required for that analysis because: (1) the HDIUT data set lacks certain things
such as very aggressive operation or malfunctioning vehicles, so a very suitable value offscaie from
this data set might not be as suitable with another data set; (2) comparing closely spaced fscaie
values does not necessarily provide a clear winner across the board because there's more than one
criteria (vehicle count, time, mass/time rates for various pollutants).
In the 2019 peer-review, one of peer-reviewers asked if we have evaluated the time distribution
from real-world data using the proposed fscaie value. In Figure G-4 through Figure G-6. we
conducted a comparison of the operating mode distributions measured from the HDIUT dataset and
295

-------
real-world operating modes estimated from MOVES3 national scale runs for the three evaluated
regulatory classes LHD, MHD, and HHD at the proposed proposed /scale values. As shown, the
distributions are similar between the HDIUT dataset and national MOVES runs, with most of the
data occurring at idle and the operating modes above 50 mph (opModelD 33-40). One notable
difference is for a national scale run MOVES estimates a higher percentage of activity in the
highest power, high speed operating mode bins. This is expected, given that the HDIUT dataset is
expected to under-represent high power operation due to steep grades, high speeds, and heavy-pay
loads (e.g. multiple trailers, over-weight trailers) compared to the in-use fleet. This comparison
supports our logic to select the fscale that maps only the most aggressive operation from the HDIUT
dataset into the highest MOVES operating mode bins.
The peer-reviewer suggested that we compare the operating mode distribution obtained from the
proposed fscale values from other in-use datasets. We agree that this would be useful to better
understand the representativeness of the HDIUT dataset, as well as further evaluate the MOVES
default activity assumptions, including the MOVES heavy-duty driving cycles. In the population
and activity report, we listed this as a project for consideration for future MOVES work.
296

-------
Table C.-l Effect of fscaie Value on Vehicle Count and Time for Light Heavy-Duty Vehicles

Number of vehicles1
Number of seconds based on a cycle with one million
seconds1'2
OpMode
fs
fs
fs
fs
fs
fs
fs
fs
fs
fs
fs
fs
2.06
4.0
5.0
6.0
9.0
17.1
2.06
4.0
5.0
6.0
9.0
17.1
0
64
64
64
64
64
64
41131
41131
41131
41131
41131
41131
1
64
64
64
64
64
64
358957
358957
358957
358957
358957
358957
11
64
64
64
64
64
64
46235
46235
46235
46235
46235
46235
12
64
64
64
64
64
64
15112
24896
29023
32878
42842
61989
13
64
64
64
64
64
64
10311
14877
16658
18290
20756
16153
14
64
64
64
64
64
43
8109
11394
12304
12431
10835
2387
15
64
64
64
64
63
18
6995
8784
8680
8092
4190
200
16
64
64
64
64
41
1
40203
20778
14065
9039
2107
1
21
64
64
64
64
64
64
45157
45157
45157
45157
45157
45157
22
59
64
64
64
64
64
7153
16366
22038
28166
47554
92823
23
64
64
64
64
64
64
9877
24704
31884
38195
49094
41467
24
64
64
64
64
64
48
12315
25290
28989
30288
26962
6425
25
64
64
64
64
64
22
13307
21400
21274
19931
11739
850
27
64
64
64
64
44
1
25844
28819
25061
18770
5^71
1
28
64
64
64
44
"> ">
">
21514
15281
8422
5020
544
5
29
64
64
42
23
0
0
16292
5808
2900
1161
i)
0
30
64
43
22
13
">
1
35269
3903
1002
40
5
i
33
64
64
64
64
64
64
26999
42467
55797
73408
142294
267976
35
64
64
64
64
64
46
16820
75707
110862
131366
129751
18220
37
64
64
64
63
44
5
33332
86600
79892
67271
13 1 77
7
38
64
63
62
44

i
46641
52178
31818
12100
^7X

39
64
62
41

1
i
49374
21416
6167
1940
1

40
64
41
23
14
1
i
113054
7852
1685
135
IS
13
Notes:
1	Values in bold are for final selected fscau. Shaded cells show instances where using an excessively high fscaie value
causes data deficit in the higher power operating modes within a speed bin.
2	Number of seconds = Average operating mode time fraction * cycle with one million seconds. The average operating
mode time fraction is the average of the time fraction (for that operating mode) across all vehicles.
297

-------
60
70
60
50
40
8 30
20
10
0
H fs = 4.00
¦ fs = 5.00
0 fs = 6.00
MIH
BIB llB ll
0 1 i 11 12 13 14 15 16 i 21 22 23 24 25 27 28 29 30 i 33 35 37 38 39 40
MOVES OpMode
Figure G-l Effect of/K* Value on Coverage and Trends of operating mode Based CO2 for Light Heavy-Duty
Vehicles
Absolute values of operating mode based emissions rates cannot be compared between series with
different fscaie values.
298

-------
Table G-2 Effect of fscaie Value on Vehicle Count and Time for Medium Heavy-Duty Vehicles

Number of vehicles1
Number of seconds based on a cycle with one million
seconds1'2
OpMode
fs
fs
fs
fs
fs
fs
fs
fs
fs
fs
fs
fs
4.0
5.0
6.0
7.0
9.0
17.1
4.0
5.0
6.0
7.0
9.0
17.1
0
58
58
58
58
58
58
36170
36170
36170
36170
36170
36170
1
58
58
58
58
58
58
349622
349622
349622
349622
349622
349622
11
58
58
58
58
58
58
32693
32693
32693
32693
32693
32693
12
58
58
58
58
58
58
20755
23825
26630
29346
34264
48508
13
58
58
58
58
58
58
11234
12492
13612
14397
15495
16683
14
58
58
58
58
58
58
8252
9101
9517
9898
10003
7687
15
58
58
58
58
58
30
6688
7014
7142
7000
6576
1914
16
58
58
58
58
58
16
28357
22855
18385
14645
8948
496
21
58
58
58
58
58
58
44291
44291
44291
44291
44291
44291
22
58
58
58
58
58
58
10132
14200
18827
23955
34436
78404
23
58
58
58
58
58
58
19195
25479
31609
37379
48401
49534
24
58
58
58
58
58
58
21109
27513
32401
34297
30771
19215
25
58
58
58
58
58
37
22136
24498
22647
20428
17498
5111
27
58
58
58
58
58
16
32912
28713
25621
22502
17220
1306
28
58
58
58
58
25
i)
18677
15863
13842
10424
4619
i)
29
58
58
46
25
In
i)
12105
10075
5243
3550
^25
i)
30
58
39
25
14
i)
i)
17304
7229
3379
1034
i)
i)
33
58
58
58
58
58
58
37996
45307
54057
64727
92859
212774
35
58
58
58
58
58
58
40077
63649
87949
107234
130599
92395
37
58
58
58
58
57
15
63932
76693
81453
80818
72135
3197
38
58
58
57
49
3d
1
56697
58308
51520
44843
1 1 144
1
39
58
53
43
25
In
1
45255
39067
25466
8307
|fi2X
1
40
53
39
26
14
1
i)
64411
25344
7923
2440

0
Notes:
1	Values in bold are for final selected fscau. Shaded cells show instances where using an excessively high fscaie value
causes data deficit in the higher power operating modes within a speed bin.
2	Number of seconds = Average operating mode time fraction * cycle with one million seconds. The average operating
mode time fraction is the average of the time fraction (for that operating mode) across all vehicles.
299

-------
JEl
Hfs= 5.00
Hfs= 6.00
¦ fs= 7.00
ML
BBL
0 1 i 11 12 13 14 15 16 i 21 22 23 24 25 27 28 29 30 i 33 35 37 38 39 40
MOVES OpMode
Figure G-2 Effect of/scale Value on Coverage and Trends of operating mode Based CO2 for Medium Heavy-Duty
Vehicles
300

-------
Table G-3 Effect of fscaie Value on Vehicle Count and Time for Heavy Heavy-Duty Vehicles

Number of vehicles1
Number of seconds based on a cycle with one million
seconds1,2
OpMod
fs
fs
fs
fs
fs
fs
fs
fs
fs
fs
fs
fs
e
9.0
10.0
11.0
12.0
14.0
17.1
9.0
10.0
11.0
12.0
14.0
17.1
0
159
159
159
159
158
159
18010
18010
18010
18010
18010
18010
1
159
159
159
159
159
159
297662
297662
297662
297662
297662
297662
11
159
159
159
159
158
159
37453
37453
37453
37453
37453
37453
12
159
159
159
159
158
159
24580
25976
27238
28427
30556
33336
13
159
159
159
159
158
159
9472
9547
9576
9576
9551
9524
14
159
159
159
159
158
159
5545
5557
5579
5533
5446
5159
15
159
159
159
159
158
159
3938
3840
3762
3704
3456
2892
16
159
159
159
159
153
152
10041
8657
7422
6337
4568
2666
21
159
159
159
159
158
159
32325
32325
32325
32325
32325
32325
22
159
159
159
159
158
159
12785
14388
15951
17580
20814
25721
23
159
159
159
159
158
159
14276
15457
16513
17395
18748
20117
24
159
159
159
159
158
159
11401
11865
12272
12453
12761
12587
25
159
159
159
159
158
159
8967
9058
9085
9044
8501
8402
27
159
159
159
159
153
154
12410
11927
11767
12026
13569
15804
28
154
154
154
153
134
27
8660
9619
10875
12433
8684
632
29
153
142
122
75
S
0
8905
9822
6562
2329
185
0
30
114
59
10
1
0
0
5861
1127
239
4
0
0
33
159
159
159
159
158
159
114214
126216
139731
154101
1SO094
23~9oO
35
159
159
159
159
158
159
139109
160667
176144
186131
189813
176111
37
159
159
159
159
153
153
115050
102440
91446
83420
74122
61219
38
154
154
153
152
131
26
55279
52010
50633
47483
27291
2421
39
152
138
122
83
<)
0
37885
33033
19207
(i5~<>
^1
0
40
114
65
11
1
0
0
16174
3344
550
0
0
0
Notes:
1	Values in bold are for final selected fscau. Shaded cells show instances where using an excessively high fscaie value
causes data deficit in the higher power operating modes within a speed bin.
2	Number of seconds = Average operating mode time fraction * cycle with one million seconds. The average operating
mode time fraction is the average of the time fraction (for that operating mode) across all vehicles.
301

-------
H fs = 9.00
¦ fs = 10.0
0fs = 11.0
niH HIH
ll
0 1 i 11 12 13 14 15 16 i 21 22 23 24 25 27 28 29 30 i 33 35 37 38 39 40
MOVES OpMode
Figure G-3 Effect of/K* Value on Coverage and Trends of operating mode Based CO2 for Heavy Heavy-Duty
Vehicles
302

-------
HHDD Opmode Distributions
Comparison of HDIUT vs. MOVES-CTI
¦ HDIUT ¦ MOVES CTI - CY2027/MY2014
n oc i i i
0.30
c 0.25
o
4-j
2 0.20
LI-
CU
"g 0.15
E
Q.
o 0.10
0.05
... Il





















ll

1
-1
1

m






.1

II il .. ..
Il ¦¦ ¦¦ ¦¦ ¦¦ ¦¦ ¦. ¦ _





1 il I
u.uu
0 1
11 12 13 14 15 16
21 22 23 24 25 27 28 29 30
Opmode

33 35
37 38 39 40
Figure G-4. Heavy Heavy-Duty Diesel Operating Mode Distribution compared between the Heavy-Duty In-Use Testing (HDIUT) Program and from
MOVES3 for a MY 2014 vehicle with an fscaie of 10 metric tons

-------
0.45
0.40
0.35
c
.2 0.30
fo,s
CL>
"E °-20
E
Q- 0.15
O
0.10
0.05
0.00
MHDD Opmode Distributions
Comparison of HDIUT vs. MOVES-CTI
¦ HDIUT ¦ MOVES CTI - CY2027/MY2014
i
1
	L




i
i









i




	i




1
	j_




i









I.
II

il 11 ¦¦ ¦¦ __ ¦¦
li ii ¦¦ .. _

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 G-5. Medium Heavy-Duty Diesel Operating Mode Distribution compared between the Heavy-Duty In-Use Testing (HDIUT) Program and from
MOVES3 for a MY 2014 vehicle with an fscaie of 7 metric tons
304

-------

0.45

0.40

0.35
c

o
0.30
o

ro
0.25
Ll-
CU

T3
0.20
O
E

Q-
0.15
O


0.10

0.05

0.00
LHDD Opmode Distributions
Comparison of HDIUT vs. MOVES-CTI
¦ HDIUT ¦ MOVES CTI - CY2027/MY2014
























!

-1

1 |

ll ll
II
II ll	
la ll ll II ¦¦ I- _
ll II II II ..
0 1 i 11 12 13 14 15 16 i 21 22 23 24 25 27 28 29 30 ; 33 35 37 38 39 40
Opmode
Figure G-6. Light Heavy-Duty Diesel Operating Mode Distribution compared between the Heavy-Duty In-Use Testing (HDIUT) Program and from
MOVES3uu for a MY 2014 vehicle with an fscaie of 5 metric tons.
uu The MOVES operating mode distribution excludes class 2b light-heavy-duty vehicles in the passenger truck and light-commercial truck source types The
vehicles included in the HDIU are all engine-certified vehicles and are class 3 or heavier.
305

-------
Appendix H THC and CO Emisssion rates from 2010 and Later
Model Year Heavy-duty Vehicles from the HDIUT
H. 1 Comparison of THC and CO Emission Rates by NOx FEL
Groups for MY 2010-2013 Vehicles in LHD and MHD
As noted in 2.1.3.2, this appendix section contains figures of the mean THC and CO emission rates
by NOx FEL Group for LHD and MHD. The figures for HHD are included in the main report.
H.l.l LHD

0.007 q

0.006 :

0.005 :
			
-

0.004 -
Ofl
-
0
1
0.003 :
1-


0.002 :

o.ooi :

0.000 :
I LHD FEL 0.20, MY 2010-2013
I LHD FEL 0.35, N=0
I LHD FEL 0.50, MY 2010-2015
N=52
N=15
j_I_i
jjjjj
ill
i« i I I
iiiili
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 H-l. Average LHD THC Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013 and
the 0.5 NOx FEL for MY 2010-2016. Error Bars are 95% Confidence Intervals of the Mean
306

-------
(/)
3
o
u
0.04
0.03
0.03
0.02 H
0.02
0.01
0.01
0.00
ii
ii
I LHD FEL 0.20, MY 2010-2013, N=52
I LHD FEL 0.35, N=0	I
I LHD FEL 0.50, MY 2010-2015, N=15

Ii ii il I
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 H-2.Average LHD THC Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013 and
the 0.5 NOx FEL for MY 2010-2016. Error Bars are 95% Confidence Intervals of the Mean
H.1.2MHD
0.006
0.005 -
^ 0.004
bo
u 0.003 -
X
I-
0.002
0.001 -|
0.000
	I	I	
I MHD FEL 0.20, MY 2010-2013, N=23
l MHD FEL 0.35, MY 2010-20}3, N=23
I MHD FEL 0.50, MY 2010-2013, N=9
liii
JUiJ
ill
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 H-3.Average MHD THC Emission Rates by Operating Mode for the 0.2,0.35 and 0.50 NOx FEL Groups
for MY 2010-2013 Vehicles. Error Bars are 95% Confidence Intervals of the Mean
307

-------
MOVES OpMode
Figure H-4. Average MHD CO Emission Rates by Operating Mode for the 0.2,0.35 and 0.50 NO* FEL Groups
for MY 2010-2013 Vehicles. Error Bars are 95% Confidence Intervals of the Mean
308

-------
H.2 Comparison of THC and CO Emission Rates between MY 2010-
2013 and MY 2014 in the 0.2 NOx FEL Group for LHD and MHD
The following figures show the comparison of the mean THC and CO emission rates between
model year 2010-2013 and 2014 and later vehicles in the 0.2 NOx FEL Group for the LHD and
MHD regulatory class by operating mode. The figures for HHD are located in the main report
(Figure 2-42 and Figure 2-43).
H.2.1 LHD
0.0040
0.0035
— 0.0030
00
u
0.0025
0.0020
0.0015
0.0010
0.0005
0.0000

¦ MY 2010-2013




¦ MY 2014-2016




















T






ii

I i
T i
i i 1
T
11
I 1
1 1

ii ii
ii Ii Ii 1
i 1
ii ii ii Ii Ii Ii

1
1. 1. 1 1
LI
Ii 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 H-5. THC emission rates for the MY 2010-2013 and MY 2014-2016 vehicles in the LHD 0.20 NOx FEL
Group
0.08
0.07
0.06
„ 0.05
¦3 0.04
O
U 0.03
0.02
0.01
0.00

¦ MY 2010-2013

r

¦ MY 2014-2016
!















|

T I
l iv ii ii ii ii ii ^

T I


« ii
		 ii i1 1

ii i1 il i
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
MOVES OpMode
Figure H-6. CO emission rates for the MY 2010-2013 and MY 2014-2016 vehicles in the LHD 0.20 NOx FEL
Group
309

-------
H.2.2 MHD
0.0014
0.0012
^ 0.0010
J25
^ 0.0008
H
0.0006
0.0004
0.0002
0.0000
MY 2010-2013
MY 2014-2015
i.
i ii il
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 H-7. THC emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the MHD 0.20 NQx FEL
Group
0.014
0.012
0.010
< 0.008
ap
g 0.006
0.004
0.002
0.000
IV ft\f ^ f~\*1 r\ n *"*


¦	MY 2010-2013
¦	MY 2014-2015























¦















TT










[



li il II1



ill
ill









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 H-8. CO emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the MHD 0.20 N()x FEL
Group
310

-------
Appendix I Analysis of 2010 and Later Model Year Heavy-duty
Gasoline Emission Rates
This appendix contains additional details of the analysis used to update the 2010 and later model
year heavy-duty gasoline running emission rates documented in Section 3.1.1.2.1.
1.1 Removal of Start Emissions from Real-World PEMS Data in
Developing Heavy-duty Gasoline Running Exhaust Emissions
The running exhaust emissions rates update for 2010 model year and later heavy-duty gasoline
vehicles is meant to include emissions from only the hot-running condition. Thus, ideally,
emissions assigned to start effects should be removed before estimating operating mode-based
average rates per test and per vehicle. This is less of a concern if each test is a full-day of operation
since the incremental start emissions might then be a small fraction of total emissions. However,
on-road tests of the three HD gasoline involved drive cycles that range from 10 to 90 minutes in
duration. Also, the idle tests, of 15 or 30 minute duration, need to have start effects removed to
ensure their contribution to OpMode 1 (idle mode) rate is unaffected by start emissions. Note that
the effect of start emissions is modeled as a separate process in MOVES and by removing them
from the running emissions, we are minimizing double-counting.
Start emissions in the Federal Test Procedure are calculated as Bag 1 minus Bag 3 of the FTP
cycle, where Bag 1 is driving after a cold start and Bag 3 is the same cycle as Bag 1 but under hot-
stabilized conditions. This method is not possible in real-world testing because it is not possible to
replicate the exact drive cycle due to varying traffic conditions. Thus, we decided to define start
emissions as the incremental emissions that occur before the TWC reaches the light-off condition
where it achieves optimal emissions reduction efficacy. We define light-off condition as the point
when the TWC first reaches 421 °C (790 °F). TWC light-off temperatures are based on design
specifics but are generally in the range of 400 °C. The selection of 421 °C as the criteria is
somewhat arbitrary at the very precise level - there is not a good reason why 421 °C is more
appropriate than say 410 °C or 430 °C. We picked 421 °C based on visual comparison of a handful
of the on-road tests for each of the three gasoline vehicles to find out at what point the TWC
temperature starts to stabilize. The effect of soak time on time to reach 421 °C catalyst temperature
and grams of emissions assigned to the start effect, thus removed from running exhaust emissions,
are shown in Table 1-1. The following figure shows the data for NOx. Interestingly, the trend for
NOx from on-road testing is comparable to the trends from previous lab-based testing, shown in
Figure 3-23. For the on-road data, grams of NOx from starts emissions for 105-minute soak is 1.15
times the 720-minute soak. For the same conditions, the ratios in Figure 3-23 are approximately
1.17 and 1.37 for the data series labeled as "MOVES" and "New Data", respectively. The trends
for THC and CO are also similar between the two figures.
311

-------
Table 1-1 Time and Pollutant Mass for Driving Assigned to Start Emissions
Soak
Time
(min)
Number
of Tests
Avg. time1 for
TWC to reach 421
°C (sec)
Avg. grams of pollutant removed
NOx
C02
CO
THC
0
109
78
0.2
356
3
0.4
3
6
42
0.02
213
1
0.03
18
6
63
0.1
265
3
0.3
30
6
91
0.8
427
9
0.8
45
8
114
1.9
493
14
1.6
75
5
122
1.8
470
16
1.8
105
7
102
2.3
463
19
1.9
180
4
107
3.0
531
22
2.7
240
2
94
1.1
424
18
1.9
360
1
1
0.00
0
0
0.00
720
48
125
2.0
662
25
3.3
Note:
1 Of the total 202 tests listed here, in three tests the catalyst never reached 421 °C, so they are not included in the
average time calculation, however, the grains of pollutant removed columns include these three tests.
¦o
cu
>
o
E
C1J
X
O
3.5
3.0
2.5
2.0
25 1.5
1.0
0.5
0.0
100 200 300 400 500
SoakTime (mins)
600
700
800
Figure 1-1. Grams of NOx from Start Emissions versus Soak Time
312

-------
1.2 Comparison of Heavy-duty Gasoline Emission Rates by Vehicle
The figures in this subsection show the emission rates calculated from the PEMS testing data
collected on three heavy-duty gasoline vehicle. The emission rates in these figures were analyzed
using an fscale of 5 metric tons used for the LHD2b3 and LHD45 regulatory class emission rates.
The error bars are the 95% confidence intervals of the mean calculated by treating the number of
routes (R) as independent random variables.
The Isuzu NPR has the highest NOx emission rates across all operating modes. The Ford E459 and
the Ram 3500 have more similar NOx emission rates, except for the high speed and power
operating modes (opModelD 39 and 40).
0.05
0.04
^ 0.03
i 0.02
0.01
0.00
I 2016_Ford_E459, R=76
I 2015Jsuzu_NPR, R=83
I 2017_RAM_3500, R=38
... ,1. ll.
jj. ii. ii.

ll ii ili ill
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 1-2. Mean Heavy-duty Gasoline NOx Emission Rates by Operating Mode and Vehicle Calculated using
fscaie of 5 metric tons.
For THC and CO, the Isuzu NPR tends to have lower emission rates. The differences between the
vehicles is more dependent on operating mode. The Ram 3500 has the highest emission rates for
the high STP and high speed operating modes, but is more comparable to the other vehicles at the
low STP and low speed operating mode bins.
313

-------
0.045
0.040
0.035
0.030
IS 0.025
^ 0.020
I-
0.015
0.010
0.005
0.000
i 2016_Ford_E459, R=76
I 2015_ISUZU_NPR, R=83
I 2017_RAM_3500, R=38
11 12 13 14 15 16
li- ill

i
jti ' I it
21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
0
I.
ii. lii Ik;,	is	a,. L ii. ii. iii l«l ii
11 12 13 14 15 16 ; 21 22 23 24 25 27 28 29 30


33 35 37 38 39 40
MOVES OpMode
Figure 1-3. Mean Heavy-duty Gasoline THC Emission Rates by Operating Mode and Vehicle Calculated using
fscale of 5 metric tons.
9.00
8.00
7.00
6.00
J/i
^2 5.00 3
g 4.00
3.00
2.00
1.00
0.00
0.60
0.50
0.40
0.30
0.20
0.10
0.00
¦	2016_Ford_E459, R=76
¦	2015_lsuzu_NPR, R=83
¦	2017 RAM 3500, R=38
= -3 rl
iiii
0 1 ; 11 12 13 14 15 16
_y_ik
J
21 22 23 24 25 27 28 29 30
33 35 3 7 38 39 40

U_L
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 1-4. Mean Heavy-duty Gasoline CO Emission Rates by Operating Mode and Vehicle Calculated using
fscale of 5 metric tons.
314

-------
1.3Extrapolating High-Power Operating Modes
When analyzing the heavy-duty gasoline data for the MHD and HHD regulatory classes (using an
fscale of 7 and 10 metric tons respectively), there was limited or no data for high power operating
mode bins. The figure below shows the mean NOx emission rates by vehicle and operating mode
when using an using mfscale of 10 used for HHD vehicles. Operating modes 29, 30, 39, 40 are
missing data from at least one of the tested vehicles. Operating mode 16 and 28 have limited data
from the Ford and Isuzu vehicles (less than 10 routes had data measured in those operating mode
bins). The small amount of data in these operating mode bins decreases our confidence in the mean
operating modes, and can lead to inconsistent trends in emission rates with power. For example,
operating mode 16 has lower mean NOx emission rates than operating mode 15.
0.05
l/l
.29
x
O
0.04
0.03
0.02
0.01
0.00
0 1
I 2016_Ford_E459, R=80
I 2015_lsuzu_NPR, R=84
I 2017 RAM 3500, R=38
11 12 13 14 15 16
_iL
£
li
21 22 23 24 25 27 28 29 30
I1, i i
33 35 37 38 39 40
MOVES OpMode
Figure 1-5. Mean Heavy-duty Gasoline NOx Emission Rates by Operating Mode and Vehicle Calculated for
HHD using an fscaie of 10 metric tons
In the case of missing data, we used the emission rates of the nearest operating mode bin with data.
In case where there were limited data, we aggregated the averages with data from the next closest
bin into into single averages. The figure below displays the resulting emission rates by vehicles for
NOx for HHD vehicles. For the Ford and Isuzu vehicles, the following operating modes were
aggregated into single averages: 15-16, 27-30, 38-40. For the RAM vehicle, operating modes 29-30
were aggregated. The same aggregation was used to calculated the updated TFIC and CO emission
rates.
315

-------
0.05
0,04
~ 0.03 H
st>
x
2 0.02 -
0.01 -
0.00
12016_Ford_E459, R=80
I 2015_lsuzu_NPR, R=84
I 2017_RAM_3500, R=38
- J. I. Ill i L
JLi
Jj
Job
11 12 13 14 15 16 ; 21 22 23 24 25 27 28 29 30
MOVES OpMode
ill I III
33 35 37 38 39 40
Figure 1-6. Mean Heavy-duty Gasoline NOx Emission Rates by Operating Mode and Vehicle Calculated for
HHD using an fscale of 10 metric tons with Aggregated Means for High Power Bins with Limited Data
The figure below shows the weighted average NOx emission rate calculated by averaging the three
vehicles together according to their production volume sales.
i/>
ao
x
O
0.040
0.035
0.030
0.025
0.020
0.015
0.010
0.005
0.000
0 1
I HHDG, MY 2016
¦ ill
11 12 13 14 15 16
¦ I
21 22 23 24 25 27 28 29 30
MOVES OpMode
lI
33 35 37 38 39 40
Figure 1-7. Weighted Average Heavy-duty Gasoline NOx Emission Rates by Operating Mode for HHD Using
Production Volumes
For CO2 emission rates, we used a different method than for TITC, CO, and NOs, Rather than
aggregate the emission rates with limited or no data, we extrapolated the higher operating modes
(30, 39, and 40) using the STP values using Equation 2-5. The assumed mid-point STP for each
operating mode bin is displayed in Table 1-2.
316

-------
Table 1-3 Assumed STP Midpoint for Each Operating Mode
OpModelD
STP midpoint
0
-
1
-
11
-
12
1.5
13
4.5
14
7.5
15
10.5
16
13.5
21
-
22
1.5
23
4.5
24
7.5
25
10.5
27
15
28
21
29
27
30
33
33
3
35
9
37
15
38
21
39
27
40
33
The figure below displays the initial mean CO2 emission rates using an fscale of 10 metric tons. Note
that the emission rates for CO2 have more consistent trends than other measured pollutants; because
of this, fewer of the high power operating modes were replaced with extrapolated rates than for the
THC, CO, and NOx emission rates. In this case, we decided to only replace one operating mode
with limited data (operating mode 39 for the Isuzu vehicle), because it was only based on one route
(which is why there are no error bars).
317

-------
l-l
0 1
I 2016_Ford_E459, R=80
12015Jsuzu_NPR, R=84
I 2017_RAM_3500, R=38
¦II IB
11 12 13 14 15 16
111
21 22 23 24 25 27 28 29 30
33 35 37 38 39 40
MOVES OpMode
Figure I-8.Mean Heavy-duty Gasoline CO2 Emission Rates by Operating Mode and Vehicle Calculated for HHD
using an fscaieof 10 metric tons
The figure below shows the mean CO2 emission rates using an fscaie of 10 metric tons with
extrapolated emission rates for the high power bins. For the Ford and Isuzu vehicles operating
modes 29 and 30 were extrapolated from operating mode 28, and operating modes 39 and 40 were
extrapolated from 38. For the RAM, operating mode 30 was extrapolated from 29.
1/1
3B
tN
o
u
I 2016_Ford_E459, R=80
I 2015Jsuzu_NPR, R=84
I 2017_RAM_3500, R=38
11 12 13 14 15 16
21 22 23 24 25 27 28 29 30
MOVES OpMode

33 35 37 38 39 40
Figure 1-9. Mean Heavy-duty Gasoline CO2 Emission Rates by Operating Mode and Vehicle Calculated for
HHD using an fscaie of 10 metric tons with Extrapolated Means for High Power Bins.
The figure below shows the weighted average CO2 emission rate calculated by averaging the three
vehicles together according to their production volume sales. Note that the CO2 have stronger
increasing trends with power compared to NOx.
318

-------

¦ HHDG, MY 2016

















¦

|



I











¦
.1





I
1 1








ll




0 1
11
12 13 14 15 1
6
21
22 23 24 25 2
7 28 29 30
33 35 37 38 39 40
MOVES OpMode
Figure X-10. Weighted Average Heavy-duty Gasoline CO2 Emission Rates by Operating Mode for HHD Using
Production Volumes
Similar calculations were repeated for the MI-ID vehicles calculated using an^i? of 7 metric tons.
319

-------
Appendix J PM Composition Measurements from Auxiliary
Power Units
Table J-l reports the organic carbon (OC), elemental carbon (EC) and total carbon (TC)
measurements conducted in the study conducted by Texas Transportation Institute (TTI, 201495).
All the measurements were collected on APU 1. TTI collected the particulate sample on quartz
fiber filters, and Sunset Laboratory Inc. analyzed the filters using thermal optical reflectance (TOR)
using the IMPROVE (Interagency Monitoring of Protected Visual Environments) procedures. Total
Carbon (TC) is the sum of Elemental Carbon (EC) and Organic Carbon (OC).
Table J-l. Organic Carbon, Elemental Carbon, and Total Carbon Measurements from the IMPROVETOR
measured on APU 1
Sample ID
Min.
DR
Test
OC
OC
uncertainty
EC
EC
uncertainty
TC
TC
uncertainty
EC/TC ratio
(Hg/cm2)
(Hg/cm2)
(Hg/cm2)
(Hg/cm2)
(Hg/cm2)
(Hg/cm2)
APU 005
10
30/1
Hot Test 1
66.35
3.42
12.98
0.75
79.33
4.17
0.16
APU 006
10
30/1
Hot Test 2
65.26
3.36
13.45
0.77
78.70
4.14
0.17
APU 007
10
30/1
Hot Test 3
59.24
3.06
10.51
0.63
69.75
3.69
0.15
APU 009
20
6/1
DPF Hot APU 1
13.85
0.79
0.86
0.14
14.71
0.94
0.06
APU 010
20
6/1
DPF Hot APU 1
14.67
0.83
1.12
0.16
15.79
0.99
0.07
APU 011
20
6/1
DPF Hot APU 1
13.18
0.76
0.93
0.15
14.11
0.91
0.07
APU 012
20
6/1
DPF Cold APU 1
16.62
0.93
1.45
0.17
18.07
1.10
0.08
APU 013
20
6/1
DPF Cold APU 1
15.86
0.89
1.40
0.17
17.27
1.06
0.08
APU 014
20
6/1
DPF Cold APU 1
17.59
0.98
1.56
0.18
19.15
1.16
0.08
APU 015
10
30/1
Cold Test 1
75.74
3.89
9.65
0.58
85.39
4.47
0.11
APU 016
10
30/1
Cold Test 2
73.83
3.79
9.61
0.58
83.44
4.37
0.12
APU_017
10
30/1
Cold Test 3
77.47
3.97
9.90
0.59
87.37
4.57
0.11
320

-------
9 References
1	USEPA (2023). Speciation of Total Organic Gas and Particulate Matter Emissions from Onroacl Vehicles in
MOVES4. EPA-420-R-23-006. Office of Transportation and Air Quality. US Environmental Protection Agency. Ann
Arbor, MI. August 2023. https://www.epa.gov/moves/moves-technical-reports .
2	USEPA (2020). Air Toxic Emissions from Onroacl Vehicles in MOVES3. EPA-420-R-20-022. Office of
Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI. November 2020.
https://www.epa.gov/moves/moves-technical-reports
3	USEPA (2023). Greenhouse Gas and Energy Consumption Rates for On-road Vehicles in MOVES4. EPA-420-R-
23-026. Assessment and Standards Division. Office of Transportation and Air Quality. US Environmental Protection
Agency. Ann Arbor, MI. August 2023. https://www.epa.gov/moves/moves-technical-reports.
4	USEPA (2023). Evaporative Emissions from Onroad Vehicles in MOVES4. EPA-420-R-23-006. Office of
Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI. August 2023.
https://www.epa.gov/moves/moves-technical-reports.
5	USEPA (2020). Brake and Tire Wear Emissions from Onroad Vehicles in MOVES3. EPA-420-R-20-014. Office of
Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI. November 2020.
https://www.epa.gov/moves/moves-technical-reports.
6	USEPA (2023). Population and Activity of Onroad Vehicles in MOVES4. EPA-420-R-23-005. Office of
Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI.
7	US EPA. Heavy-Duty Highway Compression-Ignition Engines and Urban Buses—Exhaust Emission Standards.
https://www.epa.gov/emission-standards-reference-guide/epa-emission-standards-heaw-dutv-highwav-engines-and-
vehicles. Accessed August 2023.
8	40 CFR§ 86.091(2).
9	USEPA (2022). Exhaust Emission Rates for Light-Duty Onroad Vehicles in MOVES3.R1. Office of Transportation
and Air Quality. US Environmental Protection Agency. Ann Arbor, MI.
10	Truck Trailer Manufacturers Association, Inc. v. Environmental Protection Agency, 16-1430 (D.C. Cir. 2021)
11	USEPA (2015). U.S. Environmental Protection Agency Peer Review Handbook. EPA/100/B-15/001. Prepared for
the U.S. Environmental Protection Agency under the direction of the EPA Peer Review Advisory Group. Washington,
D.C. 20460. October 2015. https://www.epa.gov/sites/production/files/202Q-
08/documents/epa peer review handbook 4th edition.pdf.
12	USEPA Science Inventory, https ://cfpub.epa. gov/si/
13	USEPA (2022). Final Rule and Related Materials for Control of Air Pollution from New Motor Vehicles: Heavy-
Duty Engine and Vehicle Standards, Office of Transportation and Air Quality. US Environmental Protection Agency,
Ann Arbor, MI., December 2022. https://www.epa.gov/regulations-emissions-vehicles-and-engines/final-rule-and-
related-materials-control-air-pollution
14. USEPA (1998), Caterpillar, Inc., Detroit Diesel Corporation, Mack Trucks, Inc., Navistar International
Transportation Corporation, Renault Vehicules Industriels, s.a., and Volvo Truck Corporation Diesel Engines
Settlement. October 22,1998. http://cfpub.epa.gov/enforcement/cases/.
15 Jack, Jason A. U.S. Army Aberdeen Test Center Support of Heavy Duty Diesel Engine Emissions Testing. U.S. Army
Aberdeen Test Center CSTE-DTC-AT-SL-E, Aberdeen Proving Ground, Maryland.
(https://gaftp.epa.gov/air/nei/ei conlcrcnce/E 115/session 1 /iack.pdD.
321

-------
16	McClement, Dennis. Reformatting On-Road In-Use Heavy-Duty Emissions Test Data. Sierra Research, Sacramento,
CA. April 2008.
17	Gautam, Mridul, Nigel N. Clark, Gregory Thompson, Daniel K. Carder, and Donald W. Lyons. Evaluation of Mobile
Monitoring Technologies for Heavy-duty Diesel-Powered Vehicle Emissions. Dept. Mechanical and Aerospace
Engineering, College of Engineering and Mineral Resources, West Virginia University, Morgantown, WV.
18	Gautam, Mridul, Nigel N. Clark, Gregory Thompson, Daniel K. Carder, and Donald W. Lyons. Development ofln-
use Testing Procedures for Heavy-Duty Diesel-Powered Vehicle Emissions. Dept. Mechanical and Aerospace
Engineering, College of Engineering and Mineral Resources, West Virginia University, Morgantown, WV.
19	Gautam, M., et al. Evaluation of In-Use Heavy-Duty Vehicle Emissions Using the Mobile Emissions Measurement
System (MEMS) for Engine Model Years 2001 to 2003 : Final Reports. Present to engine manufacturers to fulfill
testing requirements documented in Phases III and IV of the Heavy Duty Diesel Engine consent decree. Dept.
Mechanical and Aerospace Engineering, College of Engineering and Mineral Resources, West Virginia University,
Morgantown, WV. 2002 & 2007.
20	"Control of Emissions of Air Pollution From New Motor Vehicles: In-Use Testing for Heavy-Duty Diesel Engines
and Vehicles", 70 FR 34594, June 2005.
21	USEPA (2019). Manufacturer-Run In-Use Testing Program Data for Heavy-Duty Diesel Engines. Retrieved
September 17, 2020, from https://www.epa.gov/compliance-and-fuel-economv-data/manufacturer-run-use-testing-
program-data-heaw-dutv-diesel-3.
22	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
23	Data Collection of Drayage Trucks in Houston-Galveston Port Area Draft Report. EP-C-06-080. May 27, 2011
24	Bradley, Ron. "Technology Roadmap for the 21st Century Truck Program." U.S. Department of Energy: Energy
Efficiency and Renewable Energy, Washington, D.C., December 2000.
25	Rakha, Hesham and Ivana Lucic. Variable Power Vehicle Dynamics Model for Estimating Truck Accelerations. Page
6.
26	National Renewable Energy Laboratory. Development of LNG-Powered heavy-Duty Trucks in Commercial Hauling.
NREL/SR-540-25154, Golden, CO, December 1998.
27	Goodyear. "Factors Affecting Truck Fuel Economy - Section 9" Page 5.
28	Ramsay, Euan and Jonathan Bunker. Acceleration of Multi-Combination Vehicles in Urban Arterial Traffic
Corridors. PhD dissertation, Queensland University of Technology. August 2003, Page 11.
http://eprints.aut.edu.au/archive/00002359/01/RS&ETechForum2003 Ramsav&Bunker 2.pdf
29	Society of Automotive Engineers. Commercial Truck and Bus SAE Recommended Procedure for Vehicle
Performance Prediction and Charting. SAE J2188. Revised October 2003.
30	Bradley, Ron. "Technology Roadmap for the 21st Century Truck Program." U.S. Department of Energy: Energy
Efficiency and Renewable Energy, Washington, D.C., December 2000. Page 32.
31	Pritchard, Ewan G. D. and Richard R. Johnson. Hybrid Electric School Bus Preliminary Technical Feasibility
Report. Advanced Energy Corporation and Department of Mechanical Engineering, North Carolina State University,
Raleigh, NC, September 14, 2004. Page 25.
32	Hedrick, J.K. and A. Ni. Vehicle Modeling and Verification of CNG-Powered Transit Buses. California PATH
Working Paper UCB-ITS-PWP-2004-3. California Partners for Advanced Transit and Highways (PATH), Institute of
Transportation Studies, University of California, Berkeley. February 2004. Page 21.
http://repositories.cdlib.org/cgi/viewcontent.cgi?article=1169&context=its/path
322

-------
33	Motor Industry Research Association (MIRA). News from MIRA - Automotive Engineering Specialists. Hybrid
Theory: Hybrid Vehicle engineering for economy, the environment, and customer delight. Nuneaton, Warwickshire,
UK. Issue 2, Spring 2007.
34	Choi, D., J. Koupal and M. Church (2012). Analysis of Recent Heavy-Duty Vehicle Emission Test Programs.
MOVES Review Workgroup, Ann Arbor, MI. September 25, 2012.
35	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.
36	USEPA (2002). Update of Heavy-Duty Emission Levels (Model Years 1988-2004) for Use in MOBILE6. EPA-420-
R-02-018. July 2002. https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P10022RL.txt. (Table 17)
37	USEPA Office of Transportation and Air Quality. Heavy Duty Diesel Engine Consent Decree Low NOx Rebuild
Program Summary. Washington, D.C. https://www.epa.gov/sites/default/files/2015-
0l/documents/032807mstrs Ionoxsum4q07.pdf.
38	USEPA (2014). Tier 3 Vehicle Emission and Fuel Standards Program. Regulatory Impact Analysis. EPA-420-R-14-
004.	February 2014. https://www.epa.gov/regulations-emissions-vehicles-and-engines/final-rule-control-air-pollution-
motor-vehicles-tier-3
392007/2010 Heavy-duty rulemaking. 66 FR 5002, January 18, 2001
40	USEPA (2020). Exhaust Emission Rates for Heavy-Duty On-road Vehicles in MOVES3. EPA-420-R-20-018.
Assessment and Standards Division. Office of Transportation and Air Quality. US Environmental Protection Agency.
Ann Arbor, MI. November 2020. https://www.epa.gov/moves/moves-onroad-technical-reports.
41	Clark, Nigel et al. California Heavy Heavy-Duty Diesel Truck Emissions Characterization for Program E-55/59.
West Virginia University Research Corporation. Morgantown, WV. November 2005.
42	Hsu, Y., and Mullen, M. 2007. Compilation of Diesel Emissions Speciation Data. Prepared by E. H. Pechan and
Associates for the Coordinating Research Council. CRC Contract No. E-75, October, 2007.
43	Nam, Ed and Robert Giannelli. Fuel Consumption Modeling of Conventional and Advanced Technology Vehicles in
the Physical Emission Rate Estimator (PERE). EPA420-P-05-001. USEPA Office of Transportation and Air Quality,
Assessment and Standards Division, Ann Arbor, MI. February 2005.
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P1001D6I.txt
44	Kinsey, J. S., W. A. Mitchell, W. C. Squier, K. Linna, F. G. King, R. Logan, Y. Dong, G. J. Thompson and N. N.
Clark (2006). Evaluation of methods for the determination of diesel-generated fine particulate matter: Physical
characterization results. Journal of Aerosol Science, 37 (1), 63-87. DOI:
http://dx.doi.Org/10.1016/i.iaerosci.2005.03.007.
45	USEPA (2009). Development of Emission Rates for Heavy-Duty Vehicles in the Motor Vehicle Emissions Simulator
(DraftMOVES2009). EPA-420-P-09-005. Office of Transportation and Air Quality. US Environmental Protection
Agency. Ann Arbor, MI. August, 2009. https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P10050CK.txt
46	USEPA Office of Transportation and Air Quality. Update Heavy-Duty Engine Emission Conversion Factors for
MOBILE6: Analysis of BSFCs and Calculation of Heavy-Duty Engine Emission Conversion Factors. EPA420-R-02-
005,	M6.HDE.004. Assessment and Standards Division, Ann Arbor, ML
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P10022Ll.txt
47	Khalek, Imad, Thomas L Bougher and Patrick M. Merritt. Phase 1 of the Advanced Collaborative Emissions Study
(ACES). SwRI Project No. 03.13062. Southwest Research Institute, San Antonio, TX; Coordinating Research Council
(CRC), Alpharetta, GA; Health Effects Institute, Boston, MA. June 2009.
323

-------
48	US EPA. 2012. Black Carbon Report to Congress. EPA-450/R-12-001. March 2012.
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P100EIJZ.txt
49	Quiros, D. C., A. Thiruvengadam, S. Pradhan, M. Besch, P. Thiruvengadam, B. Demirgok, D. Carder, A. Oshinuga,
T. Huai and S. Hu (2016). Real-World Emissions from Modern Heavy-Duty Diesel, Natural Gas, and Hybrid Diesel
Trucks Operating Along Major California Freight Corridors. Emission Control Science and Technology, 2 (3), 156-
172. DOI: 10.1007/s40825-016-0044-0.
50	Dixit, P., J. W. Miller, D. R. Cocker, A. Oshinuga, Y. Jiang, T. D. Durbin and K. C. Johnson (2017). Differences
between emissions measured in urban driving and certification testing of heavy-duty diesel engines. Atmospheric
Environment, 166, 276-285. DOI:
https://www.sciencedirect.com/science/article/pii/S13522310173041817via%3Dihub.
51	Thiruvengadam, A., M. C. Besch, P. Thiruvengadam, S. Pradhan, D. Carder, H. Kappanna, M. Gautam, A.
Oshinuga, H. Hogo and M. Miyasato (2015). Emission Rates of Regulated Pollutants from Current Technology Heavy-
Duty Diesel and Natural Gas Goods Movement Vehicles. Environ Sci Technol, 49 (8), 5236-5244. DOI:
10.1021/acs.est.5b00943.
52	40 CFR § 86.1816-18. Table 1- Fully Phased-in Tier 3 HDV Exhaust Emission Standards.
53	Graboski, Michael S., Robert L. McCormick, Janet Yanowitz, and Lisa Ryan. Heavy-Duty Diesel Vehicle Testing for
the Northern Front Range Air Quality Study. Colorado Institute for Fuels and High-Altitude Engine Research,
Colorado School of Mines, Golden, Colorado. Prepared for Colorado State University. February 1998.
54	Energy and Environmental Analysis, Inc. Documentation and Analysis of Heavy-Duty Diesel Vehicle Emission Test
Data. Prepared for New York Department of Environmental Conservation, December 2000.
55	40 CFR § 1065.260
56	40 CFR § 1065.250
57	USEPA (2012). Updates to the Greenhouse Gas and Energy Consumption Rates in MOVES2010a. EPA-420-R-12-
025. Office of Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI. August, 2012.
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P100F3Z3.txt.
58	USEPA (2014). Greenhouse Gas Emissions from a Typical Passenger Vehicle. EPA-420-F-14-040a. Office of
Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI. May 2014.
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P100LQ99.txt.
59	USEPA (2011). Greenhouse Gas Emission Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty
Engines and Vehicles (76 FR 57106, September 15, 2011)
60	40 Code of Federal Register Volume 76 at 57216 and 57236, September 15, 2011.
61	Greenhouse Gas Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles -
Phase 2 Final Regulatory Impact Analysis, Chapter 5. EPA-420-R-16-900. August 2016
62	USEPA (2012). 2017 and Later Model Year Light-Duty Vehicle Greenhouse Gas Emissions and Corporate Average
Fuel Economy Standards (77 FR No. 199, October 15, 2012)
63	USEPA (2023). Emission Adjustments for Onroad Vehicles in MOVES4. EPA-420-R-23-021. Office of
Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI.
64	Calcagno, James A. Evaluation of Heavy-Duty Diesel Vehicle Emissions During Cold-Start and Steady-State Idling
Conditions and Reduction of Emissions from a Truck-Stop Electrification Program. PhD Dissertation, Department of
Civil and Environmental Engineering, University of Tennessee, Knoxville, December 2005.
65	40 CFR Part 86, Appendix I (f) (2)
66	Evaluating Technologies and Methods to Lower Nitrogen Oxide Emissions From Heavy-Duty Vehicles. Final
Report. SwRI Project 19503. April, 2017. https://ww2.arb.ca.gov/sites/default/files/classic/research/apr/past/13-312.pdf
324

-------
67	EMFAC 2000 Section 6.7 START CORRECTION FACTORS. 4/20/2000. On-Road Emissions Model Methodology
Documentation.
68	Glover, E.; Carey, P. Determination of Start Emissions as a Function of Mileage and Soak Time for 1981-1993
Model-year Light-Duty Vehicles. EPA420-R-01-058 (M6.STE.003). USEPA Office of Transportation and Air Quality,
Ann Arbor, MI. November, 2001.
69	USEPA (2017). Heavy-Duty Chassis Start Emissions Testing. Office of Transportation and Air Quality. US
Environmental Protection Agency, Ann Arbor, MI. July 2017.
70	USEPA (2017). On-Road Light-Duty and Heavy-Duty Vehicle Start Emissions Testing. Office of Transportation and
Air Quality. US Environmental Protection Agency, Ann Arbor, MI. July 2017
71	USEPA (2005). Energy and Emissions Inputs. EPA-420-P-05-003. Office of Transportation and Air Quality. US
Environmental Protection Agency. Ann Arbor, MI. March, 2005.
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P1001DAQ.txt.
72	USEPA (2022). Control of Air Pollution from New Motor Vehicles: Heavy-Duty Engine and Vehicle Standards,
Draft Regulatory Impact Analysis. EPA-420-D-22-001. Office of Transportation and Air Quality. US Environmental
Protection Agency, Ann Arbor, ML, March 2022. https://nepis.epa.gov/Exe/ZvPDF.cgi?Dockev=P10144K0.pdf
73	Keel-Blackmon, K., S. Curran and M. V. Lapsa (2016). Summary of OEM Idling Recommendations from Vehicle
Owner s Manuals. Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States). Fuels, Engines and
Emissions Research Center; National Transportation Research Center (NTRC).
https://info.ornl. gov/sites/publications/Files/Pub61263 .pdf.
74McCormick, Robert, M. S. Graboski, T. L. Alleman, J. Yanowitz. Idle Emissions from Heavy-Duty Diesel and
Natural Gas Vehicles at High Altitude. Journal of the Air and Waste Management Association, 50(11): 1992-8. Revised
May 3, 2000.
75Lim, Han. Study of Exhaust Emissions from Idling Heavy-duty Diesel Trucks and Commercially Available Idle-
Reducing Devices. EPA420-R-02-025. US EPA Office of Transportation and Air Quality, Certification and
Compliance Division. October 2002.
76 Irick, David K. and Bob Wilson. NOx Emissions and Fuel Consumption of HDDVs during Extended Idle. University
of Tennessee, IdleAire Technologies Inc. In: Proceedings, Coordinated Research Council 12th Annual On-Road
Vehicle Emission Workshop, San Diego, California, April 15-17, 2002.
77Lambert, Douglas, et al. Roadside Emissions Study: Preliminary Results for Stationary and On-Road Testing of
Diesel Trucks in Tulare, California. California Environmental Protection Agency. Air Resources Board. Mobile Source
Operations Division, Clean Air Technologies International, Inc May 15, 2002.
78Gautam, Mridual and Nigel N. Clark. Heavy-duty Vehicle Chassis Dynamometer Testing for Emissions Inventory,
Air Quality Modeling, Source Apportionment and Air Toxics Emissions Inventory. Phase I Interim Report, CRC
Project No. E-55/E-59, West Virginia University Research Corporation, Morgantown, July 2002.
79National Cooperative Highway Research Program (NCHRP). Heavy-duty Vehicle Emissions. NCHRP Project 25-14,
Cambridge Systematics, Inc., Battelle Laboratories, Sierra Research and West Virginia University. October 2002.
80Tang, Shida and John Munn. Internal Report - Idle Emissions from Heavy-Duty Diesel Trucks in the New York
Metropolitan Area. New York State Dept of Environmental Conservation, November 9, 2001.
81Brodrick, Dwyer. Potential Benefits of Utilizing Fuel Cell Auxiliary Power Units in Lieu of Heavy-Duty Truck Engine
Idling. Paper UCD-ITS-REP-01-01. Institute of Transportation Studies, University of California, Davis, 2001.
82Storey, John M.E., John F. Thomas, Samuel A. Lewis, Sr., Thang Q. Dam, K. Edwards, Dean, Gerald L. DeVault,
and Dominic J. Retrossa. Particulate Matter and Aldehyde Emissions from Idling Heavy-Duty Diesel Trucks. SAE
Paper 2003-01-0289. Society of Automotive Engineers, Warrendale, PA.
325

-------
83	Keel-Blackmon, K., S. Curran and M. V. Lapsa (2016). Summary of OEM Idling Recommendations from Vehicle
Owner s Manuals. Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States). Fuels, Engines and
Emissions Research Center; National Transportation Research Center (NTRC).
https://info.ornl. gov/sites/publications/Files/Pub61263 .pdf.
84	Lutsey, N., Brodrick, C-J., Sperling, D., Oglesby, C., Transportation Research Record: Journal of the Transportation
Research Board, No 1880, TRB, National Research Council, Washington, D.C., 2004, pp. 28-38
85	Hoekzema, A. (2015). Modeling Truck Idling Emissions in Central Texas. 6800 Burleson Road, Building 310, Suite
165 Austin, Texas 78744, Capital Area Council of Governments.
86	Khan, A. S., N. N. Clark, M. Gautam, W. S. Wayne, G. J. Thompson and D. W. Lyons (2009). Idle Emissions from
Medium Heavy-Duty Diesel and Gasoline Trucks. Journal of the Air & Waste Management Association, 59 (3), 354-
359.
87	Farzaneh, M., J. Zietsman, D.-W. Lee, J. Johnson, N. Wood, T. Ramani and C. Gu (2014). TEXAS-SPECIFIC
DRIVE CYCLES AND IDLE EMISSIONS RATES FOR USING WITH EPA 'S MOVES MODEL. FHWA/TX-14/0-
6629-1. Texas A&M Transportation Institute. May, 2014. http://tti.tamu.edu/documents/0-6629-l.pdf.
88	ARB (2015). EMFAC2014 Volume III - Technical Documentation. California Environmental Protection Agency, Air
Resources Board, Mobile Source Analysis Branch, Air Quality Planning & Science Division. May 12, 2015.
89	California Environmental Protection Agency. Heavy-Duty Engines and Vehicles Executive Orders. Motor Vehicle
and Engine Certification Program. Air Resources Board. Available at: https://ww2.arb.ca.gov/new-vehicle-and-engine-
certification-executive-orders
90	US EPA, Navistar Inc. Heavy-Duty Engine Recall. EPA-420-F-13-038, Ann Arbor, MI: June 2013,
https://nepis.epa.gov/Exe/ZvPDF.cgi?Dockev=P100GNLJ.PDF
91	Clark, N. and M. Gautam (2007). HEA VY-DUTY Vehicle Chassis Dynamometer Testing for Emissions Inventory, Air
Quality Modeling, Source Apportionment and Air Toxics Emissions Inventory. CRC Report. No. E55/59. Aug-07.
92	US EPA, Engine Certification Data. On-Highway Heavy Duty - Diesel and Gasoline (2012).
https://19ianuarv2017snapshot.epa.gov/compliance-and-fuel-economv-data/engine-certification-data .html
93	Khalek, I. A., M. G. Blanks, P. M. Merritt and B. Zielinska (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. DOI: 10.1080/10962247.2015.1051606.
94	Khan, A. S., N. N. Clark, M. Gautam, W. S. Wayne, G. J. Thompson and D. W. Lyons (2009). Idle Emissions from
Medium Heavy-Duty Diesel and Gasoline Trucks. Journal of the Air & Waste Management Association, 59 (3), 354-
359. DOI: 10.3155/1047-3289.59.3.354.
95	Zietsman, J. and J. Johnson (2014). Auxiliary Power Unit Testing for SmartWay Idle Reduction Verification. DRAFT
FOR REVIEW. EP-1 l-H-000527, Auxiliary Power Unit Testing for SmartWay Idle Reduction Verification. Texas
A&M Transportation Institute. August, 2014.
96	Frey, H. C. and P.-Y. Kuo (2009). Real-World Energy Use and Emission Rates for Idling Long-Haul Trucks and
Selected Idle Reduction Technologies. Journal of the Air & Waste Management Association, 59 (7), 857-864. DOI:
10.3155/1047-3289.59.7.857.
97	TTI (2012). Development of a NOx Verification Protocol and Actual Testing of Onboard Idle Reduction
Technologies. New Technology Research and Development Program. Texas Transportation Institute. Revised:
January 2012.
98	Storey, J. M., J. F. Thomas, S. A. Lewis, T. Q. Dam, K. D. Edwards, G. L. DeVault and D. J. Retrossa (2003).
Particulate matter and aldehyde emissions from idling heavy-duty diesel trucks. SAE Technical Paper.
99	See 40 CFR 1037.106(g).
326

-------
100	US EPA, Nonroad Compression-Ignition Engines: Exhaust Emission Standards, EPA-420-B-16-022, March 2015.
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P1000A05.txt
101	US EPA. Memo to Docket: Updates to MOVES for Emissions Analysis of Greenhouse Gas Emissions and Fuel
Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles - Phase 2 FRM. August 8, 2016. Pages 33-
35. Docket EPA-HQ-OAR-2014-0827-2227.
102	US EPA, Frequently Asked Questions about Heavy-Duty "Glider Vehicles" and "Glider Kits", EPA-420-F-15-904,
Ann Arbor, MI: July 2015, https://nepis.epa.gov/Exe/ZvPDF.cgi?Dockev=P100MUVI.PDF
103	US EPA, Chassis Dynamometer Testing of Two Recent Model Year Heavy-Duty On-Highway Diesel Glider
Vehicles, EPA-HQ-OAR-2014-0827-2417, Ann Arbor, MI, November 2017,
https://www.re gulations.gov/document?D=EPA-HO-OAR-2014-Q827-2417
104	USEPA (2008). Mobile Source Observation Database (MSOD): User Guide and Reference. EPA420-B-08-017.
Office of Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI. December, 2008.
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P10028RC.txt
105	EPA. 1999. Tier 2 Vehicle & Gasoline Sulfur Program Final Rule, https://www.epa.gov/regulations-emissions-
vehicles-and-engines/final-rule-control-air-pollution-new-motor-vehicles-tier
106	USEPA (2017). Light-Duty Vehicles, Light-Duty Trucks, and Medium-Duty Passenger Vehicles: Tier 2 Exhaust
Emission Standards and Lmplementation Schedule EPA-420-B-17-028. Office of Transportation and Air Quality. US
Environmental Protection Agency. September 2017.
https://nepis.epa. gov/Exe/ZvPDF.cgi/P100SMOA.PDF?Dockev=P100SMQA.PDF.
107	Development of Emission Rates for Heavy-Duty Vehicles in the Motor Vehicle Emissions Simulator MOVES2010
(131 pp, EPA-420-B-12-049, August 2012) https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P100F80L.txt
108	EPA. 2001. "Control of Air Pollution From New Motor Vehicles: Heavy-Duty Engine and Vehicle Standards and
Highway Diesel Fuel Sulfur Control Requirements; Final Rule" https://www.epa.gov/regulations-emissions-vehicles-
and-engines/final-rule-control-air-pollution-new-motor-vehicles109 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.
110	USEPA (2022) Final Rule and Related Materials for Control of Air Pollution from New Motor Vehicles: Heavy-
Duty Engine and Vehicle Standards, Regulatory Impact Analysis. EPA-420-R-22-035, December 2022. Chapter 3.2.
111	USEPA. Final Rulemaking to Establish Greenhouse Gas Emission Standards and Fuel Efficiency Standards for
Medium- and Heavy-Duty Engines and Vehicles: Regulatory Impact Analysis. EPA-420-R-13-901, August 2011.
Page 2-30.
112	USEPA. Final Rulemaking to Establish Greenhouse Gas Emission Standards and Fuel Efficiency Standards for
Medium- and Heavy-Duty Engines and Vehicles - Phase 2: Regulatory Impact Analysis. EPA-420-R-16-900, August
2016. Page 5-14.
113	USEPA (2016). Heavy-Duty Highway Spark-Lgnition Engines: Exhaust Emission Standards. EPA-420-B-16-019.
Office of Transportation and Air Quality. US Environmental Protection Agency. March 2016.
https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P1000A01.pdf.
114	Toro, C., J. Warila, D. Sonntag, D. Choi and M. Beardsley (2019). Updates to "high-power" emission rates and
start deterioration for light-duty vehicles MOVES Review Workgroup, Ann Arbor, MI. April 10, 2019
https://www.epa.gOv/sites/production/files/2019-06/documents/03-updates-ld-emission-rates-start-deterioration-2019-
04-10.pdf.
327

-------
115	USEPA (2022). Control of Air Pollution from New Motor Vehicles: Heavy-Duty Engine and Vehicle Standards,
Draft Regulatory Impact Analysis. EPA-420-D-22-001. Office of Transportation and Air Quality. US Enviromnental
Protection Agency, Ann Arbor, MI., March 2022. https://nepis.epa.gov/Exe/ZvPDF.cgi?Dockev=P10144K0.pdf
116	USEPA (2015). Exhaust Emission Rates for Light-Duty On-road Vehicles in MOVES2014. EPA-420-R-15-005.
Assessment and Standards Division. Office of Transportation and Air Quality. US Enviromnental Protection Agency.
Ann Arbor, MI. October, 2015. https://www.epa.gOv/moves/moves-omoad-teclinical-reports#moves2014.
117	40 CFR Part 86 Appendix 1(f)(1)
118	Boyce, B. 2014. Cummins Westport - Heavy Duty Natural Gas Engines for Trucks and Buses, presented at the
Southeast Alternative Fuels Conference & Expo, October 22, Raleigh, NC, USA.
119	Ayala, A., Gebel, M., Okamoto, R., Rieger, P. et al., "Oxidation Catalyst Effect on CNG Transit Bus Emissions,"
Society of Automotive Engineers, SAE Technical Paper 2003-01-1900, 2003. (http://papers.sae.org/2003-01-190Q).
1211 "Central Business District (CBD)," Emission Test Cycles. DieselNet. Last Updated: September 2000, Accessed:
August 2023. (http://www.dieselnet.com/standards/cvcles/cbd.php).
121	DieselNet, (http://www.dieselnet.com/standards/cvcles/cbd.php).
122	Hesterberg, T.; Lapin, C.; Bunn, W.; 2008. "A Comparison of Emissions from Vehicles Fueled with Diesel or
Compressed Natural Gas." Environ. Sci. Technol. 42(17): 6437-6445. (http://pubs.acs.org/doi/abs/10.1021/esQ71718i).
123	Melendez, M.; Taylor, J.; Zuboy, J. et al. Emission Testing of Washington Metropolitan Area Transit Authority
(WA1ATA) Natural Gas and Diesel Transit Buses. Technical Report NREL/TP-540-36355, National Renewable Energy
Laboratory, Office of Energy Efficiency and Renewable Energy, Department of Energy. Golden, CO. December 2005.
(http://www.afdc.energy.gov/pdfs/36355.pdf).
124	LeTavec, C., Uihlein, J., Vertin, K., Chatteijee, S. et al., "Year-Long Evaluation of Trucks and Buses Equipped with
Passive Diesel Particulate Filters," Society of Automotive Engineers. SAE Technical Paper 2002-01-0433, 2002.
(http://papers.sae.org/2002-01-0433).
125	Ayala, A., Kado, N„ Okamoto, R., Holmen B. et al., "Diesel and CNG Heavy-duty Transit Bus Emissions over
Multiple Driving Schedules: Regulated Pollutants and Project Overview," Society of Automotive Engineers, SAE
Technical Paper 2002-01-1722, 2002. (http://papers.sae.org/2002-01-1722).
126	Lanni, T„ Frank, B„ Tang, S., Rosenblatt, D. et al., "Performance and Emissions Evaluation of Compressed Natural
Gas and Clean Diesel Buses at New York City's Metropolitan Transit Authority," Society of Automotive Engineers.,
SAE Technical Paper 2003-01-0300, 2003. (http://papers.sae.org/2003-01-030Q).
127	McKain, D„ Clark, N„ Balon, T., Moynihan, P. et al., "Characterization of Emissions from Hybrid-Electric and
Conventional Transit Buses," Society of Automotive Engineers, SAE Technical Paper 2000-01-2011, 2000.
(http ://papers. sae .org/2000-01-2011).
128	Clark, N„ Gautam, M„ Lyons, D„ Bata, R. et al., "Natural Gas and Diesel Transit Bus Emissions: Review and
Recent Data," Society of Automotive Engineers, SAE Technical Paper 973203, 1997. (http ://papers. sae .org/973203).
129	McConnick, R„ Graboski, M„ Alleman, T„ Herring, A. et al., "In-Use Emissions from Natural Gas Fueled Heavy-
Duty Vehicles," Society of Automotive Engineers, SAE Technical Paper 1999-01-1507, 1999.
(http://papers.sae.org/1999-01-1507).
1311 Clark, et al., "Effects of Average Driving Cycle Speed on Lean-Burn Natural Gas Bus Emissions and Fuel
Economy," SAE Technical Paper 2007-01-0054, 2007
131	"Heavy-Duty Highway Compression-Ignition Engines and Urban Buses ~ Exhaust Emission Standards," Emission
Standards Reference Guide, USEPA. Last Updated: 5 July 2012, Accessed: 20 July 2012.
132	US EPA OTAQ, 2012. (http://iaspub.epa.gov/otaqpub/pubsearch.isp).
328

-------
133	40 CFR 86.105-94. "Clean-fuel fleet emission standards for heavy-duty engines." Code of Federal Regulations.
134	Search for compliance documents in the "Heavy-Duty Highway Compression-Ignited Engines" category on EPA's
Transportation and Air Quality Document Index System (DIS), Last Updated: 10 August 2012, Accessed: 10 August
2012. (http ://iaspub .epa.gov/otaqpub/pubsearchj sp).
135	Hajbabaei, M., G. Karavalakis, K. C. Johnson, L. Lee and T. D. Durbin (2013). Impact of natural gas fuel
composition on criteria, toxic, and particle emissions from transit buses equipped with lean burn and stoichiometric
engines. Energy, 62 (0), 425-434. DOI: http://dx.doi.Org/10.1016/i.energy.2013.09.040.
136	Yoon, S.; Collins, J.; Thiruvengadam, A.; Gautam, M.; Herner, J.; Ayala, A. Criteria pollutant and greenhouse gas
emissions from CNG transit buses equipped with three-way catalysts compared to leanburn engines and oxidation
catalyst technologies, Journal of the Air & Waste Management Association, 2013, 63:8, 926-933,
http://dx.doi.org/10.1080/10962247.2013.80017Q.
137	Jeon, J., J. T. Lee and S. Park (2016). Nitrogen Compounds (NO, N02, N20, and NH3) in NOx Emissions from
Commercial EURO VI Type Heavy-Duty Diesel Engines with a Urea-Selective Catalytic Reduction System. Energy &
Fuels, 30 (8), 6828-6834. DOI: 10.1021/acs.energyfuels.6b01331.
138	Majewski, W. A. (2005). Selective catalytic reduction. Ecopointlnc. Revision.
139	Preble, C. V., R. A. Harley and T. W. Kirchstetter (2019). Control Technology-Driven Changes to In-Use Heavy-
Duty Diesel Truck Emissions of Nitrogenous Species and Related Environmental Impacts. Environ Sci Technol, 53
(24), 14568-14576. DOI: 10.1021/acs.est.9b04763.
140	Preble, C. V., T. R. Dallmann, N. M. Kreisberg, S. V. Hering, R. A. Harley and T. W. Kirchstetter (2015). Effects of
Particle Filters and Selective Catalytic Reduction on Heavy-Duty Diesel Drayage Truck Emissions at the Port of
Oakland. Environ Sci Technol, 49 (14), 8864-8871. DOI: 10.1021/acs.est.5b01117.
141	Kean, A. J., D. Littlejohn, G. A. Ban-Weiss, R. A. Harley, T. W. Kirchstetter and M. M. Lunden (2009). Trends in
on-road vehicle emissions of ammonia. Atmospheric Environment, 43 (8), 1565-1570. DOI:
10.1016/j.atmosenv.2008.09.085.
142	Haugen, M. J., G. A. Bishop, A. Thiruvengadam and D. K. Carder (2018). Evaluation of Heavy- and Medium-Duty
On-Road Vehicle Emissions in California's South Coast Air Basin. Environ Sci Technol, 52 (22), 13298-13305. DOI:
10.1021/acs.est.8b03994.
143	Bishop, G. A., M. J. Haugen, B. C. McDonald and A. M. Boies (2022). Utah Wintertime Measurements of Heavy-
Duty Vehicle Nitrogen Oxide Emission Factors. Environ Sci Technol, 56 (3), 1885-1893. DOI:
10.1021/acs.est. Ic06428.
144	Wang, X., A. Khlystov, K.-F. Ho, D. Campbell, J. C. Chow, S. D. Kohl, J. G. Watson, S.-c. F. Lee, L.-W. A. Chen,
M. Lu and S. S. H. Ho (2019). Real-World Vehicle Emissions Characterization for the ShingMun Tunnel in Hong
Kong and Fort McHenry Tunnel in the United States. Research Report 199. Health Effects Institute. Boston, MA.
March 2019. https://www.healtheffects.org/publication/real-world-vehicle-emissions-characterization-shing-mun-
tunnel-hong-kong-and-fort.
145	USEPA (2012). Use of Data from "Development of Emission Rates for the MOVES Model, " Sierra Research,
March 3, 2010. EPA-420-R-12-022. Office of Transportation and Air Quality. US Environmental Protection Agency.
Ann Arbor, MI. August, 2012. https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P 1 OOF 1A5.txt.
146	Harvey, C. A., R. J. Garbe, T. M. Baines, J. H. Somers, K. H. Hellman and P. M. Carey (1983). A study of the
potential impact of some unregulated motor vehicle emissions. SAE transactions, 280-289.
147	Livingston, C., P. Rieger and A. Winer (2009). Ammonia emissions from a representative in-use fleet of light and
medium-duty vehicles in the California South Coast Air Basin. Atmospheric Environment, 43 (21), 3326-3333. DOI:
10.1016/j.atmosenv.2009.04.009.
329

-------
148	Thiruvengadam, A., M. Besch, D. Carder, A. Oshinuga, R. Pasek, H. Hogo and M. Gautam (2016). Unregulated
greenhouse gas and ammonia emissions from current technology heavy-duty vehicles. Journal of the Air & Waste
Management Association, 66 (11), 1045-1060. DOI: 10.1080/10962247.2016.1158751.
149	CARB (2021). EMFAC2017 Volume III - Technical Documentation. VI.0.1. April, 2021.
https://ww2.arb.ca.gov/sites/default/files/2021-08/emfac2021 technical documentation april2021.pdf.
150	Zhu, H., C. McCaffery, J. Yang, C. Li, G. Karavalakis, K. C. Johnson and T. D. Durbin (2020). Characterizing
emission rates of regulated and unregulated pollutants from two ultra-low NOx CNG heavy-duty vehicles. Fuel, 277,
118192. DOI: https://doi.Org/10.1016/j.fuel.2020.118192.
151	Jaaskelainen, H. Crankcase Ventilation. DieselNet Technology Guide. www.DieselNet.com. Copyright © Ecopoint
Inc. Revision 2012.12.
152	Zielinska, B.; Campbell, D.; Lawson, D. R.; Ireson, R. G.; Weaver, C. S.; Hesterberg, T. W.; Larson, T.; Davey, M.;
Liu, L.-J. S. 2008. Detailed characterization and profiles of crankcase and diesel particulate matter exhaust emissions
using speciated organics Environ. Sci. Technol. 42(15): 5661-5666.
153	Ireson, R.G., Ondov, J. M., Zielinska, B.,. Weaver, C. S., Easter, M. D., Lawson, D. R., Hesterberg, T. W., Davey,
M. E., Liu, L.-J. S. Measuring In-Cabin School Bus Tailpipe and Crankcase PM2.5: A New Dual Tracer Method,
Journal of the Air & Waste Management Association, 2011, 61:5, 494-503
154	Hill, L. B.; Zimmerman, N. J.; Gooch, J.; A Multi-City Investigation of the Effectiveness of Retrofit Emissions
Controls in Reducing Exposures to Particulate Matter in School Buses. January 2005. Clean Air Task Force.
155	Title 40: Code of Federal Regulations. Part 86- Protection of Environment. Control of Emissions from New and In-
Use Highway Vehicles and Engines. 86.1810-01 Subpart S—General Compliance Provisions for Control of Air
Pollution From New and In-Use Light-Duty Vehicles, Light-Duty Trucks, and Complete Otto-Cycle Heavy-Duty
Vehicles. General standards; increase in emissions; unsafe conditions; waivers.
156	40 CFR 86.004-11. "Control of Emissions from New and In-Use Highway Vehicles and Engines." Code of Federal
Regulations.
157	Khalek, I. A.; Bougher, T. L; Merrit, P. M.; Phase 1 of the Advanced Collaborative Emissions Study. CRC Report:
ACES Phase 1, June 2009.
158	Clark, N. McKain, D., Barnett, R., Wayne, S., Gautam, M., Thompson, G., Lyons, D. "Evaluation of Crankcase
Emissions Abatement Device," August 8, 2006. West Virginia University.
159	Clark, N., Tatli, E., Barnett, R., Wayne, W. et al., "Characterization and Abatement of Diesel Crankcase Emissions,"
SAE Technical Paper 2006-01-3372, 2006, doi: 10.4271/2006-01-3372.
160	Hare, C. T.; Baines, T. M.; Characterization of Diesel Crankcase Emissions. Society of Automotive Engineers, Off-
Highway Vehicle Meeting and Exhibition. MECA, Milwaukee. 1977.
161	Kalayci, Veli. "Spiracle™ Crankcase Filtration Systems: Technical Article" . Donaldson Company, Inc. January
2011.
162	{USEPA, 2020 #3833}
163	Gerhardt, M. J., D. Sonntag, G. Brown, B. Caldwell, A. Cullen, C. Hart and S. Ludlam (2020). Crankcase
Emissions for MY 2007+ Heavy-Duty Diesel Trucks. MOVES Review Workgroup, Ann Arbor, MI. October 14, 2020.
https://www.epa.gov/moves/october-2020-moves-model-review-work-group-meeting-materials.
164	Clark, N. N., et al. (2017). Pump-to-Wheels Methane Emissions from the Heavy-Duty Transportation Sector.
Environ Sci Technol, 51 (2), 968-976. DOI: 10.1021/acs.est.5b06059.
165	McClenny, W. A. (2000). Recommended Methods for Ambient Air Monitoring of NO, N02, NOy, and Individual
NOz Species. EPA/600/R-01/005. National Exposure Research Laboratory, US EPA. September 2000.
330

-------
166	Seinfeld, J. H. and S. N. Pandis (2012). Atmospheric chemistry and physics: from air pollution to climate change,
John Wiley & Sons.
167	Dunlea, E. J., S. C. Herndon, D. D. Nelson, R. M. Volkamer, F. San Martini, P. M. Sheehy, M. S. Zahniser, J. H.
Shorter, J. C. Wormhoudt, B. K. Lamb, E. J. Allwine, J. S. Gaffney, N. A. Marley, M. Grutter, C. Marquez, S. Blanco,
B. Cardenas, A. Retama, C. R. Ramos Villegas, C. E. Kolb, L. T. Molina and M. J. Molina (2007). Evaluation of
nitrogen dioxide chemiluminescence monitors in a polluted urban environment, Atmos. Chem. Phvs., 1 (10), 2691-
2704. DOI: 10.5194/acp-7-2691-2007.
168	Kurtenbach, R., K. H. Becker, J. A. G. Gomes, J. Kleffmann, J. C. Lorzer, M. Spittler, P. Wiesen, R. Ackermann, A.
Geyer and U. Piatt (2001). Investigations of emissions and heterogeneous formation of HONO in a road traffic tunnel.
Atmospheric Environment, 35 (20), 3385-3394. DOI: http://dx.doi.org/10.1016/S1352-2310(01)00138-8.
169	Khalek, I. A., M. G. Blanks and P. M. Merritt (2013). Phase 2 of the Advanced Collaborative Emissions Study.
CRC Report: ACES Phase 2. Coordinating Research Council, Inc. & Health Effects Institute. November 2013.
1711 Kim, J. H„ M. Y. Kim and H. G. Kim (2010). N02-Assisted Soot Regeneration Behavior in a Diesel Particulate
Filter with Heavy-Duty Diesel Exhaust Gases. Numerical Heat Transfer, Part A: Applications, 58 (9), 725-739. DOI:
10.1080/10407782.2010.523293.
171	Kramer, L. J., L. R. Crilley, T. J. Adams, S. M. Ball, F. D. Pope and W. J. Bloss (2020). Nitrous acid (HONO)
emissions under real-world driving conditions from vehicles in a UK road tunnel. Atmos. Chem. Phvs., 20 (9), 5231-
5248. DOI: 10.5194/acp-20-5231-2020.
172	Liang, Y., Q. Zha, W. Wang, L. Cui, K. H. Lui, K. F. Ho, Z. Wang, S.-c. Lee and T. Wang (2017). Revisiting
nitrous acid (HONO) emission from on-road vehicles: A tunnel study with a mixed fleet. Journal of the Air & Waste
Management Association, 67 (7), 797-805. DOI: 10.1080/10962247.2017.1293573.
173	Xu, Z., T. Wang, J. Wu, L. Xue, J. Chan, Q. Zha, S. Zhou, P. K. K. Louie and C. W. Y. Luk (2015). Nitrous acid
(HONO) in a polluted subtropical atmosphere: Seasonal variability, direct vehicle emissions and heterogeneous
production at ground surface. Atmospheric Environment, 106, 100-109. DOI:
https://doi.Org/10.1016/i.atmosenv.2015.01.061.
174	Trinli, H. T„ K. Imanishi, T. Morikawa, H. Hagino and N. Takenaka (2017). Gaseous nitrous acid (HONO) and
nitrogen oxides (NOx) emission from gasoline and diesel vehicles under real-world driving test cycles. Journal of the
Air & Waste Management Association, 67 (4), 412-420. DOI: 10.1080/10962247.2016.1240726.
175	Rappengliick, B., G. Lubertino, S. Alvarez, J. Golovko, B. Czader and L. Ackermann (2013). Radical precursors
and related species from traffic as observed and modeled at an urban highway junction. Journal of the Air & Waste
Management Association, 63 (11), 1270-1286. DOI: 10.1080/10962247.2013.822438.
176	USEPA (2012). Use of Data from "Development of Emission Rates forthe MOVES Model," Sierra Research,
March 3, 2010. Assessment and Standards Division. Office of Transportation and Air Quality. Ann Arbor, MI. April
2012. https://nepis.epa.gov/Exe/ZvPDF.cgi?Dockev=P100FlA5.pdf
177	USEPA (2009). Development of Emission Rates for Hea\>y-Duty Vehicles in the Motor Vehicle Emissions Simulator
(DraftMOVES2009). EPA-420-P-09-005. Office of Transportation and Air Quality. US Enviromnental Protection
Agency. Ann Arbor, MI. August, 2009. https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P10050CK.txt.
178	Belser, Evan (2020) Re: Tampered Diesel Pickup Trucks: A Review of Aggregated Evidence from EPA Civil
Enforcement Investigations. Letter to Jason E. Sloan, et al, November 20, 2020
https://www.epa.gov/sites/default/files/2021-01/documents/epaaedletterreportontampereddieselpickups.pdf
179	U.S. Census Bureau. Vehicle Inventory and Use Sun'ey. Service Sector Statistics Division, Transportation
Characteristics Branch, https://www.census.gov/programs-survevs/vius.html.
1811 Zhou, Lei. Revision of Heavy Heavy-Duty Diesel Truck Emission Factors and Speed Correction Factors. California
Air Resources Board, Sacramento. October 2006.
331

-------
181	USEPA (2022) Final Rule and Related Materials for Control of Air Pollution from New Motor Vehicles: Heavy-
Duty Engine and Vehicle Standards, Regulatory Impact Analysis. EPA-420-R-22-035, December 2022. Chapter 1.
182	Illinois Environmental Protection Agency. Effectiveness of On-Board Diagnostic I/M Testing: Report to the General
Assembly, Response to Public Act 92-0682. Bureau of Air, Springfield, IL. September 2003. Page 21.
183	Manufacturers of Emission Controls Association. Technology Details - Catalytic Converters - SCR System.
184	Song, Qingwen, and Zhu, George. Model-based Closed-loop Control of Urea SCR Exhaust Aftertreatment System
for Diesel Engine. SAE 2002-01-287. Society of Automotive Engineers, Warrendale, PA.
185	Darlington, T., Dennis Kahlbaum and Gregory Thompson. On-Road NOx Emission Rates from 1994-2003 Heavy-
Duty Diesel Trucks. SAE 2008-01-1299. Society of Automotive Engineers, April 2008.
186	Preble, C. V., T. R. Dallmann, N. M. Kreisberg, S. V. Hering, R. A. Harley and T. W. Kirchstetter (2015). Effects of
Particle Filters and Selective Catalytic Reduction on Heavy-Duty Diesel Drayage Truck Emissions at the Port of
Oakland. Environ Sci Technol, 49 (14), 8864-8871. DOI: 10.1021/acs.est.5b01117.
187	Bishop, G. A., R. Hottor-Raguindin, D. H. Stedman, P. McClintock, E. Theobald, J. D. Johnson, D.-W. Lee, J.
Zietsman and C. Misra (2015). On-road Heavy-duty Vehicle Emissions Monitoring System. Environ Sci Technol, 49
(3), 1639-1645. DOI: 10.1021/es505534e.
188	ARB (2015). Evaluation of Particulate Matter Filters in On-Road Heavy-Duty Diesel Vehicle Applications.
California Air Resources Board. May 8, 2015. https://ww2.arb.ca.gov/sites/default/files/2020-08/dpfeval O.pdf.
332

-------