Exhaust Emission Rates for Heavy-Duty
Onroad Vehicles in MOVES3
A rnA United States
Environmental Protection
^1 M ^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 MOVES3
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
NOTICE
&EPA
United States
Environmental Protection
Agency
EPA-420-R-20-018
November 2020
-------
Table of Contents
1 Introduction 1
1.1 Pollutant 1
1.2 Emission Process 2
1.2.1 Running Exhaust 2
1.2.2 Start Exhaust 2
1.2.3 Extended Idle and Auxiliary Power Exhaust 3
1.2.4 Crankcase Exhaust 3
1.2.5 Evaporative and Brake and Tire Wear Emissions 4
1.3 Fuel Type 4
1.4 Regulatory Class 4
1.5 Model Year Groups 6
1.6 Operating Modes 6
1.7 Vehicle Age 11
1.8 Updates for MOVES3 12
2 Heavy-Duty Diesel Exhaust Emissions 14
2.1 Running Exhaust Emissions 14
2.1.1 Nitrogen Oxides (NOx) 14
2.1.2 Particulate Matter (PM2.5) 47
2.1.3 Total Hydrocarbons (THC) and Carbon Monoxide (CO) 67
2.1.4 Energy 80
2.1.5 Evaluation of Fleet-average Running Rates with Real-World Measurements 90
2.2 Start Exhaust Emissions 91
2.2.1 THC, CO, and NOx 91
2.2.2 Particulate Matter (PM2.5) 98
2.2.3 Adjusting Start Rates for Soak Time 100
2.2.4 Start Energy Rates 110
2.3 Extended Idling Exhaust Emissions 112
2.3.1 1960-2006 Model Years 113
2.3.2 2007-2060 Model Years 116
2.3.3 Model Year Trends 126
2.3.4 Extended Idle Energy Rates 128
-------
2.4 Auxiliary Power Unit Exhaust 129
2.5 Glider Vehicle Emissions 134
3 Heavy-Duty Gasoline Exhaust Emissions 136
3.1 Running Exhaust Emissions 136
3.1.2 Particulate Matter (PM2.5) 153
3.1.3 Energy 160
3.2 Start Emissions 165
3.2.1 THC, CO, and NOx 166
3.2.2 Particulate Matter (PM2.5) 181
3.2.3 Soak Time Adjustments 182
3.2.4 Start Energy Rates 183
4 Heavy-Duty Compressed Natural Gas Exhaust Emissions 185
4.1 Running Exhaust Emission Rates 186
4.1.1 1960-2009 Model Years 187
4.1.2 2010-2060 Model Years 197
4.1.3 Model Year Trends 198
4.2 Start Exhaust Emission Rates 203
4.3 Ammonia Emissions 203
5 Heavy-Duty Crankcase Exhaust Emissions 204
5.1 Background on Heavy-Duty Diesel Crankcase Emissions 204
5.2 Modeling Crankcase Emissions in MOVES 205
5.3 Heavy-Duty Diesel Crankcase Emissions 206
5.3.1 1960-2007 Model Years 206
5.3.2 2007-2009 Model Years 208
5.3.3 2010-2060 Model Years 214
5.3.4 Glider Crankcase Emissions 221
5.4 Heavy-Duty Gasoline and CNG Crankcase Emissions 222
6 Nitrogen Oxide Composition 224
6.1 Heavy-Duty Diesel 225
6.2 Heavy-Duty Gasoline 225
6.3 Compressed Natural Gas 226
7 Appendices 227
Appendix A Calculation of Accessory Power Requirements 228
ii
-------
Appendix B Tampering and Mal-maintenance for Diesel Running Exhaust 229
Appendix C Tampering and Mal-maintenance for MY 2007 and Later Diesel Extended Idle. 244
Appendix D Pre-2007 Model Year Extended Idle Data Summary 246
Appendix E Developing Pre-2007 Model Year HD Diesel PM2.5 Emission Rates for Missing
Operating Modes 250
Appendix F Heavy-Duty Gasoline Start Emissions Analysis Figures 251
Appendix G Selection of Fixed Mass Factor (fscaie) values for MY 2010+ Heavy-Duty Vehicles
256
Appendix H THC and CO Emisssion rates from 2010 and Later Model Year Heavy-duty
Vehicles from the HDIUT 268
Appendix I Analysis of 2010 and Later Model Year Heavy-duty Gasoline Emission Rates 273
Appendix J Comparing Glider Vehicle and MOVES Model Year 2000 Heavy Heavy-Duty
Emission Rates 282
Appendix K PM Composition Measurements from Auxiliary Power Units 287
Appendix L Crankcase Emissions from Glider Trucks 288
8 References 291
iii
-------
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
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
iv
-------
H-GAC Houston-Galveston Area Council
HHD Heavy-Heavy-Duty Class 8 Trucks (GVWR > 33,000 lbs)
HHDD Heavy Heavy-Duty Diesel
HNO2 nitrous acid (HONO)
HP horsepower
hr hour
HV heating value
H2O water
I/M Inspection and Maintenance program
IUVP 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)
MOBILE6 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
NonECnonS04PM 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
NO2 nitrogen dioxide
NREL National Renewal Energy Laboratory
NTE Not-to-Exceed
NYSDEC New York Department of Environmental Conservation
v
-------
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 pm
PM10
particles of particulate matter with aerodynamic diameters <10 pm
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
vi
-------
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 MOVES3.
Heavy-duty vehicles in MOVES are defined as any vehicle with a Gross Vehicle Weight Rating
(GVWR) above 8,500 lbs.
Emission rates for criteria pollutants (THC, CO, NOx, and PM2.5) 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.8
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 MOVES3.
Sections 2 through 4 document the tailpipe exhaust emission rates for heavy-duty diesel, heavy-
duty gasoline, and heavy-duty CNG vehicles. Section 5 documents the crankcase emission rates
used for each fuel type of heavy-duty vehicles. Section 6 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 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
1
-------
particulate matter with mean aerodynamic diameter less than 2.5 microns (PM2.5), 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
From NOx, MOVES estimates NO, NO2, and HONO emissions as documented in Section 6.
Nitrous Oxide (N2O) emissions from heavy-duty vehicles remained unchanged from MOVES2009
and later versions and are documented in the MOVES3 Greenhouse Gas and Energy Report.3
MOVES estimates CO2 emissions from the energy rates documented in this report, using
conversion factors documented in the Greenhouse Gas and Energy Report. Ammonia (NH3)
emission rates for heavy-duty vehicles were developed for MOVES2010 and are documented in a
separate report,4 except CNG rates which are documented here in Section 4.3.
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. MOVES3 includes off-network idling as a new source of vehicle activity, but it is not a
new emission process. Off-network idling is classified as idle mode activity within the running
emission process. The running process is further delineated by 23 operating modes as discussed in
Section 1.6.
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.
2
-------
g
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. Equation 1-2 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.
Documentation of the extended idle and auxiliary power exhaust emissions for heavy-duty diesel
trucks are in Section 2.3.
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
5. 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. The emission rates for all four pollutants for all
heavy-duty source types and fuels are discussed in Section 5.
3
-------
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.5 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.5 Brake and tire wear emission rates from heavy-duty
vehicles are discussed in the Brake and Tire Wear Report.10
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
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 and electricity fuel types for light-duty vehicles, but
these fuel types are not available to be modeled for heavy-duty vehicles.9
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 two of 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.6 Urban
4
-------
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."7.
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 combination short-haul source types. As discussed in Section
2.5, the emissions are equivalent to MY 2000 HHD diesel vehicles.
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.9 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.
Regulatory Class
Description
regClassName
regClassID
Gross Vehicle
Weight
Rating
(GVWR) [lb.]
Possible Source Types
(sourceTypelD)
Light Heavy-Duty
Class 2b and 3
trucks
LHD2b3a
41
8,501 -
14,000
Passenger Trucks (31), and
Light-Commercial Trucks
(32), Buses (41, 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 Bus7
48
> 33,000
Transit Bus (42)
Gliders (Class 8
Trucks)
Glider Vehicles
49
> 33,000
Combination Trucks (61, 62)
a In MOVES3, we consolidated the MOVES2014 regulatory classes LHD<10K (regClassID 40) and LHD<14K
(regClassID 41) into LHD2b3 (regClassID 41). InMOVES2014, LHD<10K was used to model Class 2b trucks among
the light-duty source types (31,32), and LHD<14K was used to model Class 2b and 3 trucks among the heavy-duty
source types (41,42 ,43, 51 ,52 ,53, 54). InMOVES2014, the emission rates inLHD<10K and LHD<14K were
intended to be the same. However, separate regulatory classes were needed because MOVES2014 assigned the fscaie-
power scaling factor by sourcetype only, so the fscaie values (stored in the sourceUseTypePhysics table) had to differ
5
-------
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.
Running Emissions
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).
between the 2b trucks mapped to the light-duty sourcetypes (passenger and light-commercial trucks) and the 2b3 trucks
among the heavy-duty sourcetypes (single unit trucks, refuse, motorhomes and buses). Because they had different fscaies
values, they needed different emission rates. However, in MOVES3, we can now specify different road-load
coefficients, vehicle weights and fscaie values by both source type and regulatory class. Thus, we can model LHD2b3 in
both the light-duty (e.g., Source type 31, passenger trucks) and heavy-duty sourcetypes (e.g., Sourcetype 52, single-unit
short-haul trucks) using the same fscaievalue and the same emission rates, such that the LHD<10K regulatory class is no
longer needed. The vehicles that were classified as LHD<10K in MOVES2014 are now classified as LHD2b3 in
= SHO
23
x (Operating Mode Distribution
Equation 1-3
i=l
x Emission Ratet)
MOVES3.
6
-------
Start Emissions
= Starts
8
x ^\operating Mode Distributiont Equation 1 4
i=1
x Emission Ratet)
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,8 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.
Avt + Bvt + Cvf + m • vt(at + a • sindt) „ 1 n
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 6t = 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 fSCaiewa^ue is used, individual vehicles assigned to the same STP-
7
-------
defined operating mode bin are producing the same absolute tractive power, regardless of
differences in their individual source masses.
Avt + Bv} + Cvf + m ¦ vt(at + a ¦ sin0t) ^ „
STPt = —- - - —— — Equation 1-6
fscale
Where:
STPt = the scaled tractive power at time t [scaled kW or skW]
fscale = 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 (vt) 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.9
The equation for STP is generalized below in Equation 1-7, with units in scaled kW or skW:
STP= —Equation 1-7
^scale
Where: Paxie is the power demand at the axle for the heavy-duty truck.
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).
8
-------
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 fscaie 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
fscaie 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 MOVES3 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, we
revised the fscaie values for MY 2010 and later to provide more resolution in the fscaie by regulatory
class. Derivation of the new fscaie values is described in Appendix G.
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 fscaie values. For example, the OpMode 14 emission rates for MY 2012 MHD (fscaie = 7) cannot be directly
compared to the same OpMode rates from MY 2009 MHD (fscaie = 17.1) or MY 2012 HHD (fscaie =10). That is because
data assigned to an OpMode based on different fscaie 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.
9
-------
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(6t) < -2.0 OR
[at +g-sin(6t) < -1.0
AND
at-1 +g-sin(Ot-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 < STP,
1 < vt< 25
21
Coast
STP^O
25 < vt< 50
22
Cruise/Acceleration
0 < STP,< 3
25 < vt< 50
23
Cruise/Acceleration
3 < STP,< 6
25 < vt< 50
24
Cruise/Acceleration
6 < STP,< 9
25 < vt< 50
25
Cruise/Acceleration
9 < STP(< 12
25 < vt< 50
27
Cruise/Acceleration
12
-------
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
11
-------
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.
Energy rates are stored in the "EmissionRate" table, where are not distinguished by age. This table
includes 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, however, the documentation of heavy-
duty nitrous oxide3 and ammonia4 and tire and brake wear10 emission rates are documented in
separate MOVES reports.
1.8 Updates for MOVES3
This report documents the heavy-duty emission rates in MOVES3, which was updated from both
MOVES201411 and the MOVES version used to support the Heavy-duty Greenhouse Gas Phase 2
rulemaking12.
In summary, the following emission rates were updated in MOVES3 from MOVES2014:
• Heavy-Duty Diesel Vehicles
o NOx, PM2.5, THC, CO, and running exhaust emission rates for model year (MY) 2010
and later (Sections 2.1.1.5, 2.1.2.2, and 2.1.3.2)
o Energy rates for MY 2010 and later vehicles, including the impact of the Heavy-Duty
Greenhouse Gas Phase 2 rulemaking (Sections 2.1.4.3 and 2.1.4.3.2)
o Elemental carbon fraction of PM2.5 exhaust emissions from pre-2007 MY (Section
2.1.2.1.8)
o THC, CO, NOx, and PM2.5 start exhaust emission rates for MY 2010 and later (Section
2.2)
o Extended idle and auxiliary power unit (APU) exhaust THC, CO, NOx, and PM2.5
emission rates for all model years (Section 2.3 and 2.4)
o Running, start, extended idle, and APU exhaust emissions from glider trucks (Section
2.5)
o Updated crankcase emissions for 2007 and later vehicles (Sections 5.3.2, 5.3.3, and
5.3.4)
• Heavy-Duty Gasoline Vehicles
o THC, CO, NOx, and PM2.5 running exhaust emission rates for MY 2010 and later
(Sections 3.1.1.2.1 and 3.1.2.2)
o Energy rates for MY 2010 and later, including the impact of the Heavy-Duty
Greenhouse Gas Phase 2 rulemaking (sections 3.1.3.2 through 3.1.3.2.2)
12
-------
• Heavy-Duty Compressed Natural Gas (CNG) Vehicles (Section 4)
o THC, CO, NOx, and PM2.5 running exhaust emission rates for MY 2007 and later
o Emissions for vehicles using CNG fuel are now calculated for most heavy-duty source
types, not just transit buses
The significant updates made to MOVES3 since the MOVES2014 release were peer-reviewed
under EPA's peer review guidance.13 In 2017, we peer-reivewed an internal version referred to as
"MOVES201X" that included the updates to diesel running and extended idle emission rates and
updated energy consumption rates due to the incorporation of the Heavy-duty Greenhouse Gas
Phase 2 rulemaking.14 In 2019, we conducted a peer-review of the updates made for the
preliminary Cleaner Trucks Initiative (CTI) analysis, which included further updates to the heavy-
duty diesel emission rates, incorporation of glider vehicles, and updated heavy-duty gasoline and
CNG emission rates.15 In 2020, we conducted a peer-review of the updated diesel crankcase
emission rates incorporated into MOVES3.16 Materials from each peer review, including peer-
review comments and EPA responses are located on the EPA's science inventory webpage.
13
-------
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 Exhaust 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).
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 Decree17, 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
(FEL)
(or 1.25 standard or FEL, when FEL > 1.50 g/bhp-
hr)
2007-2009
1.21,2
2010+
0.2
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.18
14
-------
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.19 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
Consent Decree Testing. These data were conducted by West Virginia University using the
Mobile Emissions Measurement System (MEMS).20,21,22 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.23 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 publically available on EPA's website.24
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, 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.25 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 plan to expand
the characterization of the MY 2010+ HDIUT data set, in a future update to this report, by adding
15
-------
summary information on vehicle age distribution, odometer reading, idling time duration, and
operating mode based time and miles travelled.
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 20 1 0.26 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
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+2
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 2-10.
2 New data used to update heavy-duty diesel MY 2010 and beyond rates in MOVES3.
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 and later model year emission rates as discussed in Section 2.1.1.4.
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
16
-------
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.
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 0e^and torque zengfrom the ECU
were used to determine engine power Peng, as shown in Equation 2-1.
17
-------
Kng - 0)engTeng 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
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.
18
-------
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.
Oh
Low
Engine Access.
Alternator
Engine Access.
Alternator
Engine Access.
Alternator
CS
Air Compress
Air Compress
CS
O
O)
£
DO
Cooling Fan
Air Cond.
Cooling Fan
Air Cond.
Air Cond.
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."27
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 Pioss.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).
19
-------
Engine Power Level
(of rated power)
HHD
MHD
LHD1
(pre-2010)
LHD1
(2010+)
Urban
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, 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 rjdnveiine varies with engine speed, vehicle speed, and
vehicle power requirements. Using sources available in the literature, 28>29>30>31>32>33>34>35>36 We
estimated an average value for driveline efficiency. Table 2-5 summarizes our findings.
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 ~ V drive line (Peng Ploss,acc) Equation 2-2
20
-------
Finally, we scaled the axle power using Equation 2-3, and the STP-scaling factors fscaie presented in
Table 1-3 for every second of data.
STP=—Equation 2-3
^scale
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
data and the Houstan Drayage Program data.37,38 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.b The
Houston Drayage Program was not used to estimate any of the emission rates in MOVES.
Table 2-10 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, 2003-2006c, 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
b This analysis is described in more detail in the MOVES2014 heavy-duty exhaust report However, since we have
updated the fscaievalues used to assign STP bins, the MOVES3 rates are no longer directly comparable (see Presentation
by Choi et al. (201237).
c 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.
21
-------
emission rate for each stratification group (i.e., model year group, regulatory class, and operating
mode bin).
Where:
nj = the number of 1-Hz data points (for a given operating mode bin) for each vehicle j,
rheh = the total number of vehicles,
f/,jj = the emission rate of pollutant p for vehicle j at second i,
j = the mean emission rate (meanBaseRate) for pollutant p (for a given model year group,
regulatory class 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
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.
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. d 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, the 2009-
and-earlier LHD2b3 and LHD45 regulatory classes were assigned the same emission rates as the
LfflKlOK in MOVES2014.
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
^•veh
Equation 2-4
2.1.1.4.1 LightHeavy-Duty Class 2b3 and Classs 4&5 Trucks
2.1.1.4.2 High-Power Operating Modes
d This is consistent with the 2.06 fscaie used to develop LHD2b3 and LHD45 emission rates in MOVES2010.
22
-------
(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.
(ST PopModelD 40 \
——— I 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.
23
-------
4000
3500
„3000 H
£
GO
V2500 -
b.
X
2 2000 -
flj
ai
! 1500
1000
500
* * »
-i 1—i—i 1—i—i 1—i 1—i—i 1—i—i 1—i 1—i—i 1—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 2-1 NOx Emissions by Operating Mode from HHD Trucks for Model Year 2002. Error Bars represent
the 95 percent confidence interval of the Mean
6000
5000
"Si 4000
bQ
¦«-»
* 3000
m
ai
2000
1000
0
¦ MHD
* Bus
~ HHD
~
~ ¦
A
~
~ ¦
~
A ¦
i 4
~ ¦
. I -
~
¦ Ai,
T 1 I 1 1 1 1 1 m I 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 2-2 NOx Emissions by Operating Mode from MHD, HHD, and Urban Bus Regulatory Classes for Model
Year 2002.
24
-------
2.1.1.4.3 Missing Regulatory Class and Model Year Combinations
For regulatory class and model year combinations with missing data, we set the emissions equal to
regulatory class/model year combinations which had equivalent emission standards or
proportionally adjusted the existing emissions data using ratios of certification data or vehicle
emission standards as specified in Table 2-1. For HHD model year groups 1988-1989 and 1990, we
increased the 1991-1997 model year group emission rates by a factor proportional to the increase of
the certification levels as analyzed for MOBILE6.39 On average, the MY 1990 and MY 1988-1989
rates are 1.055 times and 1.367 times the baseline rates of MY 1991-1997, respectively. We
applied the 1988-1989 emission rates to model years 1987 and earlier. For MHD vehicles from MY
1960-1997, we used the HHD emission rates.
For model year 1998, data existed for HHD trucks but not for buses. In these cases, we calculated
the rates for Urban Buses by multiplying the Urban Bus emission rates for 1999-2002 by the ratio
of HHD emission rates between the 1998 and the 1999-2002 model year groups, as shown in
Equation 2-6.
HHD rates1998
Urban Bus rates1998 = x Urban Bus rates1999_2002 Equation 2-6
HHD rates1999_2002
For LHD2b3 and LHD45 vehicles, no data were available for the pre-2003 model year vehicles.
For MY 1999-2002, we analyzed MHD engine data using an fscaie of 2.06 as appropriate for LHD
vehicles (See Table 1-3). For MY 1998, we copied the LHD emission rates from MY group 1999-
2002. We also used thse rates as base rates to back-cast emission rates for the 1991-1997 model
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 2-10 provides a summary of the assumptions used to estimate emission rates for regulatory
class-model year groups with missing data.
2.1.1.4.4 Defeat Device and Low-NOx Rebuilds for 1991-1998 Model Year
HHD and MHD
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.17 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
25
-------
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.
Where:
operating
modes (OM)
21-30
rreflash,91-98,27 — r91-98,16
^reflash, 91-98, OMx — ^re flash, 91-98,27
r1999,27 \
^1999,16/
^91—98, OMx
r91-98,27
Equation 2-7
Equation 2-8
_ _ (r1999,37\
rreflash,91-98,27 ~ r91-98,16 I ^ I Equation 2-9
\'1999,16/ n
(^91-98,OMx \
- I Equation 2-10
'91-98,37 /
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 2008.40 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.
Where:
operating
modes (OM)
31-40
26
-------
2540
2520
2500
2480
2460
tit)
w 2440
g 2420
H 2400
2380
2360
2340
0"(
•O-G..
-O "G.\
>"G"O*©"0..(
>-o-0"Glodel 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 Rule107 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 Rule107 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
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.38
2.1.1.4.6
2007-2009 Model Year LHD45 and LHD2b3
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 standardin 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
27
-------
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 emission rates, because only one
vehicle was tested. Rather, adjustments were made to the MOVES2O1O0 MY 2003-2006 model
year group to develop emission rates for this model year group and regulatory class. During PM
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.
LNT NOx emissions
Baseline LHD2b3 (2003 — 2006) NOx emissions
(LNT normal emissions \
= (normal op. frequency) x — ;—; Equation
\ baseline emissions J 2-11
/baseline emissions\
+ (DPF req. frequency) x ;—;—
\ baseline emission J
= (0.90) x (0.10) + (0.10) x (1) = 0.19
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.
e In MOVES2014, we updated the diesel NOx emission rates for 2003-2006 based on the HDIUT program.
28
-------
2007 — 2009 LHD2b3 NOx emissions
2003 — 2006 LHD2b3 NOx emissions
( LNT NOx emissions \
= (LNT market share) ————, ITI^, „ :—:
\2003 — 2006 LHD2b3 NOx emissionsJ
( 2007 — 2009 emission standerds \
+ (non — LNT market share) ———¦—^rrr-rr^ :—: ;—r~
\2003 — 2006 NOx emissions standerds/
= (0.25) x (0.19) + (0.75) x (0.5) = 0.4225
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.37
2.1.1.5 2010-2060 Model Years
In MOVES3, the MY 2010 and later 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.
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)f 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.
f A Family Emission Limit is the maximum emission level established by a manufacturer for the certification of an
engine family.
29
Equation
2-12
-------
r-.n _ ^sec ERpol,OM,C}FEL}MYG,veh}sec ^ o 10
E Kpol,OM,C,FEL,MYG,veh ~ liquation L-lo
S^Ccount
^iveh ERpol.OM}C,FEL}MYG,veh ^ 0 . ,
ERPoi,om,c,fel,myg = 7 Equation 2-14
vencount
E Rpol,OM,C,MYG,MY
Z/ MY.FEL \ _
I ERp0i,oM,c,FEL,MYG * ^77 ) Equation 2-15
V 2jFel"VCimy,felJ
b hiL
Where:
C
ERx,y,z
FEL
MYG
MY
OM
pol
PVc.MY.FEL
SeC, SGCcount
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
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.
30
-------
0.10
0.08 -
< 0.06 -
3 ;
X
O
Z 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 2-4 NOx Emissions by Operating Mode from HHD Trucks for Model Year 2013. Error Bars represent
the 95 percent confidence interval of the Mean
More details about the selection of the NOx FEL groupings, updated fscaie 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
Upper Limit
(Excluded)
(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
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.
31
-------
100
Qc,Cft
to
Q)
I «
U- o
<1J (N
|S
E O
UJ S
05 >
.O ^
so
80
60
40
20
*\ — v'
Vp.T-0' «
_ ^"¦
+
/
<5* /
y
> gj
0.1 0.2 0.3 0.4
NOx FEL(g/bhp-hr)
0.5
0.6
Figure 2-5 Distribution of N0X 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.
Table 2-7 Number of 2010 and Later Model Year HDIUT Vehicles by NOx FEL Group and Model Year
NOx FE
Group
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. MF1D 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.
32
-------
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.
MOVES OpMode
Figure 2-6 Average LHD NOx Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013 and
the 0.5 NOx FEL for MY 2010-2015. Error Bars are 95% Confidence Intervals of the Mean
33
-------
! MHD FEL 0.20, MY 2010-2011 N=23
i MHD FEL 0,35, MY 2010-2013), N=23
1 MHD FEL 0,50, MY 2010-2013, N=9
.1 II i
I 1
ill
a .1
I ii
Hi
B
a
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
0,20
0,15
bo
' 0,10
o
z
0,05
0.00
O i 11 1/ id 14 lb lb £ 1 11 Id Z4 lb 11 £6 IH iU di is ii i'S iii 4U
MOVES OpMode
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
IHD FEL 0.20, MV 2010-2013, N=?8
IHD FEL 0,35, MY 2010-2015) N=31
IHD FEL 0,50, MY 2010-20151 M=35
I I
J
11
a
ii
_
il a
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. Both of the 2017 peer-reviewers recommended that MOVES3 consider
34
-------
varying emission rates by model year for MY 2010 and later trucks, citing studies50,51,52 that
supported the claim that "since the inception of SCR conrol in 2010, significant advancements in
thermal management, urea dosage strategy, and SCR formulations have taken place."14 For
MOVES3, 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 MOVES3 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 2-23). 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
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 differeces 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
35
-------
MOVES OpVlode
Figure 2-9 NOx emission rates for the MY 2010-2013 and MY 2014-2016 vehicles in the LHD 0.20 NOx FEL
Group
0,030
0,025
m 0,020
oT
+-*
to
02 0.015
c
o
£
o.oio
E
yj
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 OpfVtode
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
¦ i
I I
ii
36
-------
: MY 2010-2013
MY 2014-2015
la I
§1
til!
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.
37
-------
2010 2011 2012 2013 2014 2015 2016 2017 201S
2010 2011 2012 2013 2014 2015 2016 2017 201B
2010 2011 2012 2013 2014 2015 2016 2017 2018
Figure 2-12 Production Volume Contribution of Heavy-Duty Diesel Engine Families by NOx Family Emission
Limit Group for Each Regulatory Class
2010 2011 2012 2013 2014 2015 2016 2017 2018
MHDD
URBU
-"-0.35
-0-0.50
Model Year
Model Year
Model Year
Model Year
LHDD
2.1.1.5.4 Operating Modes and fscaie Values
For the updates to THC, CO, NOx, PM2.5, and energy rates for MY 2010+ HD vehicles, we used
new fscaie 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 fscaie 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.8 Note that the absolute mass/time OpMode-based emissions
rates between the two series based on different fscaie cannot be compared. The main benefit of the
new fscaie 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.
8 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
II.
¦ L H D DO. 20_f s=5.00
BLHDD 0.20 fs=17.1
liJ
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) fscaie 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 H
o
z
0.020 -
0.010 -
0.000
¦ MHDD_0.20_fs=7.00
BMHDD 0.20 fs=17.1
B IB I
1 11 12 13 14 15 16
ill
21 22 23 24 25 27 28 29 30
MOVES OpMode
LLlI
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
li
¦ H H D DO. 20_f s=10.00
BHHDD 0.20 fs=17.1
ill
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) fscaie 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 Incorporation of Tier 3 Standards for LHD2b3
In addition to regulating light-duty vehicles, the Tier 3 vehicle emission standard41 also applies to
chassis-certified light heavy-duty diesel vehicles, which constitute the majority of diesel vehicles
within class 2b and 3 vehicles (LHD2b3).h For these LHD2b3 diesel vehicles, reductions in
emission rates attributable to the Tier 3 standards are applied only to rates for NOx. For THC and
CO emissions, the emission rates measured in the HDIUT program, as described in Section 2.1.3.2,
imply that current levels on the FTP cycle are substantially below the Tier 3 THC and CO
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. Consequently, we assumed that no additional reductions in THC and CO
emissions would be realized through implementation of the Tier 3 standards on LHD diesel
vehicles.
By contrast, the Tier 3 NOx standard results in a reduction of emissions from diesel vehicles in
regulatory classes LHD2b3. Because emission standards tend to impact start and running emissions
differently, 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.38
h To estimate the impact of Tier 3 on the LHD2b3 emission rates, we assumed that all LHD2b3 diesel vehicles and
engines are chassis-certified although there are several 2b3 diesel engines produced by Cummins that are engine-
certified subject to the heavy-duty engine emission standards. All Class 4 and heavier engines are engine-certified and
therefore, not subject to the Tier 3 standards.
40
-------
Table 2-8 Phase-in Assumptions for Tier 3 NOx Standards for Light Heavy-Duty Diesel Vehicles
Model Year
Phase-in
fraction (%)
Reduction in Running
Emission Rate (%)1
Reduction in Start
Emission Rate (%)1
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. An increase in the useful life is
interpreted as an improvement in durability, which is expressed in MOVES through a delay in
deterioration effects. The deterioration effects for the diesel LHD2b3 MY 2010 and later vehicles
are derived from the tampering and mal-maintenance effects discussed in the next section. To
express this effect, the diesel LHD2b3 NOx emission rates estimated for the 0-3 year age group are
replicated to the 4-5 year age group, i.e., the onset of deterioration is delayed two years until the 6-
7 year age group (as shown in Figure 2-16). This effect is phased-in for model years 2018-2021 and
fully implemented for MY2022+. The reduction in grams per mile from the Tier 3 rulemaking
across model years is displayed in Figure 2-19.
8C
o
-c 60
m
m
m
40'
o
tn
£
E 20
HJ
r
model_years
2015-2017
- 2018
- 2019
- 2020
¦ 2021
= 2022+
0 2 4 6 8 10 12 14 16 18 20 22 24
Vehicle age, years
Figure 2-16 NOx Emission Rates for Running Exhaust Operation in a Single Operating Mode (27) vs. Age for
LHD2b3 Diesel Trucks for 2017 through 2022 and Later Model Years
41
-------
2.1.1.6 Tampering and Mal-maintenance
MOVES accounts for the fleet-average increase in emissions with vehicle age. For heavy-duty
diesel vehicles, the increase in aging is modeled through tampering & mal-maintenance (T&M)
adjustment factors. We assume that the T&M adjustment factors are the dominant cause of
emission increases with age for heavy-duty diesel vehicles and that our estimated T&M adjustment
factors also account for the emission increases from normal aging and deterioration of properly
maintained engines and exhaust aftertreatment systems.
Table 2-9 shows the estimated aggregate NOx emissions percent increases due to T&M adjustment
factors by regulatory class and model year group. As described in Appendix B, the T&M
adjustment factors in Table 2-9 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. 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 the warranty period for
the HHD, MHD, and Urban Buses are anticipated to 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,
and the LHD45 and LHD2b3 vehicles in the age 4-5 group.
We account for the emission increases with age by multiplying the MY 2007-2009 LHD emission
rates, and MY 2010 and later emission rates by the corresponding T&M adjustment factors and
scaled aged effects (Equation 7-3). Both the 2007-2009 LHD emission rates and the MY 2010 rates
are assumed to represent zero-mile emission rates. The 2007-2009 LHD emission rates are derived
from scaling emission factors by certification level. The MY 2010 and later heavy-duty emission
rates are based on the HDIUT data which tested well-maintained vehicles. The MOVES emission
rates for regulatory classes with the same zero-mile emission rates (Table 2-10) are different due to
the T&M NOx effects (Table 2-9) and phase-in of T&M effects by age (Table B-3).
Table 2-9 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
4
4
0
2010-2012
56
77
77
2013+
56
58
58
The T&M adjustments for NOx are zero for some of the model year groups because these vehicles
lack the heavy-duty advanced aftertreatment that we assume are most affected by tampering and
malmaintenance. The LHD vehicles have different T&M NOx increases than HHD, MHD, and
Urban Bus vehicles due to the 25 percent assumed penetration of lean NOx trap (LNT)
aftertreatment within the LHD2b3 vehicles beginning in MY 2007-2009, consistent with the
assumptions previously made in Section 2.1.1.4.6. Certification data shows that LNT is not actually
being used in LHD45 vehicles, however, we have not updated this assumption used to derive the
42
-------
T&M adjustment factors on the rates because the resulting LHD45 2007-2009 emission rates
compare well to the HDIUT data for 2007-2009 vehicles.37
The T&M values for model year 2010 and later vehicles account for implementation of OBD. For
LHD2b3 trucks, OBD systems were assumed to be fully implemented in MY 2010. For Class 4
through 8 trucks, (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 2-9 based on the assumptions and
calculations detailed in Appendix B.
Figure 2-17 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
age quicker, with the maximum emission rates at the 4-5 age group, while the Urban Bus do not
reach the useful life until 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 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 2-19, the age 0-3 MY 2009 NOx g/mile
emissions are over 100% higher than the MY 2010 vehicles, while the T&M adjustment factors are
less than 100% (Table 2-9). 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 LF1D45 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 MOVES3 Population
and Activity Report9 for more information. The zero-mile HHD and Urban Bus NOx emission rates by operating
mode are equivalent. Flowever, 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.
43
-------
3
|
32- Reg Class
© *¦ ^
1 HHD8
Urban Bus
| — MHD67
.1 __ ___9 LHD45
J i - ~ ^ LHD2b3
x
O
0-
0-3 4-5 6-7 8-9 10-14 15-19 20+
Age (years)
Figure 2-17 Heavy-duty Diesel NOx Emission Rates (g/mile) by Age for Model Year 2015 by Regulatory Class
Estimated using Nationally Representative Operating Mode Distribution
2.1.1.7 Summary
Table 2-10 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.
44
-------
Table 2-10 Summary of Methods for Heavy-Duty Diesel NOx Emission Rate Development for Each Regulatory
Class and Model Year Group
Model
year group
HHD
(regClass 47)
MHD
(regClass 46)
Urban Bus
(regClass 48)
LHD45
(regClass 42)
LHD2b3
(regClass 41)
1960-
1989,
1990
HHD 1991-
1997 rates
proportioned to
ratio of
certification
levels
Same rates as
HHD
Urban Bus 1991-
1997 rates
proportioned
using ratio of
HHD certification
levels
LHD 1991-1993 rates proportioned to
LHD certification levels
1991-1997
Data analysis3,0,
with
adjustments for
Low-NOx
rebuilds
Same rates as
HHD
Data analysis3
LHD 1999-2002 rates proportioned to
1991-1997 FTP standards per Table 2-1
1998
Data analysis3,0,
with
adjustments for
Low-NOx
rebuilds
Same rates as
HHD
Urban Bus 1999-
2002 rates
proportioned
using ratio of
HHD 1998 rates
to HHD 1999-
2002 rates
Same rates as 1999-2002
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
2007-2009
Data analysis'3
MHD 2003-
2006 rates
proportioned to
FTP standards
per Table 2-1°
Urban Bus 2003-
2006 rates
proportioned to
FTP standards per
Table 2-1
Percent reductions from the
MOVES2010 LHD 2003-2006 rates
(Section 2.1.1.4.6)°
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
2019-2060
Same as HHD
MY 2018
Same as MHD
MY 2018
Same as Urban
Bus MY 2018
Same as LHD45
MY 2018
Same baseline as
LHD2b3 MY 2018
with Tier 3
reductions phase-
in from MY 2019-
2022
Notes:
a Analysis based on ROVER and Consent Decree testing data
b Analysis based on HDIUT data
c Confirmed by HDIUT and/or Houston Drayage Program data
The role of the model year groups, representing a rough surrogate for technology or standards, can
be seen in Figure 2-19, which shows NOx emission rates for 0-3 age group by model yearand
regulatory class estimated in grams per mile (g/mile) using nationally representative operating
45
-------
mode distributions and average speeds. MOVES model estimates drastic reductions in NOx
emissions from pre-1990 technologies to modern 2018 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.
Figure 2-19 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
beginning in model year 2018 as documented in Table 2-10. 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 MOVES3 Population and Activity
Report.9
_30-
I
3
©
¦*-*
m
oc 20-
c
o
w
<2
E
LLl
o 10
Reg Class
Gliders
HHD8
Urban Bus
--- MHD67
LHD45
LHD2b3
0-
1970 1980 1990 2000 2010 2020 2030
Model Year
Figure 2-18 Heavy-duty Diesel NOx Emission Rates (g/mile) for the 0-3 Age Group by Regulatory Class using
Nationally Representative Operating Mode Distributions
46
-------
6-
Reg Class
HHD8
Urban Bus
— MHD67
LHD45
LHD2b3
..010
2015
2020
2025
2030
Model Year
Figure 2-19 Heavy-duty Diesel NOx Emission Rates (g/mile) for the 0-3 Age Group by Regulatory Class using
Nationally Representative Operating Mode Distributions for Model Year 2007 through 2030
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 fscaie updates to all HD regClasses for MY 2010 and later (see section 2.1.1.4.2
and Appendix G). Table 2-11 shows the model year group ranges and the applicable brake-specific
emissions standards.
2.1.2 Particulate Matter (PM2.5)
47
-------
Table 2-11 Model Year Groups Used for Analysis Based on the PM Emissions Standard
Model Year Group
Range
PM Standard (g/bhp-hr)
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
2.1.2.1.
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.42 The following description in the " Compilation of
Diesel Emissions Speciation Data - Final Report13 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:
48
-------
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
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 2-12.
49
-------
Table 2-12 Vehicle and Test Counts by Regulatory Class and Model Year Group
Regulatory
Model Year
Number of
Number of
Class
Group
tests
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 2-13.
Table 2-13 Vehicle Test Counts by Test Cycle
Test Cycle
Number of tests
CARB-T
71
CARB-R
66
CARB-I
42
UDDS_W
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).44 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.
50
-------
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 2-14
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.
Table 2-14 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
51
-------
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)45 demonstrated that time-integrated TEOM measurements compare well with gravimetric
filter measurements of diesel-generated particulate matter. Equation 2-16 shows the normalization
process for a particular one second TEOM measurement.
P
^^normalized,i,j vT~DA/f PM\\Equation 2-16
2/ ^™TEOM,i
Where:
i = an individual 1-Hz measurement (g/sec),
j = an individual test on an individual vehicle,
PMteomj,i = an individual TEOM measurement on vehicle j at second i,
PMfiiterj = the total PM2.5 filter mass on vehicle j,
PMnormaiizedjj = an estimated continuous emission result (PM2.5) emission result on vehicle j
at second i.
2.1.2.1.4 Compute A verage Rates by MO VES 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 /scaie. used for LHD. The emission rates
of pre-2010 LHD (LHD2b3 and LHD45 (regClassID 41 and 42) are based on an fscaie 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 MOVES2009.46
52
-------
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 factors47, 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 are approximating the STP-based MOVES emission rates for LHD
vehicles using these VSP-based rates.-5 Equation 2-17 used to derive the PM2.5 emission rates for
LHD regulatory class is shown below:
LHDPM2.5 emission rate
= 0.46 x MHD (VSPbased)PM2.s emission rate
Equation 2-17
Despite the uncertainty used in this approach, the representative PM2.5gram per mile emission rates
show reasonable trends (see Figure 2-27). 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, for some model year groups, the Urban Buses are subject to a different emission
standard, so we adjust the emission rates by applying a ratio of the EPA certification standards.
Table 2-15 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 are equal to the HHD emission rates multiplied by
the ratio in emission standards. In addition, the Urban Buses have different emission deterioration
effects as discussed in Appendix B.l.The gram per mile emission rates for Urban Bus presented in
Figure 2-27 show expected trends, with lower emission rates than the other regulatory classes
starting in MY 1991 through 2006.
Table 2-15 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.
J When this approximation was conducted in MOVES2010, the fscaie 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
MOVES3 to range from 3.5 to 7.8 metric tons.
53
-------
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 for the
2007-2009 vehicles), no continuous PM2.5 emissions data were availableon 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 emission rates to
estimate in-use rates for MY 2007-2009 vehicles.
An analysis of the certification data is shown in Table 2-16 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 marginthan 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.48
Table 2-16 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 largelycomposed of elemental
carbon. Elemental carbon is often used synonymously with soot and black carbon. Black carbon is
54
-------
important because of its negative health effects and its environmental impacts as a climate forcer.49
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 2-18.
PM2.5 = EC + NonECPM Equation 2-18
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
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.48 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 2-20 and Figure 2-21 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 2-20), 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 2-21), we model the large reductions
in overall PM2.5 rates and the smaller relative EC contribution to PM emissions. Figure 2-29 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.
55
-------
35
30
" 25
.c
S 20
TO
a. 15
10
c
TO
0J
I EC
InonEC
II.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 2-20 MHD Diesel PM2.5 Emission Rates for MY 2006 (age 0-3) by Operating Mode
1.4
1.2
1
JZ
CD
0.8
TO
s_
0.6
Q.
C
TO
CD
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 2-21 MHD Diesel PM2.5 Emission Rates for MY 2007 (age 0-3) by Operating Mode
2.1.2.2 2010 and Later Model 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 fscaie values
developed for the 2010+ MY NOx analysis described in Section 2.1.1.4.2 and Appendix G.
56
-------
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 2-17.
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.
Table 2-17. 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 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 2-17 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 2-18 shows the
minimum and maximum range of valid vehicle measurements by individual operating modes. The
operating modes with the smallestnumber 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.
57
-------
Table 2-18. 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
Group
(in at least one
operating mode)
any operating
mode
Measurements for any
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 2-19 and Equation
2-20, 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
treated as missing values. Despite the sparseness of the data by operating mode, due to the revised
fscaie 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
ER.
pol.OM ,C,MYG ,veh
jsec ERpol,OM,C,MYG,veh,sec
sec,
Equation 2-19
count
ER
pol,OM,C,MYG
veh ERpol,OM,C,veh
veh
Equation 2-20
count
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; seccount = a second of data (for a given veh and OM); number of seconds in that
category
veh; vehcount = 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.
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.
58
-------
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 2-17) 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 2-28.
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 2-22 through Figure 2-24 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
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).
1 As discussed in Section 2.3.2.2, we believe the reduction in THC and PlVh.swith 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 PM2.5. Additionally, SCR systems rely on oxidation catalysts and/or catalyzed
DPFs to convert NO to NO2, which also reduces PM2.5 tailpipe emissions.
59
-------
MOVES OpMode
Figure 2-22 PM2.5 Emissions by Operating Mode for LHD Model Year Groups 2010-2013 and 2014-2016
MOVES OpMode
Figure 2-23 PM2.5 Emissions by Operating Mode for MHD Model Year Groups 2010-2013 and 2014-2015
60
-------
ua
_E
I
a.
0,20 q
0.18 ^
0,16 ^
0,14 ^
0,12 ^
0,10 ^
0,08 ^
0.06 ^
0.04 ^
0.02 ^
0,00 ^
I MY 2010-2013
I MY 2014-2015
„li
ii ¦ ill 1J J.-T ±~> J.U • A.X <£...
sL2 JU •
W SJ 4^
MOVES OpMode
Figure 2-24 PM2.5 Emissions by Operating Mode for HHD Model Year Groups 2010-2013 and 2014-2015
As observed in Figure 2-22 through Figure 2-24, the PM2.5 emission rates are highest for the MHD
vehicles, followed by the HHD, and then the LHD 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 MHD vehicles, and we can only speculate on
the reasons the MHD 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 MHD than most of the other regulatory class and model year combinations, but not
all (Table 2-17).
Because much of the HDIUT PM data are missing or reported as zero, and given the additional
uncertainty regarding the MHD rates, we compared our HDIUT-based PM2.5 rates against values
reported in the literature. As shown in Figure 2-28, the MOVES age 0-3 PM2.5 rate ranges from 3 to
26 mg/mile for a MY 2010-2014 HD vehicle and from 2 to 7 mg/mile for MY 2014+. 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.50,51,52 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.53 Because the age 0-3 LHD2b3 MOVES emissions rates are well below
the Tier 3 standard, we do not estimate a reduction in PM2.5 rates with the phase-in of Tier 3 in MY
2018 and later vehicles as we do for NOx emissions (Section 2.1.1.5.5).
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+
61
-------
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
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.111
2.1.2.1 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
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
62
-------
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 7-3). The MOVES T&M adjustment factors on PM2.5 emissions over the fleet's useful
life are shown in Table 2-19 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.
Table 2-19
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
the Useful Life2
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.
2 Useful life varies by regulatory class (Table B-3)
2. 1.2.1.1 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 2-25 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
63
-------
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.
0.6-
m
c\i
Reg Class
HHD8
Urban Bus
— MHD67
LHD45
LHD2b3
0.0""
0-3 4-5 6-7 8-9 10-14 15-19 20+
Age (years)
Figure 2-25 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
2.1.2.1.2
2010-2060 Model 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 2-19 and scaled age effects (Table B-3) to estimate
emission rates for the different ages. 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 HD OBD phase-in
(see Section 2.1.2.1 and Appendix B).
Figure 2-26 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.
64
-------
3
2*0.015-
£0
^ Reg Class
¦§ „ , , . HHD8
iooio- Urban Bus
lu / MHD67
m
•*-*
o
-§ / LHD45
w S — LHD2b3
0.005
in
CM
0.000-
0-3 4-5 6-7 8-9 10-14 15-19 20+
Age (years)
Figure 2-26 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
2.1.2.2 Model Year Trends
Figure 2-27 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 2.1.1.7, 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 2-28 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 2-23 and discussed in
Section 2.1.2.2.1. The minor variation in the gram per mile emission rates within the 2010-2013
and 2014+ model year groups by model year and regulatory class are due to differences in
operating mode distributions.
65
-------
M 1-5-
E
3
©
a
an
o 1.0-
w
w
E
LU
.C
X
LU
S 0,5-
o
h-
w
CM
o.o-
197C
2010
2020
2030
Model Year
Reg Class
Gliders
HHD8
Urban Bus
MHD67
LHD45
LHD2b3
Figure 2-27 Heavy-duty Diesel PM2.5 Emission Rates (g/mile) for the 0-3 Age Group by Regulatory Class using
Nationally Representative Operating Mode Distributions
0.04-
jg
E
3
S 0.03-
(0
^ |,—Reg Class
¦§ I \ HHD8
*10 02 - / I Urban Bus
LU I I MHD67
"5 I \ LHD45
m I I LHD2b3
0.01-
CO
O
m
c\i
\
0.0
2010 2015 2020 2025 2030
Model Year
Figure 2-28 Heavy-duty Diesel PM2.5 Emission Rates (g/mile) for the 0-3 Age Group by Regulatory Class using
Nationally Representative Operating Mode Distributions for MY 2007 through 2030
66
-------
Figure 2-29 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
particle filters.
1.5- MWHWI
iiliill 11 »ll
I ¦¦¦¦¦
s lllll!l!j||||llB8
I1.0- 1111111111
ra i|||li||||||||||
^ ¦¦¦¦¦ ¦ EC
,0 '.mEC
— Illlllllllllllll
E Illlllllllllllll IH«__
0.0-
1970 1980 1990 2000 2010 2020 2030
Model Year
Figure 2-29 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:
67
-------
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 methodology8, where emission rates were calculated from 1-hz data
produced from chassis dynamometer testing. Data sources were all heavy-duty chassis test
programs:
1. CRC E-55/5942: 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)54: 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)55: 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.
68
-------
Table 2-20 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
1960-2006 Model 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
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 2-21. 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.
Table 2-21 Age Groups for which THC and
CO Emission Rates are Populated Directly Based on the Data
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
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.
69
-------
Figure 2-30 and Figure 2-31 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.
100
90
80
70
60
50
S 40
«
30
20
10
0
~ MHD
= HHD/Bus
> 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 2-30 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
70
-------
600
~ MUD
¦ HHD/Bus
1 I
0
i „ * 1
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 33 40
Operating mode
Figure 2-31 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
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 2-22
and discussed in Appendix Section B.9. As shown in Figure 2-42, 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 and Later 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)56 and a non-
dispersive infrared (NDIR) analyzer for carbon monoxide (CO)57
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
2.1.3.1.3 2007-2009 Model Years
71
-------
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 of fscaie). Figure 2-32 and Figure 2-33 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 (Figure 7-9) and especially HHD (Figure 2-32).
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.
MOVES OpMode
Figure 2-32 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
72
-------
"S 0.06 -
ilk
£l
I
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
L
i
11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES OpMode
Figure 2-33 Average HHD CO Emission Rates by Operating Mode for the 0.2 NO* FEL for MY 2010-2013 and
the 0.35 and 0.5 NO* FEL for MY 2010-2015. Error Bars are 95% Confidence Intervals of the Mean
Figure 2-34 and Figure 2-35 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.
0.0050
0.0040
aT 0.0030
-S5
(J
5 0.0020
0.0010
0.0000
¦ HDIUT, MY 2013
.ill ill
1
j I 1
1 ii
i i
. 11111. ¦ 11 i 1
1
1
ti i i 1
II
0 1
33 35 37 38 39 40
11 12 13 14 15 16 ,21 22 23 24 25 27 28 29 30
MOVES OpMode
Figure 2-34 THC Emissions by Operating Mode from HHD Trucks for Model Year 2013. Error Bars represent
the 95 percent confidence interval of the Mean
73
-------
0,03 -
1 "! ?"} ¦:> /: : "IV :\'.1 ~>7
MOVES OpMode
Figure 2-35 CO Emissions by Operating Mode from HHD Trucks for Model Year 2013. Error Bars represent
the 95 percent confidence interval of the Mean
Figure 2-36 and Figure 2-37 display the comparison of the MY 2010-2013 and MY 2014-2015
groups within the HHD 0.2 NOx FEL Groups. In general, the newer vehicles (MY 2014-2015) have
lower THC and CO emission rates than the corresponding MY 2010-2013 emission rates. Similar
model year trends are observed for LHD THC emissions, and MHD 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 NOx FEL Group to improved emission control hardware and engine
and aftertreatment operation.
MOVES OpMode
Figure 2-36 THC emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the HHD 0.20 NOx FEL
Group
74
-------
0.14
0.12
0.10
jr
tin 0.08
Q
U 0.06
0.04
0.02
0.00
0 1 11 12 13 14 IS 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
MOVES OpMode
Figure 2-37 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. 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 2-41 and Figure 2-42, respectively. The THC rates, generally
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,14 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 2-42). 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
: MY 2010-2013
MY 2014-2015
1 11
75
-------
that the CO emission rates measured from HDIUT are reasonable. Thus, we decided to accept the
reported HDIU CO emission rates as valid.
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 2-22 to CO and THC.
Table 2-22 Tampering and Mal-maintenance Effects for THC and CO over the Useful Life
Model years
Increase in THC and CO Emissions (%)
Pre-2003
300
2003 - 2006
150
2007 - 2009
150
2010 - 20121
29 (HHD, MHD, LHD45, and Bus)
22 (LHD2b3)
2013+
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-3 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-3 and applied the result to
the zero-mile (or age 0) emissions rate to get the emissions rate by age group using Equation 7-3.
Figure 2-38 and Figure 2-39 show THC and CO emission rates by age group. 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.
76
-------
100
90
80
'imm'
70
.S
jag
60
W
m
w
50
U
X
c
40
m
30
20
10
0
0-3 4-5 6-7 8-9 10-14
Age group [years]
15-19
HHD
MHD
20+
Figure 2-38 THC Emission Rates [g/hr] by Age Group for Model Year 2002 and Operating Mode 24. Error Bars
Represent the 95 Percent Confidence Interval of the Mean
J2
m
**
m
o
u
c
re
I
0-3 4-5 6-7 8-9 10-14 15-19
Age group [years]
20+
Figure 2-39 CO Emission Rates [g/hr] by Age Group for Model Year 2002 and Operating Mode 24. Error Bars
Represent the 95 Percent Confidence Interval of the Mean
77
-------
2.1.3.4
Model Year Trends
Figure 2-40 through Figure 2-42 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 2.1.1.7,
some of the minor variation in the gram per mile emission rates within the model year groups
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 differeng
gram per mile emission rates observed for 2003-2006 in Figure 2-40 and Figure 2-42.
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 2-41) is explained by the high THC emissions of the
0.5 NOx FEL group (Figure 2-32), and the high production volumes of the 0.5 NOx FEL engines in
MY 2010 (-50% of the total HHD, see Figure 2-12).
78
-------
0.5
m
1 0
3
®
¦*-»
m
^ 0.3
o
~w
w
E
ll)0,2-
O
X
«
TO
0
0.1
Reg Class
Gliders
HHD8
Urban Bus
MHD67
LHD45
LHD2b3
o.o-
1970
1980
1990
1
Model Year
_„,0
2020
2030
Figure 2-40 Heavy-duty Diesel THC Emission Rates (g/mile) for the 0-3 Age Group by Regulatory Class using
Nationally Representative Operating Mode Distributions
0.15-
J£
E
3
®
3 ' ' Reg Class
ac ojo- y
HHD8
O
X
m 0,0
ro
O
"5
<*-*
o
0.0
c
0
% 1 Urban Bus
1 ¦— MHD67
LHD45
LHD2b3
2010 2015 2020 2025 2030
Model Year
Figure 2-41 Heavy-duty Diesel THC Emission Rates (g/mile) for the 0-3 Age Group by Regulatory Class using
Nationally Representative Operating Mode Distributions (Model Year 2007-2030 only)
79
-------
Reg Class
Gliders
HHD8
Urban Bus
-«"• MHD67
LHD45
LHD2b3
0-
1970 1980 1990 2000 2010 2020 2030
Model Year
Figure 2-42 Heavy-duty Diesel CO Emission Rates (g/mile) for the 0-3 Age Group by Regulatory Class using
Nationally Representative Operating Mode Distributions
2.1.4 Energy
2.1.4.1 1960-2009 Model Years
2.1.4.1.1 LHD
In MOVES3, 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.58 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 2-46 and Figure 2-47.
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
*E
3
®
15
or
c
o
»
£ 1"
LU
O
o
80
-------
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 2-21), we used a heating value (HV) of 138,451
kj/gallon and CO2 fuel-specific emission factor {fco2) of 10,180 g/gallon59 for conventional diesel
fuel.
HV
^energy ^C02 ~~c Equation 2-21
JC02
The energy rates for the MHD, Urban Bus, and HHD vehicle classes are shown in Figure 2-43.
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 2-46 and Figure 2-47.
t ^ ^ ^ ^ »
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 2-43 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
81
-------
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 HV of of
138,451 kj/gallon, yielding a CO2 fuel-specific emission factor (fcoz) 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 2-44 shows the mean HHD 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.
100 -
80 -
*5T 60 -
m
8 40 -
20 -
0 -
0 1 ; 11 12 13 14 15 16 i 21 22 23 24 25 27 28 29 30 ; 33 35 37 38 33 40
MOVES OpMode
Figure 2-44 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 NOx FEL for MY 2010-2015. Error Bars are 95% Confidence Intervals of the Mean.
As shown in Table 2-23, the majority of vehicles within the NOx FEL groups of 0.35 and 0.5
include MY 2010-2013 vehicles, for both HHD and MHD vehicles. As discussed in the next
subsection, we expect improved energy effiencies in MY 2014 and later vehicles due to the phase-
in of the Phase 1 Heavy-duty Greenhouse Gas Emission Standards. Due to the small sample of
(¦HHD FEL 0,20, MY 2010-2Q13j N=78
(¦ HHD PEL 0.35, MY 201CWG15- N=31
¦ HHD FEL 0.50, MY 2010-2015; N=35
III
82
-------
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 2-23 HDIU Vehicles with Valid CO2 Measurements By Regulatory Class, Model Year Group and NOx
NfOx 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 2-46 and Figure 2-47.
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 2-23), 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 Rule60 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.9
The HD GHG Phase 1 rule60 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
83
-------
reductions for heavy-duty diesel energy rates from the HD GHG Phase 1 rule are shown in Table
2-24.
In MOVES3, 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 HDIU program. Instead, we renormalized the
Phase 1 GHG reductions using the MY 2014-2016 as the baseline using Equation 2-22.
Renormalized Phase 1 reductions in yeart
1 — (reductions in year,) Equation
1 — (Average reductions in 2014 thru 2016) 2-22
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-T^5%j=1-95%=1-96% = 4%
We applied the renormalized reductions to estimate the MY 2017 and later running energy rates, as
shown in Table 2-24. 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 2-24 Estimated Reductions in Diesel Engine Energy Consumption Rates from the HD GHG Phase 1
Program61
Regulatory
Class
Fuel
Model
Years
Estimated Reduction from the
MY 2013 Baseline
(applied to starts)
MOVES3 Renormalized
Reductions to MY 2014-
2016 Energy Rates
(applied to running)
HHD and
Urban Bus
Diesel
2014-2016
3%
-
2017
6%
3%
LHD and
MHD
Diesel
2014-2016
5%
-
2017-2020
9%
4%
MOVES3 also incorporates the Medium- and Heavy-Duty GHG Phase 2 rule.62 The Phase 2
program begins in 2018 model year for trailers and in 2021 for the other categories, while phasing
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)
84
-------
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."
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, it was necessary to create a new
EmissionRateAdjustment table with sourceTypelD as a primary key. The EmissionRateAdjustment
table alllows MOVES3 to model the final standards for vocational vehicles and tractor-trailers
simultaneously. The EmissionRate Adjustment table includes the following data fields, many shared
with the EmissionRate table:
1
2
3
4
5
6
7
8
polProcessID (primary key)
sourceTypelD (primary key)
regClassID (primary key)
fuelTypelD (primary key)
beginModelYearlD (primary key)
endModelYearlD (primary key)
emissionRate Adj ustment
dataSourcelD
Table 2-25 summarizes the energy rate reductions stored in the EmissionRate Adj ustment table
which are applied to the running rates in MOVES3 for MY 2018 and later heavy-duty diesel
vehicles.
n HFCs are not modeled in MOVES, and the N2O and CH4 standards are not considered technology forcing on
emissions.
85
-------
Table 2-25 Estimated Reductions in Diesel and CNG Engine Energy Consumption Rates due to the HD GHG
Phase 2 Program62
Vehicle Source Type
(Source Type ID)
Fuel
Model years
Reduction from MY 2017
Energy Rates
Long-haul Combination
Truck
(62)
Diesel
2018-2020
1.0%
2021-2023
7.9%
2024-2026
12.4%
2027+
16.3%
Short-haul Combiation
Truck
(61)
Diesel & CNG
2018-2020
0.6%
2021-2023
7.4%
2024-2026
11.9%
2027+
15.0%
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 2018 and later
model years are estimated with a chain of calculations starting with the HDIUT-based estimates by
operating mode and regulatory class for MY 2018, then reduced by applying the HD GHG Phase 1
reduction in Table 2-24 and further reduced by applying the HDGHG Phase 2 reductions listed in
Table 2-25 for model year 2018 and later. The reductions shown in Table 2-25 are a combination of
improvements to the engine and other systems excluding 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.9 The reductions in
energy use applied from both the Phase 1 and Phase 2 HD rulemakings to the MY 2014-2018
energy rates developed using HDIUT data are displayed in Figure 2-45.
86
-------
a;
+->
03
a:
c
f\l
C o
O
CN |
xs
m °
(T3 4->
rH "D
ui jy
aj
i/> ;—
ro ro
E
Q_
O
X
u
aj
0
c
1
aj
cc
1.00
0.95
0.90
0.85
0.80
0.75
0.70
Phase 2 baseline: MY 2017
Phase 1: MY 2014-2018
Phase 2: MY 2018-2027+
M0VES3 baseline:
MY 2014-2016
Phase 2 LHD2b3
baseline: MY 2018
2012
2014
2016
2018
2020 2022
Model Year
2024
2026
2028
2030
-© - LHD2b3
--B--LHD2b3
--{3— MHD (long-haul)
--EH-HHD (long-haul)
-©— LHD45 and MHD
-H— LHD45andMHD
-K— HHD (voc)
¦ X-• Bus
-©— HHD and Bus
-A— MHD (short-haul)
-A— HHD (short-haul)
Figure 2-45 HD GHG Phase 1 and Phase 2 rule reductions in running energy consumption rates for LHD2b3,
LHD45, MHD, HHD, and Urban Bus diesel vehicles from the MY 2014-2016 baseline used in MOVES3.
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. Table 2-26
describes the expected changes in CO2 emissions for diesel LHD2b3 vehicles due to improved
engine and vehicle technologies due to the HD GHG Phase 1 program. The impacts of the HD
GHG Phase 1 program on gasoline LHD2b3 energy rates are discussed in Section 3.1.3.2.1.
Because MOVES3 includes energy rate measurements from LHD vehicles for model years 2014-
2016, we renormalized the Phase 1 reductions to be adjusted to the 2014-2016 model years using
Equation 2-22. The example calculations for LHD2b3 diesel in MY 2018-2020 are provided in
Equation 2-24.
87
-------
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% 2-24
Table 2-27 shows the projected improvements in CO2 emissions due to the HD GHG Phase 2
program for diesel and gasoline LHD2b3 vehicles. 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. 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 of using the EmissionRate Adjustment table, the energy
consumption rates for LHD2b3 were lowered by the percentages shown in Table 2-26 and Table
2-27 for the corresponding model years.
The heavy-duty regulations apply to all the vehicles in the LHD2b3 regulatory class, with the
exception of medium-duty passenger vehicles (MDPVs). The fuel economy of MDPVs are covered
by the Light-duty GHG rule.63 However, MDPVs make a minor contribution to the population of
LHD2b3, and we apply the HD GHG energy reductions to all vehicles within LHD2b3.
Table 2-26 Estimated Total Vehicle Reductions in Energy Consumption Rates for LHD2b3 Diesel Vehicles due
Regulatory
Class
Fuel
Model years
Reduction from MY
2013 Energy Rates
MOVES3
Renormalized
Reductions to MY
2014-2016 Energy
Rates (applied to
running)
LHD2b3
Diesel
2014
2.3%
-
2015
3%
-
2016
6%
-
2017
9%
5%
2018-2020
15%
12%
Table 2-27 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
2021
2.50%
and Diesel
2022
4.94%
2023
7.31%
2024
9.63%
2025
11.89%
2026
14.09%
2027+
16.24%
88
-------
2.1.4.4
Model Year Trends
Figure 2-46 and Figure 2-47 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 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. 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-
©
ra
on
Reg Class
11000-
o
Gliders
HHD8
CM
O
o
LU
Urban Bus
MHD67
LHD45
¦— LHD2b3
£L
W
O
E
<
1970 1980 1990 2000 2010 2020 2030
Model Year
Figure 2-46. Heavy-duty Diesel CO2 Emission Rates (g/mile) by Regulatory Class using Nationally
Representative Operating Mode Distributions
89
-------
25
Reg Class
Urban Bus
^ MHD67
LHD45
-*- LHD2b3
Gliders
HHD8
5-
_030
Model Year
Figure 2-47. Heavy-duty Diesel Fuel Economy (miles per gallon) by Regulatory Class using Nationally
Representative Operating Mode Distributions
2.1.5 Evaluation of Fleet-average Running Rates with Real-World
Measurements
As one evaluation of the MOVES diesel exhaust running rates, Table 2-28 compares MOVES
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 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 2-28 shows that CO emission factors compare quite well between the different studies. The
NOx values are comparable to the Peralta CA location, but significantly lower than the Ft.
McHenry location. THC are also below the Peralta, CA measurements. PM2.5is lower than the Ft.
McHenry estimates, but within the standard error of the winter measurements, and close to the 95%
90
-------
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 2-28. Comparison of MOVES Emissions with Remote Sensing and Tunnel Measurements
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
O
X
12.4 ±0.6
29.6 ±4.7
17.9 ± 1.4
12.2
PM2.5
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 Exhaust 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.0
As explained in Section 1.5, 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.8 The impact of ambient temperature on cold starts is discussed in
the Emission Adjustments MOVES report.64
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, andNOx
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
91
-------
that Bag 1 starts with a cold start, and Bag 3 begins with a hot start.0 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 2-29.
Table 2-29 Average Start Emissions Increases (g/start) for pre-2010 Model Year Light Heavy-Duty Diesel
Vehicles for Regulatory Class LHD2b3 and LHD45
THC
CO
O
X
0.13
1.38
1.68
regClassID 41 and 42)
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 2-48.
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
45
40
35
30
-S
25
E "
20
15
10
5
0.00
1.00
4.00
5.00
2.00 3.00
time [hrs]
Figure 2-48 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)
0 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.
92
-------
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 2-25.
Grams per Start =
= stabilization time x (cold start average rate — hot running average rate)
= 25 minutes x
1 emissions
~~25
-I
t=25
emissions \
215 J
Equation 2-25
The results are shown in Table 2-30. The measured THC increment is zero. The NOx increment is
negative since cold start emissions were lower than warm idle emissions.
Table 2-30. Cold-start Emissions Increases (g/start) in Grams on the 2007 Cummins ISB
THC
CO
o
X
0.0
16.0
-2.3
We also considered NOx data from University of Tennessee,65 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 2-25 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 2-31
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 2-31. MOVES Inputs for Pre-2010 HHD and MHD Diesel Start Emissions (grams/start) for Regulatory
THC
CO
O
X
0.0
16.0
0.0
As discussed in the Emission Adjustments Report64, 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
93
-------
applied to CO, PM2.5, or NOx diesel start emissions because no clear trend was found with the
data.
2.2.1.2 2010-2060Model 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) cycle66. 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 this MOVES3 update includes the
following engine families from 2016 and 2017 model years shown in Table 2-32.
Table 2-32 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 2-26.
Grams per Start
= [Cold FTP Emission Results (g/(hp — hr))
— Hot FTP Emission Results(g / (hp — hr))] Equation 2-26
* 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 2-49, 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 2-49 - 0.0599 (hr) times the rated power (hp) plus
4.4297 (hp-hr).
94
-------
40
35
30
n
JZ
L_
25
O
s
SO-
lS
FTP work wer - Diesel
v = 0.6553* - 4.4257
R!= 0J5D3
250 500 350 -£00 450
Rated Pov/er(hp}
5 3D
550
Figure 2-49: Relationship between HD Diesel Engine Rated Power and FTP Cycle Work
2.2.1.2.1 Hea vy-hea vy 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 2-33. The PM2.5 emissions
are summarized in Table 2-37. 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 for MOVES3 are the mean values shown in the table. The THC
and NOx cold start emissions for HHD diesel engines are higher than those in MOVES2014, while
the CO emissions are lower.
Table 2-33: 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-hea vy 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 2-34. 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 for MOVES3 are the mean values shown in the table. Similar to the HHD engines, the THC
and NOx cold start emissions for the MHD diesel engines are higher than those in MOVES2014,
while the CO emissions are lower.
Table 2-34 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
95
-------
2.2.1.2.3 Light-hea vy duty
Analysis of five LHD diesel engines from the certification data determined the grams per start for
THC, CO, and NOx shown in Table 2-35. The sample included MY2016 and MY2017 engines,
ranging in displacement between 3.0 and 6.7 liters, and in rated power between 161 and 330 HP.
The default values for MOVES3 are the mean values. The CO and NOx cold start emissions are
higher than those in MOVES2014, while the THC emissions are lower.
Table 2-35 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 Hea vy-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 are applied to THC and CO rates.
2.2.1.2.5 Model Year Summary
Figure 2-50 through Figure 2-52 display the displays 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.
96
-------
0.20-
•e
m
•*->
m
30.15-
jg
«
tr
c
o
"®
® 0.10-
E
Lit
0
1
CO
0 0.05-
_
o
Reg Class
Gliders
HHD8
Urban Bus
MHD67
LHD45
LHD2b3
o.oo-
1380
2000
Model Year
2020
Figure 2-50 Heavy-duty Diesel THC Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class
and Model Year
15
¦E
CO
*-*
tn
ra
CD
4-*
CO
tr
c.
o
"w
w
E
LU
O
o
10-
5^
Reg Class
Gliders
HHD8
Urban Bus
MHD67
LHD45
LHD2b3
o-
1980 2000 2020
Model Year
Figure 2-51 Heavy-duty Diesel CO Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class
and Model Year
97
-------
7.5-
¦c
m
jn
3
©
13
tr
c
o
'to
E
UJ
X
O
5.0-
2.5
Reg Class
Gliders
HHD8
Urban Bus
^ MHD67
LHD45
LHD2b3
o.o-
1980
2000
Model Year
2020
Figure 2-52 Heavy-duty Diesel NOx Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class
and Model Year
2.2.2 Particulate Matter (PM2.5)
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 2-36.
Table 2-36 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.0 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,
98
-------
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 2010-2060Model Years
The cold start emissions for 2010 model year and later LHD, MHD, and HHD diesel engines have
been 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 2-37. 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 2-37 is of similar magnitude or greater than the 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 MOVES3, because of the overlap in engines and aftertreatment systems between
the two categories.
Table 2-37: 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
MOVES3 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 2-52 and Figure 2-53 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.
99
-------
0.125-
t
«
3? 0.100-
s
©
•*-*
m
CC
e
2
"m
w
E
UJ
sz
x
LU
4—»
O
£ 0.025-
c4
0.075-
0.050-
Reg Class
Gliders
HHD8
Urban Bus
MHD67
LHD45
LHD2b3
o.ooo-
13L.
Model Year
Figure 2-53 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.
2.2.3 A djusting 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.9
2.2.3.1 Adjusting Start Rates for Soak Time - MY2009 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 2-54 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.8 The soak adjustments are taken from the non-catalyst soak adjustments derived in a
CARB report67 and reproduced in a MOBILE6 report.68
100
-------
Soak Time (minutes)
Figure 2-54. 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
Report8
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
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 2-38 for THC, CO, and NOx.
Table 2-38. 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 2-39 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).
101
-------
Table 2-39. 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 - MY2010 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 new test programs were conducted to revise the 2010 MY and later soak curves for
heavy-duty diesel vehicles in MOVES3. 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 2-55. 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 2-55. At least two repeats were conducted for each soak period. The emission measurements
included dilute gaseous measurements and triplicate particulate matter filters.
102
-------
Double NREL Cycle
60
O^NOO^OtCNtt^O^NeO^Q^INN^OdnCO^O
lflHii(N05mffl9OtfrlNN00fCni/5o*Dr|Nmca*fQ
iH*HfMfMmpn,«jLninwDvDr-^r-.cococnoo'-iHfMfMfn^
rHrHr-lrHrHrHTHTH
Time (seconds)
Figure 2-55 National Renewable Energy Laboratory's Heavy-Duty Vocational Transient Cycle
The NGx, CO, THC, and PM2.5 emission results in terms of grams or mg per mile from the tests
over a range of soak periods are shown in Figure 2-56 through Figure 2-59.
Emissions (g/rnile)
op|yyiAui|jNj#
bbbbbbb bb
NOx
»
•
•
•
•
•
•
a
1
f
1
1 10 20 30 40 50 60 70 B0 90 100
SoakTime (hours)
Figure 2-56 MY 2015 Heavy-Duty Vehicle NOx Emissions by Soak Time
103
-------
O.DO m
0 1C 20 30 £0 50 §0 70 50 100
^ n s t Tlmp fhmircl
Figure 2-57 MY 2015 Heavy-Duty Vehicle THC Emissions by Soak Time
_S
i
0.50
0,45
0.40
0,35
0.30
0.25
0.20
0.15
D.1C
0.05
0,00
r
CO
10
20
30
60
70
80
00
100
Figure 2-58 MY 2015 Heavy-Duty Vehicle CO Emissions by Soak Time
104
-------
0 IS 22 30
53 73 iO 50 IOC-
Figure 2-59 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
profile from the route is shown in Figure 2-60. The route consisted of approximately 700 seconds
of driving in a neighborhood/urban environment over approximately 2.7 miles.
105
-------
Soak Route Speed Profile
0 200 440 600
T*ip Counter !4ecC*"dsi
Figure 2-60 Onroad Soak Drive Route
The emission results, in terms of total emissions over the route, from the onroad tests are shown in
Figure 2-61 through Figure 2-63.
3.0
2.5
Si
c 20
o
«" 1.5
E
LU
ra 1.0
£
05
0.0
(
Total NOx Over Cycle
•
•
•
A
•
•
•
>
0
> 5 10 15 20 25 3-
SoakTime (hours)
Figure 2-61 MY 2016 Heavy-Duty Vehicle NO* Emissions by Soak Time
106
-------
0.8
0,7
JS
0.6
;z
G
0,5
S/f
LI
0.4
J
0.5
15
o
0,2
5,1
5,0
Tots I
2D 25 3D
Figure 2-62 MY 2016 Heavy-Duty Vehicle THC Emissions by Soak Time
Total CO O »cle
1 3-c
.c
2,0
ID
D.D
• •
10
15
2D
25
3D
Soat
Figure 2-63 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
107
-------
multiplied by the ratio of 3 minutes divided by 18 minutes (the midpoint times of operating mode
101 and 102). The NQx, CO, and THC soak period ratio results for each vehicle are shown below
in Figure 2-64.
HD Diesel Truck Start Emission Ratio
1.20
1.00
.2 0.80
%
ec
o 0.60
in
£ 0.40
0.20
0.00
/
£
*
1/
f *
I %
t i
wr
P" ^
T""*1
^ 4t tusiz,4 ,V ^
4 !¦¦¦:'¦¦
Si ''
, . 4 . -J
If-'
100 200 300 400 500
Soak Time (min)
600
700
800
...,m... HD Van NOx
—Day Ccb NOx
¦ HD Van CO
—Day Cab CO
— HD Van HC
— Day Cab HC
Figure 2-64 Soak Emission Ratios from a MY 2015 HI) Day-Cab and a MY 2016 HI) Van
The 2010 MY and later heavy-duty diesel soak ratios for MOVES3 were determined by averaging
the results from the two trucks. The resulting soak adjustment ratios are shown in Table 2-40. 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.
108
-------
Table 2-40 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
The PM2.5 start rates by operating mode for MY 2010 and newer vehicles are given in Table 2-41
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.
Table 2-41 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).64 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.
109
-------
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 2-54 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.p The additive cold start adjustment for THC emission factors are displayed in
Table 2-42, 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.
Table 2-42 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
-0.0201x(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.58 Figure 2-65 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 2-24 and Table 2-26. 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).
p 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.
110
-------
500-
t?
ro
-+-»
3400-
B
or . _ _ _ Reg Class
¦2 30C \ S^. Giiders
I ]/ \ HHD8
£ Urban Bus
-------
Table 2-44. 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 Extended Idling Exhaust Emissions
In the MOVES model, extended idling is idle operation characterized by long duration idle periods
(e.g., > 1 hourq) , 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."72 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 migh 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, diesel long-haul combination trucks (sourceTypelD 62) are the only fuel
and source type combination assumed to have extended idling activity. These trucks are only
associated with diesel MHD, HHD and Glider1 regulatory classes. As an alternative to extended
idling, long-haul truck operators can also use auxiliary power units (APUs) to power their cabin
and accessories during hotelling. The emission rates for auxiliary power units (APUs) are discussed
in Section 2.4.
q The default hotelling activity in MOVES3 is estimated from telematics data assigning all idle events > 1 hr from long-
haul combination trucks to extended idling.
r Glider extended idle emission rates are documented in Section 2.5.
112
-------
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 1960-2006Model 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,
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.73 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 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.).74 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.).75 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.).75
• The California test data was collected on 42 diesel trucks in parallel with roadside smoke
opacity testing (Lambert)76. All tests conducted by the California Air Resources Board
(CARB) at a rest area near Tulare, California in April 2002 are described in the Lambert76
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.77
• The National Cooperative Highway Research Program (NCHRP)78 obtained the idling
portion of continuous sampling during transient testing to determine idling emission rates
on two trucks.
113
-------
• A total of 33 heavy-duty diesel trucks were tested in an internal study by the City of New
York (Tang et al.)79. 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).80 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.).81 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).65
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.82 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.83 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)84 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, 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
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."
114
-------
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-duracton 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. (20 09)85 found that MHD and
HHD trucks had similar emission rates during extended idle. Consequently, MOVES3 applies the
same extended idle emissions rates to MHD and HHD, as shown in Table 2-45.
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
115
-------
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 2-45.
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 emssion deterioration, they do include real-
world vehicles at a variety of ages and conditions and thus it would be "double-couting" to apply
the exhust running T&M effects to these rates.
Table 2-45. Pre-2007 Extended idle emission rates (g/hour) in MOVES by pollutant for MHD and HHP
Model Year
Groups
NOx
THC
CO
PM2.5
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 2007-2060Model Years
The 2007 and later model year extended idle emission rates were updated in MOVES3 using 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 MOVES3 were updated 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 2005s to 2012.86 Another study conducted by California Air Resources
Board (ARB)87 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 becaue 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 2-46. The last three columns in Table 2-46 are taken from the California Executive Order
certification database.88 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
s 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.
116
-------
of nonconformance penalty (NCP) engines in 20 1 2.89 In these cases, the family emission limit for
which the vehicle was certified is reported in Table 2-46. 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
2-46 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 2-46. HHP Diesel Tractors Used to Update the MY 2007 and Later Extended Idle Emission Rates
Clean
Study
Engine
MY
Engine
Odometer
NOx cert
(g/bhp-hr)
Idle
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 2-47) 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.
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
117
-------
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
conditions1. 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 2-66 through Figure 2-70.
Table 2-47. 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)
30V (-i.it:)
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.90 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 A/C 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
To develop the revised extended idle emission rates for MOVES3, 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 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
Tab' temperatures.
The individual test results and the average emission rates by model year group are presented in the
following figures (Figure 2-66 to Figure 2-70). Within each figure, the tests are distinguished
'As discussed earlier, our assumptions for pre-2007 trucks are different.
118
-------
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 2-66 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
119
-------
CQ2
A
15000-
2008
2008
2010
2012
Figure 2-66. 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 2-67 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
MOVES3 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.
120
-------
CO
2006
2008
2010
2012
2014
Figure
2-67. CO Emission Rates from the TTI and ARB Programs by Engine Model Year, and Average
Emission Rate (line) Based on All the Data
Figure 2-68 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 and later model years) as
represented by a solid line in Figure 2-68. 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 2-71.
121
-------
NOx
2006
2008
2010
2012
Figure 2-68. 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 2-69 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 trucks11 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.91 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.
11 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.
122
-------
THC
^10-
%
o
m
"m
'E
©
©
T3
i
condition
= hot
¦= cold
iab
SCR
• 0
A -
2006
2008
2010
2012
2014
Figure 2-69. 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 2-70 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.v 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.
v The MY 2005-2006 PM2.5 emission rates measured from the TTI data are only ~3 times higher than the MY 2007-
2009 PM2.5rates, and roughly -10 times smaller than the PM2.5rates for MY 1990-2006 (2.5 g/hr). 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
informatin from the TTI dataset due to limitations on time and resources.
123
-------
PM
Figure 2-70. 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 2-48 to estimate the EC and PM2.5 emission rates.
Table 2-48. Baseline elemental carbon to PM2.5 fraction assumed for extended idling
Model Year
Group
EC/PM
Source
MOVES2014 Extended
Pre-2007
0.26
Idling11™
2007-2009
0.10
ACES Phase I90
2010+
0.16
ACES Phase II92
2.3.2.3 Tampering and Mal-maintenance
As discussed in Section 2.3.1.2, we did not incorporate tamperaing and mal-maintenance effects on
the pre-2007 extended idle rates. For the 2007 and later extended idle rates, we incorporatd 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-maintened vehicles with elevated emission rates. In addition,
14 of the 20 vehicles had odometer readings with less than 100,000 miles (Table 2-46).
w 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.
124
-------
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-maintaned.
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 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 2-49. As shown, the MOVES EC/PM emission rates for MY
2007+ trucks are slightly higher than the 'Baseline' EC/PM fractions in Table 7-2, because the fleet
emissions are assumed to include some emission contribution from trucks with failed DPFs, which
have a higher EC/PM fraction.
Table 2-49. 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)
FC
(g/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 5 percent of long-haul
combination trucks in the US and therefore, they are a minor contributor to the emissions from
extended idling trucks. Second, Khan et al. 200993 evaluated extended idle emission rates of pre-
125
-------
2007 MHD engines and did not observe a pronounced difference in extended idle emission rates
between MHD and HHD trucks. 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.
Extended idle emission rates were updated for all model years, as described in Sections 2.3.1 and
2.3.2. Figure 2-71 through Figure 2-74 illustrate the extended idle emission rates in MOVES3 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 PM2.5rates between the pre-2007 and
post-2007 extended idle rates corresponding to the implementation of the DPFs, which is consistent
with the -27 decrease in PM2.5 running exhaust emission rates from PM2.5 certification data as
discussed in Section 2.1.2.1.7.
2.3.3 Model Year Trends
Extended Idle NOx Emission Rate
150
13
DC 75
o ^
to 50
E
m 25
0
1960
1980
2000
2020
2040
Model Year
Figure 2-71. Extended Idle NOx Emission Rates for HHD and MHD Diesel Vehicles by Model Year
126
-------
Extended Idle THC Emission Rate
150
J 125
O)
j 100
"5
K 75
<75 50
w
£ rtr
lu 25
0 ¦
1980 1980 2000 2020 2040
Model Year
Figure 2-72. Extended Idle THC Emission Rates for HHD and MHD Diesel Vehicles by Model Year
Extended Idle CO Emission Rate
150
JT 125
£100
tr 75
m 50
w
E
m 25
0
1960 1980 2000 2020 2040
Model Year
Figure 2-73 Extended Idle CO Emission Rates for HHD and MHD Diesel Vehicles by Model Year
127
-------
Extended Idle PM Emission Rates
5
4
Elemental Carbon
I
3
Non-Elemental Carbon PM
1960
1980
2000
2020
2040
Model Year
Figure 2-74. Extended Idle PM2.5 Emission Rates HHD and MHD Diesel Vehicles by Model Year
2.3.4 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 2-75. 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 2-49 and are also plotted
in Figure 2-75. The extended idle energy consumption rates are the same for regulatory class MHD
and HHD diesel vehicles.
128
-------
1,5e+Q
regClassName
~ HHD8
* MHD67
modelYearlD
Figure 2-75. Extended Idle Energy Emission Rates for HHD and MHD Diesel Trucks
2.4 Auxiliary Power Unit Exhaust
Auxiliary power unit (APU) exhaust is a separate emission process in MOVES. APU usage only
applies to the vehicles with hotelling activity, which are the heavy-duty regulatory classes (MHD,
HHD, and Gliders) within the combination long-haul truck source type (sourceTypelD 62). The
APU emission rate for MHD, HHD and glider regulatory classes are the same for each model year.
The projected use of APUs during hotelling due to the HD GHG Phase 2 program, shown below in
Table 2-50, were used to revise the "hotellingactivitydistribution" table in MOVES362, as is also
discussed in the Population and Activity Report.9
Table 2-50: 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%
For MOVES3, the APU emission rates have been updated to reflect new standards, data, and
analysis. The APU emission rates were updated in MOVES3 based on two studies that measured
in-use APU emission rates. The Texas Transportation Institute (TTI, 2014)94 tested two diesel APU
129
-------
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.100*
The second study used to update APU emission rates was by Frey and Kuo (20 09),95 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 20 1 296 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. 200397 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
20 0995 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.
Table 2-51. 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
TTI 201494
2
Kubota Z482
2006
0.48
10.9/8.1
Tier 2
Frey and Kuo
200995
3
Kubota Z482
2006
0.48
10.9/8.1
Tier 2
Frey and Kuo 2009
4
Confidential Information
TTI 201296
5
Confidential Information
TTI2012
6
Kubota Z482
Storey et al. 200397
Table 2-52 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
x 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.
130
-------
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 criteria emissions (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 in the criteria emissions between the hot and cold conditions.
Table 2-52. In-Use APU Emission Rates
APU
C02
CO
O
X
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:
a Frey 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 2-53.
131
-------
Table 2-53. Average APU Emission Rates from non-DPF APU IDs 1, 2, and 3 according to Cold, Hot, and Mild
Ambient Conditions
co2
CO
O
X
THC
PM
Fuel
Ambient
Temperature
DPF
fe/hr)
fe/hr)
fe/hr)
fe/hr)
fe/hr)
feal/hr)
condition
(°F)
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 2-55.
We estimated elemental carbon (EC) fraction of PM from composition measurements made on
APU ID 1 as reported in Appendix K. 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 2-54. 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 2-55.
Table 2-54. 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 2-55. Fleet-Average Non-DPF Equipped APU Emission Rates in MOVES3
C02
fe/hr)
CO
fe/hr)
NOx
fe/hr)
THC
fe/hr)
PM2.5
fe/hr)
EC
fe/hr)
NonEC
fe/hr)
EC/PM
Fuel
feal/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).98 The APU PM standards along with the current Tier 2 and Tier 4 nonroad standards for
nonroad diesel engines 8
-------
the engine information and emission levels obtained from the publicly available US EPA nonroad
certification database. The development of these rates are described in the HD GHG Phase 2
MOVES documentation and summarized here.100
We anticipate that the APU manufacturers will meet the 2021 PM standard by modifying the
engine control strategy (such as using leaner air fuel mixture) rather than by using an aftertreatment
such as a diesel particulate filter. Such a strategy is likely to lead to increased NOx emissions - the
decrease in PM emissions between the 2012 and 2013 certified APU engines was accompanied by
25 percent increase in NOx emissions. Thus, we estimated a slight NOx disbenefit in obtaining a
lower PM standard. We estimated the in-use APU NOx emissions for 2021-2023 by multiplying the
baseline emissions by 1.25 (15.6 * 1.25 = 19.5 g/hr). We do not anticipate any increases to occur in
CO2, CO, or THC emissions with the 2021 standard, and expect the emissions will not change in
2021 for these pollutants.
To achieve the APU PM standard for MY 2024, we anticipate APU manufacturers will be required
to use DPF aftertreatment. The average PM emission rate from the DPF-equipped APU ID 1 tests
was 0.025 g/hr (Table 2-52), which is similar to the extended idle PM emission rate for 2013+
trucks (Table 2-49) of 0.021 g/hr. We do not believe the data are sufficient to determine a
difference in PM emission rates between APU and main engine extended idling when both engines
are equipped with diesel particulate filters. Thus, we used the MY 2013+ extended idle PM2.5
emission rate as the APU emission PM2.5 emission rate for 2024 and later model years (Table
2-49). We used the EC/PM split measured from the DPF-equipped APU (Table 2-54) 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 2-57.
Table 2-57 APU Emission Rates in MOVES3 with APU PM Controls in the HD GHG Phase 2 Program
Model Year
C02
(g/hr)
CO
(g/hr)
NOx
(g/hr)
THC
(g/hr)
PM2.5
(g/hr)
EC
(g/hr)
NonEC
(g/hr)
EC/PM
Fuel
(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.
133
-------
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+. 101
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.9
For modeling purposes, all glider vehicles are presumed to be combination trucks (sourceTypelD
61 and 62) running on diesel fuel. As detailed in 0, EPA's in-house glider vehicle emission testing
data102 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). CO
from glider vehicles is higher. Based on this analysis, the MOVES running, start, and extended
idling exhaust rates for gliders of all model yearsy are set equal to those of the model year 2000
heavy heavy-duty vehicles.
For example, Figure 2-76 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.
y Glider emission rates are the same for all model years from 1960 through 2060. As discussed in the activity report,
glider activity in MOVES begins in model year 2008.
134
-------
1
2
regClassID B47B49
3
4
5
6
3000-
2000-
1000-
•E
o
TO
DC
o
CD
CD
e
CD
II II II II .. II II II
.. ¦¦ .. >¦ II II II II .. II II II
II II II
II II II
t—I—I—i—i—i—i—i—i—I—i—I—i—i—i—i—I—i—i—r~
MY2008
. .. .i ¦! il il ll .. .i il J ll ll I
¦ ill
jJJ.jjjjJ.
.1 .1 .1 ll ll
-------
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 Exhaust Emissions
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 and later emission rates based on the latest testing data. 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)103, 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
Regulatory
Age group
group
class
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
(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.
136
-------
30
25
•? 20
a.
JZ
=9
m
X 15
O
Z
o 10
o
-1 i-
• CO
«. Nox
~ HC
hiil
¦rhTjT.
Si?
14
12
1
h 0.8 1"
Q.
0.6 5
2
0.4 x
0.2
0
¦"0.2
1980
1985
1995
Model year
2000
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.
137
-------
4000"
3000"
2000-
1000 -
o-
150-
™ 100
.2 50"
0"
500-
A A ~ i A a
~
I
400 "
300-
200-
100 -
0"
• A
• A *
::; i»1 •4: J i1: i ¦" * 1 i4
¦ ¦ ¦ ¦
A
«
A
i tl1
~
~
~
* *
~ ¦ ¦ ¦
« ¦ 1
Model.Years
~ 1960-1989
* 1990-1997
¦ 1998-2007
i—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.
138
-------
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
O
X
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.
139
-------
2500
2000-
1500 "
1000
500
0 J
I 100-
I
!> 75-
! 50
25 ~\
0
200
150-
100 -
50"
0"
A A
~ A
A A
~ . 4
• ~ A A A j. •
~ ~ ~ • ~
A * ~
• I *
~ i
~ A
. ~
* ~ ~ ~
A •
• ~
A A
A
A •
* * * *
* A
A ~ ~
~
i ~ ~
• ~
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 LIID45
3.1.1.1.2 MHD and HHD
Like the 1J ID 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.
140
-------
30000-
20000-
10000 -
3
o
to
w
o-
600"
® 400"
£
c
o
E
uj
200
o-
2000-
1500-
1000-
500-
0-
X
Model, fears
1960-1989
1990-1997
19§8-2007
a
I I 1 I I I I ! I I I i I ! I j ! I
0 1 11 121314151621 2223242527 282930 333537383940
opModeD
Figure 3-4. Emission Rates for MY 1990-1997 at age 0-3 years for Regulatory Class MHD and HHD
Table 3-3 displays the multiplicative age effects by operating mode for MHD, and HHD gasoline
vehicles. While these age effects were derived from the same data as those for the LHD vehicles,
these heavy-duty age effects are slightly different, because the operating modes are defined with
the STP scaling factor of 17.1. For operating modes that do not depend on the scaling factor
(opModelD 0, 1, 11, and 21), the age effects are the same as the LHD age effects. Also, because the
vehicles tested were LHD2b3 and MHD vehicles, no data were available in the high STP power
modes (typically only a HHD truck would reach these). Thus, the higher operating modes
(opModelD 13-16, 24-30, and 35-40) use the same values as the closest operating mode bin with
data.
141
-------
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
O
X
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.
142
-------
7500-
5000-
2500"
o-
2 500-
o
* *
* #
• +
£
400-
300-
age
6+
0-5
. ~ A * . *
* . # i • ~ t ~ * * ~ * t * • ~
A
*
.1 200"
| 100-
"E
UJ 0
1000
750-
500-
250-
~ — i—r~
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40
opModelD
Figure 3-5. Emission Rates by Operating Mode and Age Group for MY 1998-2007 Vehicles in Regulatory Class
MHD and HHD Gasoline Vehicles
3.1.1.2 2008-2009 Model Years
3.1.1.2.1 LHD
The MY 2008 and 2009 LHD emission rates are updated from the MY 2007 LHD emission rates to
account for the phase-in of the Tier 2 and HD 2007 rulemaking which set emissions standards for
medium-duty passenger vehicles (MDPV), Class 2b, and Class 3 chassis-certified vehicles.
Medium duty passenger vehicles fall within the LHD2b3 regulatory class in MOVES. The useful
life emission standards for these vehicles are shown in Table 3-4.
143
-------
Table 3-4 Useful Life FTP Standards from the Tier 2 Rulemaking104 and the HP 2007 Rule106
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
O
X
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.8 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
standard11. Consequently, we relied on the MOVES2014 analysis of in-use Tier 2 Bin 8 vehicles
conducted for the light-duty emission rates.8 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 dependant on the individual power and weight of the vehicle, and should scale approximately
to the g/mile emission standards.2
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.1133 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
O
X
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.8 The "split" ratios for
z 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 tons9, which differs from the fscaie value of 2.06 metric tons.
aa This analysis assumed that 5% of the gasoline LHD2b3 engines were engine-certified, but, actually, all engine-
certified LHD2b3 vehicles are diesel-fueled. However, the engine-certification standard has a small impact on the
aggregate standard used to develop these rates shown in Table 3-5.
144
-------
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
O
X
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 20 09.105 For estimating emission rates in MOVES, we
145
-------
used the same assumptions to estimate the MY 2008 emission rates. The MY 2008 running
emission rates are interpolated between the MOVES 2007 and 2009 emission rates by operating
mode and age group.
opMotieiD
0
25
1
27
11
28
12
— 29
13
— 30
14
33
15
— 36
16
— 37
21
38
22
— 33
23
40
24
12 16 20 24 0 4 8 12 >4048 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
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.bb 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 MOVES2010.106 The rates are modeled
by applying a a 70 percent reduction to the MY 2007 running rates starting in MY 2008, which is
bb Negligible portions are run on other fuels. The figures are aggregated from data supplied by Polk.
146
-------
consistent with the emission standard reduction with the "Heavy-Duty 2007 Rule."107,cc 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 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-dyanamometer 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)
5. Air conditioning on or off
6. Cabin windows down or up
cc The engine-certified standards for heavy-duty gasoline were reduced by 93% (THC), 80% (NOx), and 61% (CO)
with the MY 2008 standard.
147
-------
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 fscaie values (Table 1-3) of 5
(LHD2b3 and LHD45), 7 (MHD), and 10 (HHD). The selection of these new fscaie 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.
148
-------
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.dd
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 fscaie
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 fscaie 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 fscaie 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. 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 rates for MY 2007-2022 from the
MOVES2014 database, 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.108
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.
dd Sales of Class 2b gasoline trucks are much larger than for Class 3, 4, 5, and 6.
149
-------
Table 3-9 Tier 3 Reductions by Model Year for Gasoline LHD2b3
Model Year
THC
CO
X
O
2018
35%
38%
41%
2019
44%
48%
52%
2020
53%
59%
63%
2021
62%
68%
74%
2022-2060
71%
78%
85%
3.1.1.4 Model Year Trends
Figure 3-7 through Figure 3-9 display the THC, CO, and NOx 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.
150
-------
1,5-
©
1
3
©
ro 1-0"
02 Reg Class
|
,| MHD67
I \ LHD45
So 5^ LHD2fa3
i
o
V5
0,0""
Model Year
Figure 3-7. Heavy Duty Gasoline Running Exhaust THC Emission Rates for the 0-3 Age Group by Model Year
and Regulatory Class using Nationally Representative Operating Mode Distributions
The CO emission rates are shown in Figure 3-8. 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.
151
-------
60-
I
3
J8
«
tr
c
g
E
Lit
O
o
40-
20-
Reg Class
HHD8
MHD67
LHD45
— LHD2b3
1
Model Year
....j
Figure 3-8. Heavy Duty Gasoline Running Exhaust CO Emission Rates for the 0-3 Age Group by Model Year
and Regulatory Class using Nationally Representative Operating Mode Distributions
Figure 3-9 shows that the NOx emission rates follow decreasing trends with model years that
correspond with tighter emission standards.
|6-
D3
1
Reg Class
HHD8
MHD67
LHD45
LHD2b3
1970
1980
1990
2000
Model Year
2010
2020
2030
Figure 3-9. Heavy Duty Gasoline Running Exhaust NOx Emission Rates for the 0-3 Age Group by Model Year
and Regulatory Class using Nationally Representative Operating Mode Distributions
152
-------
3.1.2 Particulate Matter (PM2.5)
The available studies from which to develop PM2.5 emission for heavy-duty gasoline are
partiuclarly 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 1960-2009Model 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 medium- and 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/mile107
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 3-10.
Table 3-10 Summary of Data Used in HD Gasoline PM Emission Rate Analysis
Vehicle
MY
Age
Test cycle
GVWR
[lb]
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
153
-------
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 1x10 4
65 22 —
mg/mile for low sulfur fuels). This value leads to the computation of the ratio: —'—— = 1 679.
38.84
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 1x10 5 mg/mile
2 ji "g
for low sulfur fuels, this value leads to the computation of the ratio: —¦—= 0.578.
4.687
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:
154
-------
Ratiofinal = RatioolderWtFrac + Rationewer(l — WtFrac)
= 1.679x0.75 + 0.578x0.25 = 1.40
Equation 3-1
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.ee
3.1.2.1.3 MHD and HHD
For MHD and HHD regulatory classes, the emission rates are based on a fscaie 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 fscaie THC ratios, and the calculated 17.1
fscaie based EC emission rates are displayed in Table 3-11. 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.
ee 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 tons9, which differs from the fscaie value of 2.06 metric tons.
ff 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.
155
-------
Table 3-11. Derivation of MHD and HHD Elemental Carbon Emission Rates from LHD2b3 Rates using fscale
17.1/2.06 THC emission ratios. Using Model Year 2001 as an Example
f scale
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 unepectected 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 and Later 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) 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 PM2.5 rates
from a draft MOVES3 version used for the preliminary Cleaner Trucks Initiative (CTI) analysis.15
156
-------
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 3-12 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 2-28). Since those numbers are
comparable given the uncertainty of the PM2.5 emission rates, and the fact that no modal HD
gasoline PM2.5 data 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 and later
(see Appendix B.8).
Table 3-12 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 CTI analysis15. 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 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.ff LHD45 and LHD2b3 are higher
than 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
2-28). We intend to update the HD gasoline rates to be consistent with the updated HD diesel rates
in an upcoming MOVES update.
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
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.
ff 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 MF1D gasoline PM2.5 rates are lower than the comparable MF1D diesel PM2.5 emission
rates.
157
-------
The MOVES heavy-duty gasoline PM2.5 emission rates are constant for MY 2013 and later model
years. There are differences between the 2010-2012 and 2013-2060 model year groups due to
different tampering and mal-maintenance assumptions applied to the diesel emission rates (see
Section 2.1.2.1), 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.53 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 3-12) 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 Model Year Trends
Figure 3-10 and Figure 311 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.
Reg Class
HHDS
MHD67
— LHD45
— LHD2b3
_Q)
|
30.15-
52
"S
Q£
c
.9.
U 0.10-
£
LU
sz
X
LU
2 0.05
o
I-
csi
0.00-
1970
2000 2010 2020 2030
Model Year
Figure 3-10. Heavy Duty Gasoline Running Exhaust PM2.5 Emission Rates for the 0-3 Age Group by Model Year
and Regulatory Class using Nationally Representative Operating Mode Distributions
158
-------
JB
I
3
©0.015-
15
g Reg Class
| \ HHD8
"E 0,010- \ MHD67
LLi I
!c \ LHD45
lS 1 __ _ LHD2b3
_
+-*
° 0.005-
m
o.ooo-
2010 2015 2020 2025 2030
Model Year
Figure 3-11 Heavy Duty Gasoline Running Exhaust PM2.5 Emission Rates for the 0-3 Age Group by Model Year
and Regulatory Class using Nationally Representative Operating Mode Distributions (Model Year 2007 through
2030 Only)
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.
159
-------
0.0
I
O)
-------
0 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 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.
161
-------
3.1.3.2.1 LHD2b32014-2060Model 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 2-26 (Phase 1) and Table 2-27
(Phase 2) in Section 2.1.4.3.
Table 3-13 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 emissionRateAdjustmenttable 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 Rule109 discussed in
Section 2.1.4.3 and shown in Table 3-14.
Table 3-14 Heavy-Duty Gasoline Reductions due to the Heavy-Duty GHG Phase 1 Rule
109
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 3-15,
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 emissionRateA djustment table in the MOVES database.
162
-------
Table 3-15 Heavy-Duty Gasoline Reductions due to the Heavy-Duty GHG Phase 2 Rule
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 gasolinegg 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-198 5) - 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
88 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.
163
-------
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.
2000-
©
I
3
©
15(
<0
tr
c
0
'm
m
E
1000-
UJ
CM
O
O
O
*E
©
-C
500-
O.
m
O
1
<
Reg Class
HHD8
MHD67
LHD45
LHD2b3
0-
19
I
1990
2 01 £
2020
30
Model Year
Figure 3-14. Heavy-Duty Gasoline Running Exhaust CO2 Emission Rates by Model Year and Regulatory Class
164
-------
E
O
CO
O 10~
®
CL
w
©
Reg Class
HHD8
MHD67
LHD45
— LHD2b3
0"
1970 1980 1990 2000 2010 2020 2030
Model Year
Figure 3-15. Heavy-Duty Gasoline Miles per Gallon by Model Year and Regulatory Class
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 3-17, Table 3-22), 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 standards111 for the Federal Test Procedure (FTP)
are shown in Table 3-16. 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.
165
-------
Table 3-16 FTP Standards (g/hp-hr) for Heavy-Duty Gasoline Engines for Model Years 1990-2008+111
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).103 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 3-17 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 fscaievalues as was done for running exhaust rates. As
166
-------
discussed later, start emission rates are separated by regulatory classes to account for differences in
the emission standards and/or available test data.
Table 3-17 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 (%) was
calculated in terms of the logarithmic mean (xi) as shown in Equation 3-2.
Xg = elnX/ Equation 3-2
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 7-3 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 (xi). 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
167
-------
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 3-3.
s2
~ ~ „-k- Equation 3-3
Aa — -*
-------
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)
169
-------
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 3-18. 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 s, as
shown in Equation 3-4.
S = ^|x|es2(es2 — 1) Equation 3-4
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 3-18 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.
170
-------
Table 3-18. Cold-Start Emission Rates (g) for Heavy-Duty Gasoline Trucks, by Age Group (Italicized Values
Replicated from Previous Age Groups)
Age Group
n
Pollutant
CO
THC
NOx
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
171
-------
300
250
@200
$ 150
2
8 ioo
P=
50
(a) CO
k
10 15
Age (years)
20
25
18
16
14
@12
% 10 -
8 -
6
2
o
o
fi-
ll 4
(b
; ino
^
10 15
Age (years)
20
25
12 -
10 -
§ 8
* c
ex e
2
2
0
(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.
172
-------
3.2.1.1.4 2005-2007Model Yean
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 3-18 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 3-18 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 Ac as shown in
Equation 3-5.
' al4g/hP-hr 1(1.0 g/hp - hr)
f _ v (0.14 + 0.20) g/hp- hr J _n57 Equation 3-5
l\jn — — U.O I
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 3-6.
f =.
-'NOx
0.20 g/hp - hr
(0.14 + 0.20) g/hp - hr
1.0 g/hp - hr
4.0 g/hp - hr
= 0.147
Equation 3-6
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 Yean
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 applied the age effects
used for light-duty truck (LDT) start emission rates in a draft MOVES3 version used for the
preliminary Cleaner Trucks Initiative (CTI) analysis. In that version, we updated the deterioration
173
-------
effects for LDT start NOx exhaust from MOVES2014112,hh by applying the ratios shown in Table
3-19.
Table 3-19. Deterioration Reduction Ratio for NOx Starts from MY 2009-2017 LDT Vehicles
ageGroupID
Draft MOVES3 Verson
/MOVES2014
3
1.00
405
0.85
607
0.79
809
0.73
1014
0.62
1519
0.62
2099
0.62
For THC and CO, the age effects are unchanged from the LDT multiplicative age effects from
MOVES2014.113 The resulting multiplicative age effects for start emission rates for LHD2b3
vehicles used in MOVES3 for model years 2009-2017 are shown in Table 3-20. 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 3-20 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 the LDT deteriorartion rates for starts results in the MOVES 3 start emission rates for MY
2010+ gasoline LHD2b3 vehicles having a higher relative deterioration than running emission rates
(compared to Table 3-3).11 We recognize this is inconsistent with our knowledge of light-duty start
deterioration.8 We plan to address this data gap with data collected on LHD2b3 in future versions
of MOVES.
hh 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.8
11 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.
174
-------
3.2.1.1.6 Incorporating Tier 3 Standards: Model years 2018 and later
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.108 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.
rhe 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. 113'-ii For MOVES3, we adjusted the NOx
start emission rates from those developed in MOVES 3 by applying the deterioration ratios in Table
3-19 to the MOVES2014 NOx start rates. The 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 3-21 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.108
Table 3-21 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
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.
JJ In MOVES3, the deterioration effects for all model year light-duty vehicles were updated on updated data and
analyis. 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.
175
-------
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 3-7.
37.1g/hp-hr
/co " 14.4g/hp - hr ~ ^ Eq,",t">n "
The corresponding ratio for THC for 1990-2004 model year vehicles is 1.73, as shown in Equation
3-8.
„ 1.9e/hp-hr 1 „„
f"c ~ l.lg/hp-hr " 1 73 Eq<,ati0n 3"8
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 3-24 ). Note that the ratios for CO and THC do not vary by model year
group because the standards do not; See Table 3-15.
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 3-24 .
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.114 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 this MOVES3 update includes the
following engine families from the 2016 and 2017 model years shown in Table 3-22.
Table 3-22 Engine Data Analyzed to Revise the Cold Start Emission Rates for HD 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 3-9.
Grams per Start
= [Cold FTP Emission Results (g/(hp - hr)) Fn..ation 3 9
- Hot FTP Emission Results(g/(hp - hr))] 4
* FTP Cycle Work (hp — hr)
176
-------
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 HD gasoline engines determined
the grams per start for THC, CO, NOx, and PM2.5. The mean and standard deviation of the THC,
CO, NOx, and PM2.5 emission levels for the three engines are shown in Table 3-23. 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 HP. The new default cold start emissions values for MOVES3 are the mean
values shown in Table 3-23. The THC, NOx and PM2.5 cold start emissions for HD gasoline
engines are higher compared to MOVES2014, while the CO emissions are lower.
Table 3-23 Cold Start Emissions for MY 2008 and Later Heavy-Duty Gasoline Engines
Grams per Start
THC
CO
NOx
PM2.5
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 THC, CO and
PM2.5 as was used for the previous model year groups (which is based on the gasoline LHD2b3
1960-2004 model years). For NOx, we applied the relative age deterioration as was used for
LHD2b3 vehicles for MY 2008 and later vehicles shown in Table 3-20 and Table 3-21.kk The start
rates for THC, CO, and NOx for this model year group for each age are graphed in Figure 3-21.
3.2.1.3 Summary
Table 3-24 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.
kk 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.
177
-------
Table 3-24 Summary of Cold Start Emission Rates for Heavy-Duty Gasoline Vehicles
Regulatory
Class
Model Year
Group
CO
THC
NOx
LHD2b3
1960-2004
Data analysis, values from
Table 3-18
2005-2007
Data analysis, values
from
Table 3-18
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
LHD45,
MHD,
HHD
1960-2004
Increased
in proportion
to standards from LHD2b3
Same values as
LHD2b3
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
Updated based on FTP
certification data,
deterioration based on
the 2008 + LHD2b3
vehicles
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.
178
-------
•e
m
m 9
3
jg
«
tr
c
.2 R
CO O
®
E
Lit
0
1
CO
ra 3-
0
_
o
o-
Reg Class
HHD8
MHD67
LHD45
— LHD2b3
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.
t 200-
m
ja
3
B
ro
tr
c
o
"m
| 10C
LU
o
o
Reg Class
HHD8
MHD67
LHD45
LHD2b3
1980
2000
Model Year
2020
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.
179
-------
e4" 1
3 I
« is
S '
©3- 1 Reg Class
§ 1 HHD8
g MHD67
m ! LHD45
wo- \
E \ — LHD2b3
UJ
X 'i1
O
o-
1380 2000 2020
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.
0-L . . , , L , , , , . JL , , , . , L, , , , , JL
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.
180
-------
3.2.2 Particulate Matter (PM2.5)
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 3-1 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).
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, andHHD
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 3-23. 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 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
from the 2010 and later model year groups since they are based on certification results from these
engines.
181
-------
0.125-
tT !
15 I
-» 0,100- I
3 :i
© i
•+~*
03 i
0C i (
g 0,075- | ' Reg Class
! 5-, HHD8
E \ I MHD67
So 050- A LHD45
x LHD2b3
w
-t—»
o
In 0.025-
c4
0.000-
1980 2000 2020
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
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 report8 and shown in Figure 2-54. These are the same adjustments used for
heavy-duty gasoline vehicles in MOVES2014.
To evaluate these adjustment ratios for MOVES3, 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, MOVES3 retains the start emission adjustement
ratios used in MOVES2014.
182
-------
HD Gasoline Truck Start Emission Ratio
1.6
1.4
ru
+-»
1/1 12
T3
o
u 1.0
o
O O.s
4->
fU
^ 0.6
£=
o
8 0.4
E
m 0.2
0.0
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
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.58 Figure 2-26 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 MOVES 3 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 3-25 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 using a draft version of MOVES3 used for the preliminary CTI analysis15. Like diesel
vehicles, starts from gasoline vehicles are estimated to be a relatively small contributor to the total
183
-------
energy demand of vehicle operation. Due to the small contribution to the total energy inventory, we
have not prioritized updating the heavy-duty gasoline start emissions rates.
Table 3-25 Relative Contribution of Total Energy Consumption from Each Pollutant Process by Regulatory
Class for Heavy-Duty Gasoline Vehicles in Calendar Year 2016
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 2-43. 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 2-24 and Table 2-26. As discussed in Section 2.2.4, the start energy rates are not
projected to change due to the HD GHG Phase 2 standards.
500 ~
Reg Class
— HHD8
— MHD67
— LHD45
— LHD2b3
o
ffi
f 100-
o
<
o-
1980 2000 2020
Model Year
Figure 3-24 Heavy-Duty Gasoline Cold Start CO2 Emission Rates (g/start) by Model Year and Regulatory Class
(OpmodelD 108)
184
-------
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 prevelant use
of in CNG vehicles has occurred among city transit bus fleets and in solid waste collection or
refuse truck fleets.11' 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 MOVES3.9
30.00%
25.00%
Q
cc
£ 20.00%
_o
Js
%• 15.00%
jh
Tj
*3 10.00%
>
o
z
U 5.00%
Refuse Truck
-Single Unit Short-haul Truck
- 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 for MOVES3 include:
185
-------
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.
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 CH4rate. The methane
fraction from CNG vehicles is 89% and 96% for model year groups 1960-2004 and 2002-2060
respectively, as documented in the Speciation report.1
All criteria 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)116 discussed later in this section. All other conversion
factors to energy were taken from Melendez et al. (2005).120
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.
4.1 Running Exhaust Emission Rates
The pre-2010 running emission rates are relatively unchanged from MOVES201411, and are based
on cycle average rates as discussed in Section 4.1.1. In MOVES3, we updated 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 as discussed in Section 4.1.2.
11 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.
186
-------
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)127 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
Figure 4-2).117 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).
187
-------
25
20
i 15 H
s>
Hi
§; 1°H
5 -
0
CBD
0
100
200
400
500
600
300
Time, s
Figure 4-2 Driving Schedule Trace of the Central Business District (CBD) Cycle118
60
-C
t
40
ft 30
V
3 20
10
0
0
500
1000
1500
2000
Time (sec)
Figure 4-3 Driving Schedule Trace of the Washington Metropolitan Area Transit Authority (WMATA) Cycle120
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.119 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)122 and Ayala et al. (2003).116 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.
188
-------
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 2005120
National Renewable Energy Laboratory
(NREL)
WMATA
2001 (4),
2004 (3)
Ayala 2003116
California Air Resources Board (CARB)
CBD
2000 (4),
2001 (2)
LeTavec 2002121
Atlantic Richfield Company (ARCO)
CBD
2001 (1)
Ayala 2002122
CARB
CBD
2000 (2)
Lanni 2003123
New York Department of Environmental
Conservation
CBD
1999 (3)
McKain 2000124
West Virginia University (WVU)
CBD
1999 (3)
Clark 1997125
wvu
CBD
1996 (10)
McCormick
1999126
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 criteria pollutant rates (NOx, CO, PM2.5 , and THC) 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.
189
-------
4.1.1.2.2
2002-2006 Model Years
Of the surveyed data, only one study had vehicles newer than MY 200 l.mn1,127 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.128
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 fscaie 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.129 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.
mm Several papers have discussed more recent vehicles. Examples include Clark et al. (2007).126 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.
190
-------
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
Group
Number of Engine
Families3,13
Certification Emission Rate (g/bhp-hr)c
NOx
CO
PM2.5
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
2010-20176
159 for NOx and CO, 153
for THC, and 120 for
PM2.5
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 FID CNG engine families.
b Some engine families did not report emission data for TF1C and/or PM2.5.
c MY 2002-2006 group emission rates are projected sales weighted average of FID CNG urban bus certification
emission rates. MY 2007-2009 and 2010-2017 group emission rates are simple average of all FID CNG
certification emission rates (no weighting for projected sales).
d Certification data has measurements of organic material non-methane hydrocarbon equivalent (OMNMF1CE).
For this analysis they were treated as NMF1C values.130
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 F1DIUT 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 vehcles 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.
191
-------
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 Q were similar to those of the tested buses.
350
300
250
S, 200
§3
| 150
P
100
50
I
ii
~1 3 1 3 3 1 1 3 1 1 I 1 1 I I 1 I I I 3
0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 33 40
Operating Mode
Figure 4-4 Operating Mode Distribution for the CBD Cycle
900
800
700
- 600
J* 500
| 400
P 300
200
100
0
r r r r r r r r r r r r r r r r 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 4-5 Operating Mode Distribution for the WMATA Cycle
4.1.1.3.2 Simulating Cycle A verage Emission Rates
With the operating mode distributions determined above, and the emission rates in the
MOVES2010b database, we simulated the gasoline MHD emissions for each pollutant for each
cycle. Dividing by the cycle total distance we calculated the simulated cycle-average distance-
specific rate for that cycle (Esimcycie, g/mile), as shown in Equation 4-1. Using this method, the
simulated cycle emission aggregates were calculated as a function of the following parameters:
192
-------
• fuel type,
• driving cycle,
• age group,
• regulatory class,
• model year, and
• pollutant and process.
I ^p.OM * '''oM.cycle
I'p.simcvcie n Equation 4-1
D
cycle
Where:
Dcycie = distance of the cycle, in miles
Rp,om= emission rate of pollutant p in operating mode OM, in g/hr
Tom,cycle = 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 1960-2006 Model Years
The operating mode based emissions rates for MOVES2010bnn 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
nn The PM exhaust emission rates were subsequently updated from MOVES2010b for MHD gasoline vehicles for pre-
2010 model years.
193
-------
term), which in turn will lead to larger adjustment ratio (Ecng/Emdg). For MY group 2002-2006,
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 ECNGiCBDi1994_2001 Equation 4-2
kCNG,OM, 1994-2001 — KMDG,OM,1997 * ~B
^MDG,simCBD,1997
n _ n 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
deterioration131 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.
194
-------
RcNG.OM, 2007-2009 —
n ECNGiWMATAi2004 CCNG,2007—2009
kMDG,OM,2004 * c * r
^MDG,simWMATA,2004 LCNG,2002-2006
Equation 4-4
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
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 criteria 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.
195
-------
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
PM_NonEC
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
PM_NonEC
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+e
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
cThe 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
e 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.
196
-------
4.1.2 2010-2060Model 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
gaseous pollutants, the 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.
197
-------
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 rule60 (see Table 2-24).
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 rule62 (see Table 2-25).
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 2-25). 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
198
-------
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 CH4 and 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 catatalyst (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.132,133 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~
E
3
©
CO
cr
c
o
"w
m
E
LU
o
X
®
CO
0
"w
¦*-»
o
7.
5.0-
2,5-
Reg Class
HHD8
Urban Bus
0.0-
2000
2010
Model Year
2030
Figure 4-6. Heavy Duty CNG THC Exhaust Emission Rates for the 0-3 Age Group by Model Year and
Regulatory Class using Nationally Representative Operating Mode Distributions
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.132'133
199
-------
15-
3
©
«
tr
c
o
'«
»
E
LU
O
o
10-
Reg Class
HHD8
Urban Bus
2000
Model Year
„20
2030
Figure 4-7. Heavy Duty CNG CO Exhaust Emission Rates for the 0-3 Age Group by Model Year and
Regulatory Class using Nationally Representative Operating Mode Distributions
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.132,133
200
-------
15-
I
310-
®
•*—»
ro
a.:
c
o
'«
.2
E
LU 5-
X
O
Reg Class
HHD8
Urban Bus
V « «
„J0
2010
Model Year
2020
2030
Figure 4-8. Heavy Duty CNG NOx Exhaust Emission Rates for the 0-3 Age Group by Model Year and
Regulatory Class using Nationally Representative Operating Mode Distributions
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
emissions133 and the other study showing stoichiometric engines emitting higher PM2.5 rates.132
201
-------
JB
I
3
©0.075-
15
a:
c
o
w
w
"E 0.050-
LU
£2
X
LU
_
° 0.025-
in
csi
Reg Class
HHD8
Urban Bus
0.000-
J000
2010
Model Year
2020
2030
Figure 4-9. Heavy Duty CNG PM2.5 Exhaust Emission Rates for the 0-3 Age Group by Model Year and
Regulatory Class using Nationally Representative Operating Mode Distributions
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.
i 1000-
E
UJ
CM
o
o
o
'EI
©
JC
Cl
m
Q
500
Reg Class
HHD8
Urban Bus
0
Model Year
2020
2030
Figure 4-10. Heavy Duty CNG CO2 Exhaust Emission Rates by Model Year and Regulatory Class using
Nationally Representative Operating Mode Distributions
202
-------
We have more confidence in the model year 2010 and later emission rates because the emissions
rates are derived directly from in-use second-by-second data. Unfortunately, the differences in the
methodology likely contributed to the differences in emission rates for THC and possibly PM2.5
that are not explained by shift to stoichiometric TWC vehicles for 2010 and later vehicles. Future
updates to MOVES could revisit the pre-2010 CNG exhaust running emission rates to address these
inconsitencies.
4.2 Start Exhaust Emission Rates
In the absence of any measured start exhaust emissions from CNG vehicles, their start rates are
copied from the pre-2010 model year heavy-duty diesel start rates for all pollutants including
energy rates. MOVES still estimates that the majority of emissions from CNG vehicles are from
running emissions, which are based on CNG test programs. We acknowledge that the diesel start
rates may not accurately represent CNG start emissions.
4.3 Ammonia Emissions
Similar to carbon monoxide emissions, CNG vehicles with stoichiometric three-way catalysts have
been shown to emit higher ammonia emissions than CNG vehicles with lean-burn
combustion.132,133 Since ammonia measurements were not part of the HDIUT program used to
update criteria pollutant and particulate matter emission rates, the ammonia emission rates for CNG
vehicles are simply set equivalent to the ammonia emission rates from heavy-duty gasoline
vehicles.4
203
-------
5 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.134
5.1 Background on Hea vy-Duty Diesel Crankcase Emissions
Federal regulations permit 2006-and-earlier heavy-duty diesel-fueled engines equipped with
"turbochargers, pumps, blowers, or superchargers" to vent crankcase emissions to the
atmosphere.135 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'.134
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 136 137 138.
Beginning with 2007-model-year heavy-duty diesel vehicles, federal regulations no longer permit
crankcase emissions to be vented directly to the atmosphere, unless they are included in the
certification exhaust measurements.139 Many manufacturers have adopted open crankcase filtration
systems. 134These 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.134 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.140
A summary of published estimates of diesel crankcase emissions as percentages of the total
emissions (tailpipe + crankcase) are provided in Table 5-1. 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 emissions, resulting in an increase in the relative
crankcase contribution for THC, CO, and PM2.5 emissions.
204
-------
Table 5-1 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, 1977143
1973
None
2
3.9%
0.4%
0.1%
2.8%
Zielinska et al. 2008136
2000,
13.5% -
Ireson et al. 2011137
2003
None
2
41.4%
Clark et al. 2006142
Clark et al. 2006141
2006
None
1
3.6%
1.3%
0.1%
5.9%
Khalek et al. 200948
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 tailpipe and crankcase exhaust. The crankcase ratios used in MOVES
are a ratio of crankcase to tailpipe exhaust.
5.2 Modeling Crankcase Emissions in MO VES
MOVES calculates crankcase emissions using a gaseous and a particulate matter crankcase
emission calculators. 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 the crankcase emission rates to the tailpipe exhaust
emission rates, but it does not change the tailpipe exhaust emission rates. Unlike the gaseous
calculator, the particulate matter calculator accounts for tailpipe MOVES exhaust emission rates in
the emissionratebyage table (processID 1, 2, 90) that may include crankcase emissions and has the
ability to divide the exhaust emission rates into components representing the contributions from
tailpipe exhaust and crankcase emissions. Exhaust ratios are used to adjust the particulate matter
tailpipe exhaust emission rates to account for the crankcase contribution. For most model years, 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. However, the 2007-2009 model
year tailpipe particulate matter emission rates are assumed to include crankcase emissions, and the
exhaust ratios are less than one to account for the contribution as discussed in Section 5.3.2. More
details on the crankcase calculator 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 5-2. 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
205
-------
Table 5-2 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
O
X
NO, N02) HONO
PM2.5
EC, SO4, H20(aerosol), NonECNonSC^PM
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 emissons. 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 emisson control system. If crankcase emissions do not exhibit the
same increase in deterioration as tailipipe emissions, this method would lead to an overestimation
of crankcase emissions in older vehicles..
5.3 Hea vy-Duty Diesel Crankcase Emissions
Different crankcase emission ratios are derived for pre-2007, 2007-2009, and 2010 and later heavy-
duty vehicles as documented in the following sections. The model year distinctions capture large
changes in the requirements for crankcase emissions, as well as the control of tailpipe exhaust
emissions which impact the crankcase to tailpipe emission ratios.
As discussed in the following sections, the data on crankcase emissions are limited. For example,
the 2010 and later heavy-duty diesel crankcase emission rates are based on tests from two vehicles.
As such, the data have considerable uncertainties. In general, the crankcase emission rates are only
a small fraction of the tailpipe emissions as presented in the following sections, so the uncertainties
in crankcase emissions do not extend to total exhaust emissions. For glider trucks, we had no
crankcase emissions data, we conducted sensitivity testing to evaluate the impact of the
assumptions used in MOVES on the heavy-duty emissions inventory as discussed in Section 5.3.4.
5.3.1 1960-2007Model Years
Table 5-3 displays the crankcase/tail-pipe emission ratios used for pre-2007 diesel exhaust. For
THC, CO, and NOx, we selected the values measured on the MY2006 diesel engine reported by
Clark et al. 2006.142 These values compare well with the previous HC, CO, NOx values reported
much earlier by Hare and Baines (1977),143 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 well represented as a fraction of the exhaust
emissions. The THC crankcase ratios presented in Table 5-3 are also used for methane, total
206
-------
organic gases (TOG), non-methane hydrocarbons (NMHC) and all the chained pollutants to
NMHC listed in Table 5-2.
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. 2008136
and Ireson et al. 2011137 reported crankcase contributions to total PM2.5 emissions as high as 40
percent. Jaaskelainen (2012)134 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.144 The crankcase emission ratios shown Table 5-3 are applied to running, start
and extended idle exhaust to estimate the corresponding crankcase exhaust emissions.
Table 5-3 MOVES pre-2007 Diesel Crankcase Ratios for HC, CO, NOx, and PM2.5
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
PM2.5
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
for important differences in the PM speciation between tailpipe and crankcase emissions.
The pre-2007 diesel ratios are derived such that the total crankcase PM2.5/tailpipe exhaust PM2.5
ratio is 20 percent, and the crankcase emissions EC/PM fraction reflects measurements from in-use
crankcase emissions. Zielinska et al. 2008136 reported that the EC/PM fraction of crankcase
emissions from two pre-2007 diesel buses is 1.57 percent. 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.136,137 The crankcase emission factors shown in Table 5-4 are derived such that the
crankcase PM2.5 emissions are 20 percent of the PM2.5 exhaust measurements 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 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).
207
-------
Table 5-4. MOVES Exhaust and Crankcase Ratios for Pre-2007 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
go
O
Exhaust
1
1
1
H20
1
1
1
EC
0.007
0.004
0.007
nonECnon SO4PM
Crankcase
0.367
0.937
0.367
go
O
0.367
0.937
0.367
H20
0.367
0.937
0.367
5.3.2 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 Section
5.3.2.
The crankcase ratios for 2007-2009 HDD emissions are based on the ACES Phase 1 study140,
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 a 125 hours of degreening process before the test program. Thus, they represent low-
mileage, 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 samples of the hot FTP cycles that included tailpipe exhaust emissions and repeat tests
that included both the tailpipe exhaust and the crankcase exhaust routed to the sampling system.
The crankcase emission rates for each engine were calculated by taking the difference in the total
mass emitted between hot-FTP cycles with and without crankcase routed into the sampling system
and then dividing by the testing time as shown in Equation 5-1.
CCemissions,i — Exhaustwjthcc,i — Exhaustw/0cQj Equation 5 1
Where:
^emissions,i= crankcase emissions (grams per hour) for engine i
Exhaustwithcc,i=hot-FTP cycle average emission emissions (grams per hour) for each
engine i with the crankcase routed into the sampling system
Exhaustw/oCC i= hot-FTP cycle average emission emissions (grams per hour) for each
engine i without the crankcase
208
-------
Then the crankcase emission rates were averaged together in Equation 5-2 with the backup engine
being treated as additional tests of engine B.
CC
(CCengjneA "I"
CCengineB CCengineB'.sitel '-"'-'engineB',site2
'ACESavg
+ CCengineC + CCengineD) Equation
5-2
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. Based upon
data from 2010+ HDD vehicles using the same crankcase technology, the crankcase emission rates
for extended idling are similar to the running rates, so we also use the hot-FTP rates from ACES
Phase 1 for the crankcase extending idling rates. The crankcase rates are listed in Table 5-5.
Equation 5-3 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
Equation 5-3
Where each pollutant rate is in g/hr and the ratio of NMOG/NMHC is 1.343 from the hcspeciation
table in MOVES.1
Table 5-5 The MY 2007 crankcase exhaust rates (CCACESavg) from ACES Phase 1 FTP cycle used for running and
CO
O
X
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 5-1 are calculated using Equation 5-4 where
the ExhaustACES is calculated using Equation 5-2, but using Exhaustw/oCC,i in place of the
crankcase values.
CC
CC
ACES ratio —
ACESavg
Equation 5-4
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 in short and
long-haul single-unit and combination trucks (sourcetypes 52,53,61,62) in each operating mode
estimated from a MOVES national scale run. The values for ExhaustMOvES are listed in Table 5-6.
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.
209
-------
Table 5-6: The MY 2007-2009 MOVES exhaust base rates (g/hr) for running (weighted by operating mode
activity) and extended idling
CO
O
X
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 5-5 to estimate the base crankcase to tailpipe exhaust ratio for 2007-2009
HD vehicles.
CCAGESavg Equation 5-5
'"'-•base " ExhaUStM0VES
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 fleet-wide crankcase emissions, 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.
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.00 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 vehicle weight categories and across categories for a heavy-duty diesel
total as shown in Table 5-7. 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 67% of all heavy-duty engines produced in these
model years had open crankcase systems. The average value across model years and regulatory
classes was used for all 2007 and later engines.pp
00 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.
pp 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 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
results.
210
-------
Model Year
LHD
MHD
HHD
MY Total
2016
0
0.912
0.725
0.696
2017
0
0.919
0.635
0.663
2018
0
0.884
0.640
0.660
average
0
0.905
0.666
0.672
Finally, to estimate the MY 2007-2009 crankcase emission ratios for MOVES, we assume that
crankcase emissions are proportional to the exhaust emissions across regulatory classes and source
types. In this calculation, we must account for the fact that many LHD2b3 trucks are chassis-
certified rather than engine-certified, and required to have closed crankcases. To estimate the final
crankcase emissions for each source type, we multiplied the open crankcase ratio, CCratio base, by
the open crankcase fraction of all heavy-duty regulatory classes, OpenCCfrac, and by the fraction of
engine-certified vehicles, engine_certratio,sourcetype,modeiyear as shown in Equation 5-6.
CC
ratio,Sourcetype,modeiyear
^ratio,base * OpenCCfrac
(engine_cert
ratio,Sourcetype,m
odelyear)
Equation 5-6
Where
CCratio,sourcetype,model year= the crankcase ratio used in MOVES by source type and model
year
CCratio,base= the heavy-duty open crankcase ratio calculated from Equation 5-5
OpenCCfrac= the fraction of open crankcase systems determined from certification data and
manufacture production volume as shown in Table 5-7 =0.672
engine_certratio,Sourcetype,modelyear= the fraction of engine-certified vehicles within each
source type and model year, as explained below.
The MOVES 2007-2009 crankcase ratios for THC, CH4, NMHC, TOG, NOx, CO, and PM2.5 for
each process for heavy-duty source types with 100 percent engine-certified vehicles are shown in
Table 5-8.
211
-------
Table 5-8 MOVES 2007 to 2009 Diesel Crankcase/Tailpipe Ratios for the Heavy-duty Source Types
(41,42,43,51,54,61,62) except the short and long-haul single-unit trucks (source types 52,53)
Crankcase/Tailpipe ratio
Baseline
Final values to
Process
Pollutant
values
(Equation
5-5)
account for closed
crankcase
(Equation 5-6)
THC
0.60
0.40
NMHC
1.14
0.77
Crankcase
ch4
0.22
0.15
Running
TOG
0.67
0.45
Exhaust
CO
0.71
0.48
O
X
0.003
0.002
PM2.5
0.04
0.03
THC
0.00
0.00
NMHC
0.00
0.00
Crankcase
ch4
0.00
0.00
Start
TOG
0.00
0.00
Exhaust
CO
0.00
0.00
O
X
0.00
0.00
PM2.5
0.00
0.00
THC
0.21
0.14
Crankcase
Extended
Idle
Exhaust
NMHC
0.39
0.26
ch4
0.08
0.05
TOG
0.23
0.15
CO
0.21
0.14
O
X
0.008
0.005
PM2.5
0.44
0.29
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, MOVES uses the
same speciation profile for both crankcase and tailpipe emissions, and all intermediate PM2.5
species. The resulting crankcase emission ratios for MY 2007 to 2009 heavy-duty diesel by
intermediate PM2.5 species are provided in Table 5-9 and the tailpipe ratios are calculated as one
minus the crankcase ratio. As explained in Section 5.2, the exhaust crankcase emission factor is
less than one for 2007-2009 diesel vehicles to account for the contribution of crankcase emissions
in the base exhaust emission rates.
212
-------
Table 5-9. MOVES Exhaust and Crankcase Ratios for Model Year 2007-2009 Diesel by Pollutant and Process
for PM2.5 Species for Heavy-duty Source Types (41,42,43,51,54,61,62) except the Short and Long-haul Single-
unit Trucks (source type 52,53)
Process
Model Year Group
Pollutant
Start
Running
Extended Idle
Tailpipe
Exhaust
2007-2009
EC
1
0.971
0.7071
nonECnonS04PM
1
0.971
0.7071
go
O
1
0.971
0.7071
H20
1
0.971
0.7071
Crankcase
2007-2009
EC
0
0.0290
0.2929
nonECnon SO4PM
0
0.0290
0.2929
go
O
0
0.0290
0.2929
H20
0
0.0290
0.2929
The fractions of engine-certified vehicles in single-unit short- and long-haul trucks (sourceType 52
and 53) are calculated using Equation 5-7. We assumed that 5.1 percent of the diesel vehicles in
LHD2b3 are engine-certified based on the national vehicle registration data and the vehicle activity
data. The remainder of the LHD2b3 are chassis-certified vehicles which must comply with light-
duty standards requiring closed crankcase systems. The other heavy-duty regulatory classes contain
only engine-certified vehicles. The MOVES sampleVehiclePopulation table is used to estimate the
fraction of LHD2b3 vehicles in each source type for each model year. The fraction of LHD2b3 in
diesel-fueled short-haul and long-haul combination trucks (sourceType 52 and 53) by model year
are shown in Table 5-10. None of the other heavy-dutyqq source types include LHD2b3 vehicles for
2007 and later model years.
engine_certrati0 Sourcetype modeiyear Equati
— [l — LHD2b3sourcetXype mocjei year x (l — engine_certratio LHDD)]
Where
engine_certratio Sourcetype modelyear= the fraction of engine-certified vehicles within sourcetype
52 and 53 by model year (Column C in Table 5-10)
LHD2b3sourcetype mocjei year= the fraction of LHD2b3 diesel vehicles within each source
type and model year group of diesel vehicles (Column A in
Table 5-10)
engine_certratio LHDD= the fraction of engine-certified vehicles within light-heavy-duty
diesel (LHDD) vehicles (Column B in Table 5-10)
qqNote that two of the light-duty source types (passenger truck and light-commercial trucks) contain a small fraction of
LHD2b3 vehicle, but these vehicles are all expected to have closed crankcase systems as documented in the light-duty
exhaust report8
213
-------
Table 5-10 Calculation of the Engine Certification Ratio for Diesel Short- and Long-haul Combination trucks
(A)
(B)
(C)
Model Year
Fraction of LHD2b3
Engine cert ratio within LHD2b3
Engine cert ratio
2007
33.3%
5.1%
68.4%
2008
47.3%
5.1%
55.2%
2009
32.9%
5.1%
68.8%
2010
39.8%
5.1%
62.3%
2011
55.2%
5.1%
47.6%
2012
50.5%
5.1%
52.1%
2013
42.3%
5.1%
59.9%
2014
35.8%
5.1%
66.1%
2015+
35.8%
5.1%
66.1%
Using Equation 5-6, we multiply the engine cert ratios shown in Table 5-10 by the crankcase ratios
shown in Table 5-8 and Table 5-9. This reduces the crankcase emission ratios for source type 52
and 53 by approximately one-third to one-half depending on the model year.
5.3.3 2010-2060Model Years
For 2010-and-later diesel vehicles, the HDIUT program (see Section 2.1.1.1) is used for the
baseline exhaust emission rates. 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 (NFVEL) 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 an ARB transient (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 5-1.
214
-------
Figure 5-1 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 5-11.
Table 5-11: 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 5-12 reports the averages of the repetitions by truck and phase, and by phase only (by
weighting the two trucks equally).
215
-------
Table 5-12 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 used for the preliminary CTI analysis for calendar year 2015.15 Phase 2
had transient operation below 50 mph and was matched to operating modes 0-30 (36.7% of total
activity) and Phase 4 had constant high speed data and was matched to operating-modes 33-40
(63.3% of total activity) as shown in Table 5-13. HHD diesel vehicles in short- and long-haul
single-unit and combination trucks vehicles were chosen as to match the NVFEL testing vehicles.
216
-------
Table 5-13. 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 (STP^,
skW)
Vehicle
Speed
(vt, 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^O
1 < vt< 25
4.55
12
Cruise/Acceleration
0 < STPk 3
1 < vt< 25
6.62
13
Cruise/Acceleration
3 < STPk 6
1 < vt< 25
1.64
14
Cruise/Acceleration
6 < STPk 9
1 < vt< 25
1.00
15
Cruise/Acceleration
9 < STPfC 12
1 < vt< 25
0.77
16
Cruise/Acceleration
12 < STP^
1 < vt< 25
1.36
21
Coast
STP^O
25 < vt < 50
3.37
Phase 2
22
Cruise/Acceleration
0 < STPk 3
25 < vt < 50
1.34
23
Cruise/Acceleration
3 < STPfC 6
25 < vt < 50
1.54
24
Cruise/Acceleration
6 < STPk 9
25 < vt < 50
1.52
25
Cruise/Acceleration
9
-------
2f=l(CCphasel,j -CCphase2,j) Equation5-8
^start —
Where
CCstart= average crankcase (g/start) emission from all the test runs
CC phase i,j = crankcase (g/hr) emission rate from test run j and Phase 1
n= the number of tests (13)
In the case of THC and NOx, Equation 5-8 yielded negative start emission rates, because the
average Phase 2 (g/hr) emission rates for these pollutants are higher than the Phase 1 (g/hr)
emission rates. There are physical explanations that can lead to an observed increase in crankcase
emission rates in Phase 2. Lubricating oil and diesel fuel in the crankcase can evaporate as the
engine block heats during operation, which is measured as THC and PM2.5 if the vapors condense
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. Similarly, we set the
start crankcase emission rate to zero for THC and NOx. The average CH4 start emission rate was
also positive, but because methane is estimated in MOVES as a fraction of THC, we also set it
equal to zero as shown in Table 5-14.
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 5-3 to
calculate TOG crankcase emissions, with the MOVES NMOG/NMHC ratio for 2010+ diesel
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.145 Thus, by using tailpe 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
5-2).
218
-------
Table 5-14 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)
0.268
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 5.3.2. 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.
Table 5-15 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
O
X
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
As was done for the development of the 2007-2009 crankcase emission ratios documented in
Section 5.3.2, we assume that the crankcase ratio derived for 2010+ heavy heavy-duty diesel
(HHD) engines applies to the other heavy-duty regulatory classes (MHD, LHD) 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 5-6 to account for the fraction of open crankcase systems within
each source type, with the fraction of open crankcase systems assumed to be 67.2%. The fraction of
engine-certified vehicles among short and long-haul single-unit trucks (source type 52 and 53)
varies by model year, see Table 5-10.
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 5-16. For long-haul and short-haul single unit trucks
(sourceType 52 and 53) these values are multiplied by the fraction of engine certified vehicles in
each model year, provided in Table 5-10.
219
-------
Table 5-16 MOVES 2010+ Diesel Crankcase/Tailpipe Ratios for Heavy-Duty Sourcetype (41,42,43,51,54,61,62)
Except the Short- and Long-haul Single-unit Trucks (Sourcetype 52,53)
Crankcase/Tailpipe ratio
Process
Pollutant
Model Year
Baseline
values
(Equation 5-5)
Final values to account for closed crankcase
Group
(Equation 5-6)
THC
2010-2013
0.36
0.24
2014+
0.37
0.25
NMHC
2010-2013
0.57
0.39
2014+
0.58
0.39
ch4
2010-2013
0.016
0.011
2014+
0.016
0.011
Running
TOG
2010-2013
0.37
0.25
2014+
0.38
0.25
CO
2010-2013
0.003
0.002
2014+
0.007
0.005
NOx
2010-2013
0.024
0.016
2014+
0.024
0.016
PM2.5
2010-2013
0.26
0.18
2014+
0.32
0.22
THC
2010-2013
0
0
2014+
0
0
NMHC
2010-2013
0
0
2014+
0
0
ch4
2010-2013
0
0
2014+
0
0
Starts
TOG
2010-2013
0
0
2014+
0
0
CO
2010-2013
0.89
0.60
2014+
0.89
0.60
NOx
2010-2013
0
0
2014+
0
0
PM2.5
2010-2013
0
0
2014+
0
0
THC
2010-2013
0.17
0.11
2014+
0.28
0.19
NMHC
2010-2013
0.26
0.18
2014+
0.44
0.30
ch4
2010-2013
0.006
0.004
2014+
0.009
0.006
Extended
TOG
2010-2013
0.17
0.11
Idle
2014+
0.28
0.19
CO
2010-2013
0.006
0.004
2014+
0.006
0.004
NOx
2010-2013
0.017
0.011
2014+
0.017
0.011
PM2.5
2010-2013
1.11
0.75
2014+
1.82
1.22
220
-------
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 5-17. Since PM2.5 uses ACES Phase 1 data,
starts are assumed to be included in the running and the crankcase starts are set to zero.
Table 5-17. MOVES Exhaust and Crankcase Ratios for 2010+ Diesel by Pollutant, Process, and Model Year
Group for PM2.5 Species for the Heavy-duty Source types (41,42,43,51,54, 61,62)
Except the Short- and Long-haul Single-unit Trucks (Source type 52,53)
Process
Model Year Group
Pollutant
Start
Running
Extended Idle
EC
1
1
1
2010-2013
nonECnonS04PM
1
1
1
go
O
1
1
1
Tailpipe
H20
1
1
1
Exhaust
EC
1
1
1
2014+
nonECnonS04PM
1
1
1
00
O
1
1
1
H20
1
1
1
EC
0
0.1761
0.7465
2010-2013
nonECnonS04PM
0
0.1761
0.7465
00
O
0
0.1761
0.7465
Crankcase
H20
0
0.1761
0.7465
EC
0
0.2168
1.2203
2014+
nonECnonS04PM
0
0.2168
1.2203
00
O
0
0.2168
1.2203
H20
0
0.2168
1.2203
5.3.4 Glider Crankcase Emissions
Glider trucks in MOVES are anticipated to have emissions equivalent to those from MY 2000 HD
diesel engines (Section 2.5). As such, the crankcase fractions derived for pre-2007 vehicles should
also be used to estimate crankcase emissions from vehicles in the glider regulatory class. However,
MOVES models crankcase emissions by sourcetype (e.g., long-haul combination truck) and model
year, rather than by regulatory class (e.g., Glider, HHD). As such, glider vehicles use the same
crankcase ratios as the other HHD and MHD vehicles in the long-haul and short-haul combination
truck sourcetypes and model year groups.
In Appendix L, we evaluated the impact of accounting for gliders in the crankcase ratios for several
calendar years (2015, 2020, 2030, and 2045) using current estimates of glider populations9. The
impact on crankcase emissions from gliders is significant for THC, CO, and NOx emissions.
However, looking at the broader inventory, the difference for total emissions of heavy-duty diesel
vehicles for CO, NOx and Total PM2.5 is less than a 0.2%, and for THC, the difference is less than
3%.
221
-------
By not explicitly accounting for glider crankcase ratios, we are unable to accurately speciate the
PM from glider crankcase emissions (i.e., elemental carbon, organic carbon, and other species).
This causes the EC/PM fraction from crankcase emissions for glider trucks in MOVES3 to be
dominated by EC emissions, whereas the EC/PM fraction of crankcase emissions from pre-2007
HD trucks is anticipated to be less than 1%. When aggregated to all HD regulatory classes and
exhaust emission processes, this causes the EC/PM fraction to be slightly higher in MOVES3, by
1% in 2015 to 6% in 2045. See more details in Appendix L. We anticipate using this information as
we consider revisions to the crankcase calculations in future versions of MOVES.
5.4 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 5-18; these are documented in the MOVES
light-duty emission rates report.8
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 5-9. The resulting
fleet-wide crankcase to exhaust emission ratios for 1969 and later vehicles are shown in Table
5-18. 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.
Crankcase Ratio1969+ = PCV failure rate (4%) x Crankcase Ratioprel969 Equation 5-9
Table 5-18 Crankcase to Tailpipe Exhaust Emission Ratio for Heavy-Duty Gasoline and CNG Vehicles for HC,
CO, NOx, and PM2.5
Pollutant
pre-1969
1969 and later
HC
0.33
0.013
CO
0.013
0.00052
O
X
0.001
0.00004
PM (all species)
0.20
0.008
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 5-19. No information is
available to estimate separate speciation between exhaust and crankcase, so the factors are the same
for all PM subspecies.
222
-------
Table 5-19 MOVES Exhaust and Crankcase Ratios for Heavy-Duty Gasoline and CNG Vehicles by Pollutant,
Pollutant
Process
1960-1968
1969-2050
EC
Exhaust
1
1
nonECnonS04PM
1
1
00
O
1
1
h2o
1
1
EC
Crankcase
0.2
0.008
nonECnonS04PM
0.2
0.008
00
O
hi*
0.2
0.008
h2o
0.2
0.008
223
-------
6 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 + NO2.146,147 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.146 NOz compounds are formed in the atmosphere
as oxidation products of NOx.147
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. N02is 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.148
MOVES produces estimates of 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, for the heavy-duty vehicle source
types, the NOx fractions vary only according to fuel type, model year, and emission process. The
NOx fractions in MOVES were developed from a literature review from emission test programs
conducted in the laboratory with constant volume sampling dilution tunnels.4 As for crankcase
emissions, we will consider updating the MOVES structure to use separate NO/NO2 and NO2/NOX
fractions by regulatory class to accommodate different emissions from gliders from other heavy-
duty regulatory classes.
MOVES also produces estimates of 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.149 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 unity.
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
224
-------
MOVES users developing air quality inputs of NO, NO2, and HONO, should estimate NOx as the
sum of NO + NO2 (pollutantlDs 32 and 33), rather than using the direct NOx output in MOVES
(polluantID 3).
Future work is needed to (1) update MOVES to model NOx and HONO fractions according to
regulatory class, (2) update the NOx and HONO fractions in MOVES based on more recent
measurements, (3) reconcile the definition of NOx in MOVES, while also correctly accounting for
the emissions of NOz species that may impact NOx measurements and (4) reconcile measurement
differences that may occur between NOy species measured at the tailpipe, with NOy species
measured on road side measurements.150
6.1 Hea vy-Duty Diesel
Table 6-1 shows the NOx and HONO fractions for heavy-duty diesel vehicles. The 1960-2006
model year heavy-duty diesel NOx fractions were estimated as the average reported fraction from
three studies of heavy-duty vehicles not equipped with diesel particulate filters.4 The 2010+ NO2
fractions are based on the average from three diesel test programs of diesel vehicles measured with
diesel particulate filters. The 2007-2009 values are an average of the 1960-2006 and 2010-2050
values, based on the assumption that the NOx fractions changed incrementally, as trucks equipped
with catalyzed diesel particulate filters were phased-into the fleet. We intend to incorporate data on
NOx fractions from SCR-equipped diesel trucks in the next version of MOVES.
The NOx fractions are the same across all diesel source types (including light-duty) and across all
emission processes (running, start, extended idle), except for auxiliary power units, which use the
conventional NOx fractions (1960-2006) for all model years because it is assumed that the APUs
are not fitted with diesel particulate filters. The NO2 fractions originally developed from a
MOVES2010 Report4 were reduced by 0.008 to account for the HONO emissions. 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 pre-2007 NOx fractions are
more relevant for this regulatory class.
Table 6-1 NOx and HONO Fractions for Heavy-Duty Diesel Vehicles
Model Year
NO
N02
HONO
1960-20063
0.935
0.057
0.008
2007-2009
0.764
0.228
0.008
2010-2060
0.594
0.398
0.008
Note:
a All Model Year of Auxiliary Power Units (APUs) use the 1960-2006 NOx and HONO fractions
6.2 Heavy-Duty Gasoline
The NOx fractions for heavy-duty gasoline are based on the MOVES values used for light-duty
gasoline measurements. Separate values are used for running and start emission processes. As
stated in the MOVES2010 report4 the values are shifted to later model year groups to be consistent
with emission standards and emission control technologies. These values are shown in Table 6-2
for both light-duty and heavy-duty gasoline vehicles. The NO2 fractions originally developed from
the MOVES2010 report4 were reduced by 0.008 to account for the HONO emissions.
225
-------
Table 6-2 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
6.3 Compressed Natural Gas
We used the average of three NO2/ NOx fraction reported on three CNG transit buses with DDC
Series 50 G engines by Lanni et al. (2003)123 with the 0.008 HONO fraction assumed for other fuel
types, to estimate the NOx fractions of NO, NO2, and the HONO fraction. These assumptions yield
the NOx and HONO fractions in Table 6-3, which are used for all model year CNG heavy-duty
vehicles.
Table 6-3 NOx and HONO Fractions CNG Heavy-Duty Vehicles
Model Year
NO
N02
HONO
1960-2060
0.865
0.127
0.008
226
-------
7 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
228
-------
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 MOVES2010151
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 MOVES2014.38 No changes were made to the T&M assumptions or data
between MOVES2014 and MOVES3. We acknowledge that T&M adjustment factors in MOVES
should be re-evaluated and updated in the future, particularly to incorporate data on the durability
and emissions performance of advanced aftertreatment systems on modern heavy-duty diesel
vehicles.
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
included in Appendix C.
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.
229
-------
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 i, and the corresponding T&M emission effect, as shown in Equation 7-1.
fr&M,p = / < (T&M frequency; x T&M emission effectp i) Equation 7-1
Where:
^t&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 7-2.
230
-------
EREnd of useful life,p,r,o — ERzero mile,p,r,o ^ (l ^T&M.p)
Equation 7-2
Where:
EREnd of warranty,] = the heavy-duty diesel emission rate at the end of warranty for each
pollutant p, regulatory class, r, and operating mode, o
ERZero mile= the zero-mile heavy-duty diesel emission rate for each pollutant p, regulatory
class, r, and operating mode, o
fx&M= the tampering and mal-maintenance adjustment factor for each pollutant p (Equation
7-1)
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 VIUS152 (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.
Regulatory class
Warranty agea
(Requirement:
100,000 miles or 5 years)
Useful life mileage/age
requirement
Useful life
agea
HHD
1
435,000/10
4
MHD
2
185,000/10
5
LHD45
4
110,000/10
4
LHD2b3
4b
110,000/10
4
BUS
2
435,000/10
10
Notes:
aThe 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 LHD2b3 diesel vehicles actually have a warranty of 50K miles. We realized later that MOVES emission rates for
LHD diesel were incorrectly developed using a warranty period of 100K instead of 50K. We will update the warranty
period for LHD diesel in a future update to MOVES.
While both age and mileage metrics are given for these periods, whichever comes first determines
the applicability of the warranty. As a result, since MOVES deals with age and not mileage, we
needed to convert all the mileage values to age equivalents, as the mileage limit is usually reached
before the age limit. The data show that on average, heavy heavy-duty trucks accumulate mileage
much more quickly than other regulatory classes. 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 the scaled age effect 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
231
-------
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.
Table B-2. Calculation of sa
Where:
0
age < end of warranty age
{age — end of warranty age)
end of warranty age < age < useful life
{Useful life age — end of warranty age)
1
age > useful life
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. We assumed that there is 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 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-3 shows the average scaled age effect by age group. In this table, a value
of 0 indicates no deterioration, or zero-mile emissions level (ZML), and a value of 1 indicates a
fully deteriorated engine, or maximum emissions level, at or beyond useful life (UL). We
multiplied the scaled age factor by the T&M adjustment factor of each pollutant and age in
Equation 7-1.
Table B-3. Scaled Age Effect, Sa
Age Group
LHD
MHD
HHD
Bus
0-3
0
0.083
0.25
0.03125
4-5
1
0.833
1
0.3125
6-7
1
1
1
0.5625
8-9
1
1
1
0.8125
10-14
1
1
1
1
15-19
1
1
1
1
20+
1
1
1
1
ERp,r,a,o — ERzero mile,p,r,o ^ (1 ^ ^T&m) Equation 7-3
Where:
ERp ri0,a = the heavy-duty diesel emission rate for each pollutant p, regulatory class r, age a,
operating mode, o,
ERzero mile= the zero-mile heavy-duty diesel emission rate for each pollutant p, regulatory
class r, operating mode, o
sa= scaled age effect at age a
fx&M= the tampering and mal-maintenance adjustment factor (Equation 7-1)
232
-------
Sections B.2 through B.9 discuss the data sources and assumptions used to determine the T&M
failure frequencies and T&M emission effect for each pollutant and model year range of vehicle,
used to derive the T&M adjustment factor in Equation 7-1.
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 Memo153 (2006). The
basic EMFAC occurrence rates for tampering and mal-maintenance were developed
from Radian and EFEE reports and internal 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 Categories
EPA generally adopted the categories developed by CARB, with a few exceptions. The high fuel
pressure category was removed. We added a category for misfueling to represent the use of
nonroad diesel in cases when ULSD onroad diesel is required. We combined the injector categories
into a single group. We reorganized the EGR categories 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 category.
EPA grouped the LHDD, MHDD, HHDD, and Diesel bus groups together, except for model years
2010 and beyond. We assumed that the LHDD group will primarily use Lean NOx Traps (LNT) for
the NOx control in 2010 and beyond. On the other hand, we also assumed that Selective Catalyst
Reduction (SCR) systems will be the primary NOx aftertreatment system for HHDD. Therefore, the
occurrence rates and emission impacts will vary in 2010 and beyond depending on the regulatory
class of the vehicles.
233
-------
B.4 T&M Model 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.
B. 5 T&M Failure Frequency Rates and Differences from
EMFAC2007
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 hardware would evolve through 2010, rather than be replaced with
completely new systems that would justify a higher rate of failure. We assumed that many of the
2010 changes would occur with the aftertreatment systems which are accounted for separately.
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 LDV I/M program154 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-4).
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
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.
234
-------
Table B-4. 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
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.
235
-------
B. 6 Tampering & Mal-maintenance Failure Frequency Rate
Summary
Table B-5. T&M Failure Frequency Rate by Model Year Group
Frequency Rates
1994-97
1998-2002
2003-2006
2007-2009
2010+ HHDT
2010+ LHDT
Timinq Advanced
5%
2%
2%
2%
2%
2%
Timinq Retarded
3%
2%
2%
2%
2%
2%
Injector Problem (all)
28%
28%
13%
13%
13%
13%
Puff Limiter Mis-set
4%
0%
0%
0%
0%
0%
Puff Limiter Disabled
4%
0%
0%
0%
0%
0%
Max Fuel Hiqh
3%
0%
0%
0%
0%
0%
Cloqqed Air Filter - EPA
8%
8%
8%
8%
8%
8%
Wronq/Worn Turbo
5%
5%
5%
5%
5%
5%
Intercooler Cloqqed
5%
5%
5%
5%
5%
5%
Other Air Problem - EPA
6%
6%
6%
6%
6%
6%
Enqine Mechanical Failure
2%
2%
2%
2%
2%
2%
Excessive Oil Consumption
5%
3%
3%
3%
3%
3%
Electronics Failed - EPA
3%
3%
3%
3%
3%
3%
Electronics Tampered
10%
15%
5%
5%
5%
5%
EGR Stuck Open
0%
0%
0.2%
0.2%
0.2%
0.2%
EGR Disabled/Low Flow - EPA
0%
0%
10%
10%
10%
10%
Nox Aftertreatment Sensor
0%
0%
0%
0%
10%
10%
Replacement Nox Aftertreatment Sensor
0%
0%
0%
0%
1%
1%
Nox Aftertreatment Malfunction - EPA
0%
0%
0%
0%
13%
16%
PM Filter Leak
0%
0%
0%
5%
5%
5%
PM Filter Disabled
0%
0%
0%
2%
2%
2%
Oxidation Catalyst Malfunction/Remove - EPA
0%
0%
0%
5%
5%
5%
Mis-fuel - EPA
0.1%
0.1%
0.1%
0.1%
0.1%
0.1%
B. 7 NOx T&M Emission Effects
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 LHDD are 80 percent efficient and the selective
catalyst reduction (SCR) systems in HHDD 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.155 Visteon reports a 60-80
percent NOx reduction with open loop control.156
In testing, the failure of the NOx aftertreatment system had a different impact on the NOx emissions
depending on the type of aftertreatment. The HHDD vehicles with SCR systems would experience
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 LHDD vehicles with
LNT systems would experience 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.
236
-------
Table B-6. NOx T&M Emission Effect by Model Year Group
1994-97
1998-2002
2003-2006
2007-20C9
2010+ HHDT
2010 LHDT
Federal Emission Standard
5.0
5.0
4.0
2.0
0.2
02
Timina Advanced
80%
R0°/.
60'^
§n*
12%
Trminq Retarded
-20%
9ri-,:
...
"i
20 ¦¦¦:¦
y n : i
Injector Problem fall)
-5%
-1 .;
¦" ¦".
.. 1 j
Puff Limit er Mis-set
0%
.
(3 :¦
•v..;.
{).
U'": i
Puff Limiter Disabled
0%
G'i-'L
(J>.
'J •
r- :
o':
Max Fuel Hiqh
10%
()%
0%
i.ij
0";'.
Ciooaed Air Filter
0%
0%
0%
m
0%
0%
Wrong/Worn Turbo
0%
0%
0%
0%
0%
0%
Intercooier Clogged
25"
A/
v.-.
'*£ 70
3%
5%
Other Air Problem
0
OH'!'-
0 ¦:
riy
0%
o%.
EmirM? Failure
-" 0"/:
-" 0:'
•
--1 C ¦
.10%
-in°/> l
Oil Uori>j;
0
0 :•
ij "V"i i.
;
'¦> ¦ ¦ i
Fiec-tro:! cs ?ailed
;y
no.'
0:/
'k.ii "ji'i't
/'
Ij |
[¦¦¦ lectr:>n 1 ;VT!:::efed
R
<5f
hi:
L{) ¦:
!¦¦>/ i
bUr\ Mt.iCK Up&n
o:,.
ov.
--2C
-vij
-20 ,¦ i
fIGR Disabie:; / low 1 o
0 ••••;¦
0'"'-
3 j ¦:
:b 0 :•• •.
j
Nox A'ter'estniert Sensor
0
0':-;.
0;.
w'
V 'J !_
200 i
Re|jlK:^O Ticrt N Aftoilicaiircr Sens?"
'"'i
cy/~.
IV
0
i/ i._
200 : j
Nox A'torV:vviorf.
;
i'i ¦''/.
0 '¦¦¦
0- .
300 ¦¦¦ j
PVi F He' L.:~iik
0
C
O-'-.
•• •
0 i
PM F:;!i-:" Dis;
-------
Combining the NOx emission effects with the frequency results in the initial T&M adjustment
factor shown in the Table B-7 below. This methodology estimated a small (9-14%) T&M NOx
adjustment factor for 2009 and earlier models due to NOx effect of the following failure modes:
electronics tampered, timing advances, intercooler clogged, and ERG disabled/Low Flow. MOVES
does not use the estimated NOx T&M emission effects initially estimated for 2009 and earlier
model years, and assumes no NOx increase. This is incorporated into the 3rd column of Table B-7
labeled with (Remove 2009 and earlier).
MOVES assumes NOx increases only in the 2010+ vehicles with NOx exhaust aftertreatment
technologies, in contrast to the initial estimate of T&M due to 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.157 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, 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.
Table B-7. Tampering & Mai-Maintenance NOx adjustment factor (fr&M,NOx) (Percent) for MHD, HHD and
Urban Buses
Model
years
fT&M,NOx,nonOBD
(Initial)
fr&M,NOx ,nonOBD
(Removed 2009
and earlier)
foBD
fl&M.NOx.OBD
1994-1997
10
0
0
-
1998-2002
14
0
0
-
2003-2006
9
0
0
-
2007-2009
11
0
0
-
2010-2012
SCR
87
87
0.33
77
2010-2012
LNT
72
72
1
48
2013+ SCR
87
87
1
58
The T&M adjustment factors (fr&M,NOx) are calculated using the NOx emission effects and
frequencies shown above. The fr&M,NOx factors incorporate the OBD assumptions discussed in
Section B.10, including the assumed penetration of OBD (foBD)
Lean NOx trap (LNT) aftertreatment are assumed to penetrate 25 percent of LHD2b3 trucks starting
in 2007, consistent with the assumptions previously made in Section 2.1.1.4.6.
The T&M rates for LHD2b3 in 2007-2009 are calculated by adjusting Equation 3-10 to account for
T&M of LNT aftertreatment, as shown in Equation 7-4:
-------
2007 - 2009 LNT NOx emissions (T&M)
2003 — 2006 LHD2b3 NOx emissions
/LNT normal emissions \ Equation 7-4
= (normal op. frequency) x — x (T&M effect)
\ baseline emissions J
/baseline emissions\
+ (DPF reg. frequency) x
\ baseline emission J
= (0.90) x (0.10) x (1.72) + (0.10) x (1) x (1) = 0.2548
The ratio of 2007-2009 LHD2b3 (with T&M) over the baseline 2003-2006 NOx rates is calculated
by adjusting Equation 3-11 to account for the T&M effects of LNT, as shown in Equation 7-5.
2007 - 2009 LHD2b3 NOx emissions (T&M)
2003 — 2006 LHD2b3 NOx emissions
/2007 — 2009 LNT NOx emissions (T&M)\
= (LNT market share) ( T TTT^, ^ I Equation 7-5
V 2003 — 2006 LHD2b3 NOx emissions I
( 2007 — 2009 emission standards \
+ (non — LNT market share) ————„ ——
\2003 — 2006 NOx emissions standards/
=0.25x0.2548 +0.75x0.5=0.4387
Then, the overall T&M effect for 2007-2009 LHD2b3 is calculated in Equation 7-6 by dividing
Equation B-2 by Equation 3-11.
2007 - 2009 LHD2b3 NOx emissions (T&M)
2007 — 2009 LHD2b3 NOx emissions (zero mile)
Equation
_ / 2007 - 2009 LHD2b3 NOx (T&M) \ //2007 - 2009 LHD2b3 NOx (zero mile)\ 7-6
\2003 — 2006 LHD2b3 NOx emissions )/ V 2003 — 2006 LHD2b3 NOx emissions J
= 0.4387/0.4225 = 1.04 = 4% increase due to T&M
For 2007-2009, LHD45 uses the same emission rates and T&M factors as for the LHD2b3 rates
As noted earlier, we assume no NOx increase for the pre-2007 for the other heavy-duty regulatory
classes.
For 2010+, 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). Thus, for calculating the T&M NOx effects for 2010-2012,
we weighted the LNT-specific and 2013+SCR-specific T&M effects (from Table B-7) according to
the market shares, as shown in Equation 7-7:
2010+ LHD2b3 NOx emissions T&M Equation 7-7
239
-------
=LNT market share x (2010 thru 2012 LNT T&M) + (non-LNT market
share) x (2013+SCR T&M)
=0.25x0.48+0.75x0.58=56%
For the MHD, HHD, and Urban Buses regulatory classes we use the SCR T&M effects from Table
B-7. We assume only 33 percent OBD penetration in 2010-2012, and full penetration for 2013+
model years. Note that for 2010+ LHD45, we use the same T&M as the other heavy-duty
regulatory classes, that assume no LNT penetration, and a phased-in OBD penetration. The NOx
T&M adjustment factors used for each MOVES regulatory classes and model year groups are
shown in Table B-8.
Table B-8. NOx T&M Adjustment Factors (fr&M,NOx) by MOVES Regulatory Classes and Model Year Groups
Model Year
Group
LH2b3
(RegClass 41)
LHD45
(RegClassID
Other HD
(RegClassID
46,47,48)
Gliders
(RegClass
49)
2007-2009
4%
4%
0%
0%
2010-2012
56%
77%
77%
0%
2013+
58%
58%
58%
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 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 effects along with the OBD effects discussed in 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 2-19).
240
-------
Table B-9. PM2.5 T&M Emission Effect by Model Year Group
1!)!) 1
21111:;
20(17
l!)!)7
211(12
21 lll(')
2(1(1!)
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/Worn 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%
|i nr..
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%
nrr..
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 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.
241
-------
Table B-10. THC T&M Emission Effect by Model Year Group
1994-97
1998-2002
2003-2006
2007-2009
2010+ HHDT
2010 LHDT
Federal Emission Standard
1.3
1.3
1.3
0.2
0.14
0.14
Timing Advanced
0%
0%
0%
0%
0%
0%
Timing Retarded
50%
50%
50%
50%
10%
10%
Injector Problem (all)
1000%
1000%
1000%
1000%
200%
200%
Puff Limiter Mis-set
0%
()"
0°/
0%
0%
()¦¦'•.
Puff Limiter Disabled
0%
()"
0°/
0%
0%
()¦¦'•.
Max Fuel High
10%
()¦¦¦
0°/
0%
0%
()¦¦'•.
Clogged Air Filter
0%
0'/i
0%
0%
0%
0'/i
Wrong/Worn Turbo
0%
0%
0%
0%
0%
0%
Intercooler Clogged
0%
0%
0%
0%
0%
0%
Other Air Problem
0%
0%
0%
0%
0%
0%
Engine Mechanical Failure
500%
500%
500%
500%
100%
100%
Excessive Oil Consumption
300%
300%
300%
300%
60%
60%
Electronics Failed
f>0A-.'.
50%
50%
50%
10%
10%
Electronics Tampered
()¦>
0°/
0%
0%
0%
EGR Stuck Open
()¦>
0°/
1 OO":'.
1 ()()¦¦:'.
20%
20%
EGR Disabled / Low Flow
()¦>
0°/
0%
0%
0%
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%
0%
0%
PM Filter Disabled
()¦>
0°/
0°/
0%
0%
0%
Oxidation Catalyst Malfunction/Remove
50%
50%
50%
Mis-fuel
-------
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 2-22 (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.
Specifically, we reduced the emissions increases for all pollutants due to tampering and mal-
maintenance 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 engines would have
OBD in the 2010, 2011, and 2012 model years, and 100 percent would have OBD by 2013 model
year and later. Equation 7-8 describes the calculation of the percent increase in emission rate
through useful life (T&M adjustment factors (fx&M))> where fonu represents the fraction of the fleet
equipped with OBD (0 percent for model years 2009 and earlier, 33 percent for model years 2010-
2012, and 100 percent for model years 2013 and later). The result from this equation can be
plugged into Equation 7-2 to determine the emission rate for any age group.
^T&M.p = fT&M,nonOBD,p X (1 — f0Bo) c .. , „
. Equation 7-8
+ 0.67 X rx&Mnon0BD,p x Mobd)
Where:
fx&M,p= the tampering and mal-maintenance adjustment factor for pollutant, p, that accounts
for the phase-in of OBD
fx&M,p= the tampering and mal-maintenance adjustment factor for pollutant, p, for engines
without OBD; calculated in Sections B.7 through B.9
f0BD= penetration of the fleet equipped with OBD
These OBD impacts apply to any truck in GVWR Class 4 and above (regulatory classes LHD45,
MHD, HHD, and Urban Bus). LHD2b3 trucks are assumed to have the OBD fully-phased in
beginning in model year 2010 as shown in Equation 7-7. As data for current and future model years
become available, we may consider refining these estimates and methodology.
-------
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
7-1. 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 7-1. References Used to Support In-Use DPF Failure Rate Assumption for Extended Idling Emissions
Study
Relevant Information
US EPA (2015)108
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)158
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)159
3% of 2007+ trucks at Port of LA have PM emissions 3x the standard. 9% of
2008+ trucks at Cottonwood site have PM emissions 3x the standard
CARB (2015)160
35% to 4% of trucks submitted warranty claims related to the PM filter between
2007 and 2011
CARB (2015)160
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. (20 1 4)159 and Preble et al. (20 1 5)158 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 7-2. 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.
244
-------
Table 7-2. Baseline and deteriorated THC and PM2.5 emission rates to account for failure of diesel particulate
filters (DPFs) by model year groups
Engine
Model
Year
Baseline
Deteriorated
THC
fe/hr)
PM2.5
fe/hr)
EC
fe/hr)
nonEC
fe/hr)
Failure
rate
THC
fe/hr)
PM2.5
fe/hr)
EC
fe/hr)
nonEC
fe/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.161 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)/1.53 = 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 7-3. As shown, the MOVES EC/PM emission rates for MY 2007+
trucks are slightly higher than the 'Baseline' EC/PM fractions in Table 7-2, 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 7-3. Emission Rates Calculated from Weighting the 'Baseline' and 'Deteriorated' Emission Rates from
Engine
Model
Year
MOVES
Deteriorated
Fraction
THC
fe/hr)
PM2.5
fe/hr)
EC
fe/hr)
nonEC
fe/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.
245
-------
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.
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
246
-------
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
247
-------
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
248
-------
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
249
-------
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
group
Speed Class (mph)
Type
Medium
Heavy-Duty
Heavy Heavy-
Duty
1960-87
1-25
Intercept (fio)
-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 (fii)
0.02821
0.0968
Transformation
Coefficient
0.5864
0.84035
(0.5c2)
ln(PM) = pQ+ /^STP + 0.5cr2
Where :
/?o = an intercept term for a speed class within a model year group, as shown in the table above,
= 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).
250
-------
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 (hO< = MK)
CO slor ts *s. Age by MYG
(a) CO
o [ ; : I
*
"" 111!
modelycargroup
II199? 19982004
FTP Coid-Starts (g). HD SI (HD< = 14K)
THC star-ts vs. Ago by HYG
(b) THC
! j
I9E0I989 n D C 19901990
FTP Cold-Starrs (g). HD SI (HD< =
NOx starts vs. Age by P
11 12 13
1911997 AAA 19902004
Iyeargroup OOP 19601989 CCD 19901990 I ' I 1991199? AAA 19982004
Figure 7-1 Cold-Start FTP Emissions for Heavy-Duty Gasoline Vehicles, Averaged by Model-year and Age
Groups
251
-------
FTP Cold-Starts (g), HD SI (HD< = 14K)
CO GEO-mean starts vs. Age by MYG
node I year group 13601389 B-B-B 19901330 "•—I—~ 1991 1937 AAA 19982004
FTP Cold-Starts (g). HD SI (HD< = 14K)
THC GEO-mean starts vs. Age by MYG
node I year group e-e-e I960 1389 B-B-B 19901930 H—II—~ 19911997 A A A 19982004
FTP Cold-Starts (g). HD SI (HD< = 14K)
NOx GEO-mean starts vs. Age by MYG
Figure 7-2 Cold-Start FTP Emissions for Ileavy-Duty Gasoline Vehicles, GEOMETRIC MEANS by Model-year
and Age Groups
252
-------
CO I n_SD vs. Age by HYG
node 1 ycargroup O-O O 13G0I383 B-B-B 13301390 •*—I—«" 19911337 ft-A-6 13382004
FTP Cold—Starts (g), HD SI (hD< = 14K)
THC In.SD vs. Age by MYG
node 1 yeargroup O O O 13601383 O 13 ~ 13301330 H—I—I" 19911397 a & & 13382004
FTP Cold-Starts (g), HD S! (HD< = 14 K)
NOx In.SD vs. Age by WYG
nodes I yeargroup e-Q-e 13601389 B-B-B 19901330 "»—t-H- 1931 1997 tr-ts-A 13382004
Figure 7-3 Cold-start FTP Fmissions for Heavy-Duty Gasoline Trucks: LOGARITHMIC STANDARD
DEVIATION by Model-year and Age Groups
253
-------
CD ARITH-meon sloits vs. Age by M*C
Figure F-l. Cold-Start Emissions for Heavy-Duty Gasoline Trucks: RECALCULATED ARITHMETIC
MEANS by Model-year and Age Groups
Table F-l Emission Standards for Heavy-Duty Spark-Ignition Onroad Engines
FTP Coid—Starts (g). HD SI (l-0< = 14K)
THC ARITH-mean starts vs Age by MYQ
FTP Coid—Starts (g). HD SI (hO< = MK)
NO* ARITH-meon starts vs. Age by MYG
254
-------
Regulatory Class
Model Year
Emissions Standards (g/hp-hr)
CO
THC
NMHC
X
O
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
255
-------
Appendix G Selection of Fixed Mass Factor (fSCaie) 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 fscaie 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 fscaie 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 of fscaie 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 criteria pollutants and CO2 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 of fscaie 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,
256
-------
we expect LHD, MHD, and HHD vehicles 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 fscaie between candidate values that look reasonably good for both vehicle
count and time distribution.
Looking at Table G-l for LHD vehicles, fscaie = 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 fscaie 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 fscaie value, for LHD, should be between
2.06 and 9.00. Based on further analysis, the final fscaie 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 fscaie 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 the fscaie 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, fscaie 5.00 yields much more aggressive driving behavior in
operating modes 30 and 40 compared to fscaie 4.00, with mean CO2 emissions rates approximately
30% higher in these operating modes. As stated earlier, our objective is to select the fscaie that yields
the most aggressive operation in the highest operating modes while still providing sufficiently
robust estimates of the emission rates. Because fscaie 5.00 still provides mean CO2 emissions rates
that are robust for the high operating mode bins, fscaie 5.00 is preferable to an fscaie of 4.00. fscaie
5.00 provides strong and expected increasing trends in CO2 emission rates between operating
modes 29 and 30, whereas fscaie 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 fscaie
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 of fscaie 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,15 one of peer-reviewers asked if we have evaluated the time distribution
from real-world data using the proposed fscaie value. In Figure 7-4 through Figure 7-6 we conducted
257
-------
a comparison of the operating mode distributions measured from the HDIUT dataset and real-world
operating modes estimated from MOVES3 national scale runs for the three evaluated regulatory
classes LHD, MHD, and HHD at the proposed proposed fscaie 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 fscaie 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 fscaie values from other in-use datasets. We agree that this would be useful to better
understand the representativenss 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.
258
-------
Table G-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
1 1
1
25844
28819
25061
18770
:>r>71
1
28
64
64
64
44
¦>¦>
V
21514
15281
8422
5020
."">11
j
29
64
64
42
23
I)
l)
16292
5808
2900
1161
I)
I)
30
64
43
22
13
V
1
35269
3903
1002
40
.)
1
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
1 1
.)
33332
86600
79892
67271
13177
7
38
64
63
62
44
¦>¦>
1
46641
52178
31818
12100
978
V
39
64
62
41
23
1
1
49374
21416
6167
1940
1
7
40
64
41
23
14
1
1
113054
7852
1685
135
IS
13
Notes:
1 Values in bold are for final selected fscaie• 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.
259
-------
H fs = 4.00
¦ fs = 5.00
B fs = 6.00
HIH
win Ml ll I
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 fscaie 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.
260
-------
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
4.0
fs
5.0
fs
6.0
fs
7.0
fs
9.0
fs
17.1
fs
4.0
fs
5.0
fs
6.0
fs
7.0
fs
9.0
fs
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
2.")
I)
18677
15863
13842
10424
1(111)
I)
29
58
58
46
25
in
I)
12105
10075
5243
3550
62:)
I)
30
58
39
25
14
I)
I)
17304
7229
3379
1034
1)
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
:?n
1
56697
58308
51520
44843
1 1 1 11
1
39
58
53
43
25
in
1
45255
39067
25466
8307
1628
1
40
53
39
26
14
i
i)
64411
25344
7923
2440
7
I)
Notes:
1 Values in bold are for final selected fscaie• 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.
261
-------
mi
Hfs= 5.00
Hfs= 6.00
¦ fs= 7.00
ML
ill
hhi
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 fscaie Value on Coverage and Trends of operating mode Based CO2 for Medium Heavy-Duty
Vehicles
Table G-3 Effect of fscaie Value on Vehicle Count and Time for Heavy Heavy-Duty Vehicles
262
-------
Number of vehicles1
Number of seconds based on a cycle with
million seconds1,2
one
OpMo
de
fs
9.0
fs
10.
0
fs
11.
0
fs
12.
0
fs
14.
0
fs
17.
1
fs
9.0
fs
10.0
fs
11.0
fs
12.0
fs
14.0
fs
17.1
0
159
159
159
159
158
159
18010
18010
18010
18010
18010
18010
1
159
159
159
159
159
159
29766
2
29766
2
29766
2
29766
2
29766
2
29766
2
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
8
0
8905
9822
6562
232!)
18f>
0
30
114
59
10
1
0
0
5861
1127
239
1
0
0
33
159
159
159
159
158
159
11421
4
12621
6
13973
1
15410
1
18609
4
23796
0
35
159
159
159
159
158
159
13910
9
16066
7
17614
4
18613
1
18981
3
17611
1
37
159
159
159
159
153
153
11505
0
10244
0
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
6576
391
0
40
114
65
11
1
0
0
16174
3344
550
0
0
0
Notes:
1 Values in bold are for final selected fscaie. 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.
263
-------
H fs = 9.00
¦ fs = 10.0
Bfs = 11.0
imh 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 fscaie Value on Coverage and Trends of operating mode Based CO2 for Heavy Heavy-Duty
Vehicles
264
-------
1
035
0,30
0,25
0.20
0.15
0.10
0.05
0.00
¦1
HHDD Opmode Distributions
Comparison o rs. MOVES-CT1
¦ HDIUT IOVES CT1 - CY2027/MY2014
11 12 13 14 15 16 ! 21 22 23 24 25 27 28 29
35 37 38 39 4t
2
3
4
5
6
Figure 7-4. 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
035
c
o
0,30
u
m
LL.
0,25
T3
0,20
O
E
a.
0.15
O
0.10
0.05
0.00
node Distributions
Comparison of HDiUT vs. MOVES-CTI
¦ HDIUT 10VES CTI - CY2027/MY2014
11 li
I
0 1 j 11 12 13 14 15 21 22 23 24 25 27 28 29 30
Opmode
33 35 37 38 39 40
Figure 7-5. 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
266
-------
0,45
0.40
0,35
0,30
0,25
0.20
0.15
0,10
0.05
0.00
LHDD Opmode Distributions
Comparison of HDIUT vs. MOVE5-CTI
¦ H MOVES €11 - CY2027/MY2014
I I
I.
I I
I
IIII.,
Opmode
Figure 7-6. Light Heavy-Duty Diesel Operating Mode Distribution compared between the Heavy-Duty In-Use Testing (HDIUT) Program and from
MOVES3ss for a MY 2014 vehicle with an fscaie of 5 metric tons.
ss 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.
267
-------
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 MY2010-2013 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.1.1LHD
0,007
0.006
0,005
< 0,004
0,003
H
0,002
0,001
0.000
MOVES OpMode
Figure 7-7 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
ill LHD PEL 0,20, MY 2010-2013
ill LHD FEL 0.38, N=0
¦ LHD FEL 0,50, MY 2010-2015
!\: ¦¦¦¦¦
T
I 1 - 1 I ! 1!1
1 - I
-------
MOVES OpMode
Figure 7-8 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
—. °-004
•5*
m
003
X
l_
0.002
0.001
0.000
0 1 11 12 li 14 lb lb 21 22 2i 24 2b li 28 29 iO ii is 4/ iX is 40
MOVES OpMode
Figure 7-9 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
: i
20, MY 2010-20P, N=2
35, MY 2010-2013, U=2
l MHO f
l MHD F
i MHD PEL 0,50, fc
' 2010-20
N=9
li
ill
I I
It
269
-------
0,05
0.04 -
"sr o,o3
m
o
u 0,02 -\
0,01 -
0,00
ill ill^ili 11
( MHD PEL 0,20, MY 2010-2013, N*23
I MHD FEL0.35, MY 2010-2013, N=23
I MHD FEL0.5Q, MY 2010-2013, N±9
1 ill ill ill ill II
J
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 7-10 Average MHD CO 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
270
-------
H.2 Comparison of THC and CO Emission Rates between MY 2010-
2013 and MY2014 in the 0.2 N0X FEL Group for LHD andMHD
Figure 7-11 through Figure 2-37 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.
H.2.1 LHD
3
u
x
: MY 2010-2013
MY 2014-2016
0.0040
0.0035
0.0030
0.0025
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 7-11 THC 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
P I1 I1
i1 «] i1 i1 1' i
ii il il I 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 7-12 CO emission rates for the MY 2010-2013 and MY 2014-2016 vehicles in the LHD 0.20 NOx FEL
Group
271
-------
H.2.2MHD
0.0014
0.0012
J" 0,0010
3
Sjj 0,0008
l-
0.0008
0.0004
0.0002
0,0000
1
1
1
1
MY 2010-2013
MY 2014-2015
1 A1 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 7-13 THC emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the MHD 0.20 NOx FEL
Group
0.014
0.012
0.010
5 0.008
m
0 0.006
0.004
0.002
0.000
: MY 2010-2013
MY 2014-2015
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 7-14 CO emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the MHD 0.20 NOx FEL
Group
272
-------
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 7-4. Figure 7-15 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.
273
-------
Table 7-4 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
X
O
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 grams 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
o.o
100 200 300 400 500
SoakTime (mins)
600
700
800
Figure 7-15. Grams of NOx from Start Emissions versus Soak Time
274
-------
1.2 Comparison of Heavy-duty Gasoline Emission Rates by Vehicle
Figure 7-16 through Figure 7-18 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 fscaleof 5 metric tons used for the LHD2b3 and LI 11)45 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 ¦
*22
1 0.0, H
0.01 -
0.00
I 2016_Ford_E459, R=76
I 2015_Isuzu_NPR, R=83
12017_RAM_3500, R=38
i
.i- B-
I
1
1
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 7-16. 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.
275
-------
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 1
iu_
ii- t, ll.
»- li- li,
¦ iil
11 12 13 14 15 16 ; 21 22 23 24 25 27 28 29 30
33 35 37 38 39 40
MOVES OpMode
Figure 7-17. Mean Heavy-duty Gasoline THC Emission Rates by Operating Mode and Vehicle Calculated using
fscale oi 5 metric tons.
9.00
0.60
8.00 :
0.50
7.00 :
0.40
^5
6.00 :
0.30
5.00 :
0.20
O
u
4.00 :
0.10
3.00 :
0.00
2.00 :
1.00 :
0.00
¦ 2016_Ford_E459, R=76
¦ 2015_buzu_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 7-18. Mean Heavy-duty Gasoline CO Emission Rates by Operating Mode and Vehicle Calculated using
fscale of 5 metric tons.
276
-------
I. 3Extrapolating High-Power Operating Modes
When analyzing the heavy-duty gasoline data for the MHD and HHD regulatory classes (using an
fscaieof 7 and 10 metric tons respectively), there was limited or no data for high power operating
mode bins. Figure 7-19 shows the mean NOx emission rates by vehicle and operating mode when
using an using an fscaie 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.
o.o5 q ¦ i r.
0,04 : - ¦,¦¦¦¦, ; I ¦ II
"m 0.03 : I II III r I
S 1 111 ...ill! II
X ; i i III' III!! II
§ 0-02 - j j | | | | | |L I I
o.oi - • 1 Ti II 1 ; I I J J] I I 11 111 II I
o.oo J -—^—*
MOVES OpMode
Figure 7-19. Mean Heavy-duty Gasoline NOx Emission Rates by Operating Mode and Vehicle Calculated for
HHD using an fof 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. Figure 7-20 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 THC and CO emission rates.
277
-------
0,05 -
0.04 -
JT 0.03 :
S
x
§ 0.02 -
0.01 :
0.00 -
0 1 ; 11 12 13 14 15 16 ; 21 22 23 24 25 2? 28 29 30 ; 33 35 37 38 39 40
MOVES OpMode
Figure 7-20. Mean Heavy-duty Gasoline NOx Emission Rates by Operating Mode and Vehicle Calculated for
HHD usinga fscaieoi 10 metric tons with Aggregated Means for High Power Bins with Limited Data
Figure 7-21 shows the weighted average NOx emission rate calculated by averaging the three
vehicles together according to their production volume sales.
0.040
0.035
0.030
— 0.025
m
0.020
X
0.015
0,010
0.005
0,000
MOVES OpMode
Figure 7-21. 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 THC, CO, and NOx. 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 7-5.
12016_Ford_E459, R=80
; 2015Jsuzu_NPR, R=84
I 2017 RAM 3500, R=38
278
-------
Table 7-5. 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
Figure 7-22 displays the initial mean CO2 emission rates using an fscaieoi 10 metric tons. Note that
the emission rates for CO2 are more well behaved than the criteria pollutants, because of this, less
of the high power operating modes were replaced with extrapolated rates, than were binned 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).
279
-------
MOVES OpMode
Figure 7-22. Mean Heavy-duty Gasoline CO2 Emission Rates by Operating Mode and Vehicle Calculated for
HHD using an fof 10 metric tons
Figure 7-23 shows the mean CO2 emission rates using an fscaieof 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.
m
m
r*
O
u
90
80
70
60
50
40
30
20 :
10 :
0
f 2016_Ford_E459, R=80
12015Jsuzu_NPR, R=84
12017 RAM 3500, R= 38
I
I 1
I, 1
I, 1
I
t I
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
MOVES OpMode
Figure 7-23. Mean Heavy-duty Gasoline CO2 Emission Rates by Operating Mode and Vehicle Calculated for
HHD using an fscaieoi 10 metric tons with Extrapolated Means for High Power Bins.
Figure 7-24 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.
280
-------
80
60
BO
S 40
fj
O
u
20
U ± i XI ±/L ±Z> JLS- ±J> ID i Z.J. £.£. Z. 3 <£.«+ Z. Z. / Z.O Z. ^ OU i DO DJ O / 30 33 «+U
MOVES OpMode
Figure 7-24. Weighted Average Heavy-duty Gasoline CO2 Emission Rates by Operating Mode for HHD Using
Production Volumes
Similar calculations were repeated for the MHD vehicles calculated using an fscaie of 7 metric tons.
281
-------
Appendix J Comparing Glider Vehicle and MOVES Model Year
2000 Heavy Heavy-Duty Emission Rates
Glider vehicles are new vehicles with older engines. Tailpipe exhaust emissions were measured
from two heavy heavy-duty glider vehicles, one each of MY 2016 and MY 2017, at the US EPA's
National Vehicle Fuel and Emissions Laboratory (NVFEL) heavy-duty chassis dynamometer. Both
glider vehicles were sleeper cab tractors and lacked EGR, DOC, DPF, or SCR emissions control
devices. Each glider vehicle was tested on three drive cycles - (1) cold start Heavy-Duty Vehicle
Urban Dynamometer Driving Schedule (UDDS) sequence; (2) World Harmonized Vehicle Cycle
(WHVC) sequence; (3) Super Cycle (a combination of urban and freeway driving). The vehicles
were tested with two sets of road-load coefficients representing 60,000 lbs and 80,000 lbs total
vehicle weight. The MOVES operating mode based results shown here are time-weighted average
of the various drive cycle and weight combinations. More details about the vehicles, drive cycles,
test procedures, and results are available in the EPA report for the glider test program.162 The data
that was analyzed for MOVES included only the tests where the vehicles were operating free of
any malfunction indicator lights because MOVES treats tampering and malmaintenance separately.
Since the measured glider emission rates are from only two glider vehicles tested in lab conditions,
we decided not to use those rates as-is to estimate fleet-wide real-world glider vehicle emissions
from MOVES. Instead, we used the data to identify which model year emission rates in MOVES
can best represent contemporary glider vehicles. We know that contemporary glider vehicles
typically have re-built powertrains that do not require DOC, DPF, and SCR controls, and likely
lack even EGR systems. This meant glider vehicles should compare well with MY 2006 or older
emission rates. Thus, we compared the emission rates from the two glider vehicles with regClass
47 (HHD) rates for applicable model year groups in MOVES. MOVES defines emission model
year group by pollutant and has emissions rates for one or more model year groups within the MY
1998-2006 period. For example, in MOVES, there are different NOx rates for MY 1998, MY 1999-
2002, and MY 2003-2006. On the other hand, PM2.5 and CO2 (from energy) rates are the same
across MY 1998-2006.
A comparison of glider test results versus MOVES base emissions rates for NOx, THC, CO, CO2,
and PM2.5 are presented in Figure J-l though Figure J-6. Given that glider vehicles lack emissions
control devices that provide significant reductions in NOx and PM2.5 , our primary focus was to find
the best model year match for these pollutants, with THC and CO as secondary comparison to pick
between model years that match equally well for NOx and PM2.5 . The decision to pick a
representative model year mostly came down to NOx because emissions rates for PM2.5 are the
same for MY 1998-2006. From Figure J-l, it is seen that the NOx emissions from the measured
glider vehicles are best represented by the MY 1999-2002 rates in MOVES. Similarly, for THC
(Figure J-2), MOVES MY 1998-2002 rates are a better match compared to the MY 2003-2006
rates. The CO rates from the measured glider vehicles are significantly higher than MOVES
(Figure J-3), so it is likely that modeled CO emissions for glider vehicles in MOVES will be
underestimated. The CO2 rates from glider vehicles are comparable to the MOVES MY 1998-2006
rates (Figure J-4). When comparing the rates between measured glider vehicles and MOVES, we
paid attention to differences in the comparison across operating modes. Typically, within each
speed-bin, vehicles spend more time in and thus generate a larger fraction of total emissions from
the low and medium power operating modes versus high-power operating modes. Using NOx and
282
-------
C02 as examples, the comparison between measured glider vehicle versus MOVES MY 2000
emission rate for operating modes 27 and 28 are not as good as lower operating modes. However,
the difference in these high power operating modes has a lower impact on total inventory because
vehicles spend less time in the operating mode. For PM2.5, we have gravimetric filter-based
measurement for both gliders and instantaneous measurement using an AVL Micro-Soot Sensor
(MSS) for only Glider-2, as seen in Figure J-5 and Figure J-6. We applied MOVES operating mode
based PM2.5 rates to the operating mode based time distribution for Glider-1 and Glider-2 test
cycles to estimate cycle total emissions for MOVES and compared those to the total PM2.5 from
the gravimetric filter measurements. While Glider-2 filter and instantaneous rates are lower than
MOVES MY 1998-2006 rates, the comparison for Glider-1 is much closer. Overall, we believe
MOVES MY 1998-2006 PM2.5 rates are reasonable for use as glider vehicle PM2.5 rates because
they are derived from vehicles without DPF, a key PM2.5 emissions control device that is absent in
most glider vehicles.
Based on these comparisons, we decided to apply the MY 2000 HHD (regClass 47) emission rates
for THC, CO, NOx, PM2.5 , and Energy (for CO2) in MOVES to glider vehicles (regClass 49).
MOVES OpMode
Figure J-l Comparison of NOx Emission Rates from Glider Vehicle Lab Testing to MOVES regClass 47 (HHD)
1998-2006 Model Years
283
-------
60
u
60
50
40
30
20
10
—A- Glider-1
-B- Glider-2
—e-MOVES_MY1998-2002
-O- MOVES_MY2003-2006
J ©
a /p---e..
>0--®—er
A A^j£s—
'J&—& ° B"
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 J-2 Comparison of THC Emission Rates from Glider Vehicle Lab Testing to MOVES regClass 47 (HHD)
1998-2006 Model Years
900
800
700
'Lr
600
-C
W>
500
o
u
400
300
200
100
0
Glider-1
Glider-2
-MOVES_MY1998-2002
MOVES MY2003-2006
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 J-3 Comparison of CO Emission Rates from Glider Vehicle Lab Testing to MOVES regClass 47 (HHD)
1998-2006 Model Years
284
-------
400000
350000
300000
£ 250000
M
g 200000
u
-A- Glider-1
-B Glider-2
-0-MOVES MY1998-2006
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 J-4 Comparison of CO2 Emission Rates from Glider Vehicle Lab Testing to MOVES regClass 47 (HHD)
1998-2006 Model Years
Glider 1 O MOVES_1998-2006
Glider 2 O MOVES JL998-2006
30
_ 25
re
§ 20
_gj
S" 15
^ 10
2
<>- 5
0
O
0
~
O
0
A O
~ 0
v. % v
v. x v
60,000 lbs
80,000 lbs
30.0
25.0
re
20.0
O
a;
15.0
>•
10.0
5
5.0
Q.
0.0
0
O
¦
¦
0
0
0
¦
u
¦
¦ ¦
V
\f X V
60,000 lbs
80,000 lbs
Figure J-5 Comparison of Gravimetric Filter-based PM2.5 Emissions from Glider Vehicle Lab Testing to
MOVES regClass 47 (HHD) 1998-2006 Model Years
285
-------
180
160
140
-£• 120
-C
-S2 ioo
80
60
40
20
0
Figure J-6 Comparison of Instantaneous PM2.5 Emissions from Glider-2 Lab Testing to MOVES regClass 47
(HHD) 1998-2006 Model Years
-B-Glider-2
-©-MOVES_MY1998-2006 A
/ 1 1
/ Q 1
1 jit P
Q—(SH
~
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
286
-------
Appendix K PM Composition Measurements from Auxiliary
Power Units
Table 7-6 reports the organic carbon (OC), elemental carbon (EC) and total carbon (TC)
measurements conducted in the study conducted by Texas Transportation Institute (TTI, 201494).
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 7-6. Organic Carbon, Elemental Carbon, and Total Carbon Measurements from the IMPROVE_TOR
measured on APU 1
Sample ID
Min.
DR
Test
OC
OC
uncertainty
EC
EC
uncertainty
TC
TC
uncertainty
EC/TC ratio
(lig/cm2)
(lig/cm2)
(lig/cm2)
(|ig/cm2)
(lig/cm2)
(|ig/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
287
-------
Appendix L Crankcase Emissions from Glider Trucks
The MOVES CrankcaseEmissionRatio table stores crankcaseEmissionRatios (ratio of crankcase to
tailpipe exhaust) by fuel type, source type and model year. MOVES applies the crankcase emission
ratios to tailpipe exhaust emissions (start exhaust, running exhaust, and extended idle exhaust) by
fuel type, source type, and model year. In the current MOVES structure, MOVES is unable to
apply separate crankcase emission ratios by regulatory class, including to distinguish crankcase
emission ratios between glider trucks (regulatory class ID 49) and other heavy-duty regulatory
classes (HHD, MHD, LHD45, LH2b3).
Because glider trucks as defined by MOVES to use pre-2007 engines and emission control
technologies, the crankcase emissions should also use the pre-2007 crankcase emission ratios. Due
to the higher tailpipe exhaust emission rates for pre-2007 vehicles, the crankcase emission ratios
are lower for THC, CO, and NOx, but comparable for total PM2.5 emissions. (See Table 5-3, Table
5-8 and Table 5-16). The crankcase emission ratios are quite different for the PM2.5 species,
including elemental carbon, because the pre-2007 diesel trucks have much higher EC/PM fractions
in their exhaust, whereas the EC/PM fraction for crankcase emissions is low.
In developing the crankcase emission ratios for MY 2007 and later heavy-duty trucks, we did not
consider the impact of glider trucks. Developing crankcase emission ratios in the current structure
would have made the crankcase ratios much more complex and quality assure. For example, rather
than having constant crankcase emissions for model year groups and across source types, the
emission ratios would need to vary for each model year and source type according to the population
of glider trucks. Also, for some use cases of MOVES, it may not be desired to have the nationally
representative population of glider trucks, incorporated into the crankcase emission ratios.
This appendix shows the impact of this choice by comparing glider crankcase emissions calculated
by MOVES3 and a "Sensitivity Case," glider crankcase emissions calculated by post-processing
the MOVES3 output to apply pre-2007 HD diesel crankcase ratios to the glider emissions. Table K-
1 displays the percent differences between MOVES3 and the "Sensitivity Case." As expected
MOVES3 overestimates the emissions from gliders for THC, CO,and NOx for most cases. The total
PM2.5 values are quite similar, however the elemental carbon (EC) compoment of PM2.5 is
overestimated substantially.
Table K-l also displays the percent difference for crankcase emissions aggregated across all heavy-
duty source types. As expected, the overestimation of crankcase emissions is much less when
considering the other regulatory classes, limited to 30% overestimation for THC and NOx, and
differences of less than one percent for CO and PM2.5. When running, start and all other exhaust
processes are considered for heavy-duty diesel trucks the impact is less than 3% for THC
emissions, and less than 0.2% for CO, NOx, and PM2.5.
288
-------
Table L-l. Percent Differences Between M0VES3 and Sensitivity Case Applying pre-2007 Crankcase Ratios to
Year
Process and
Regulatory Class
% Differences between MOVES3 and Sensitivity
Case
THC
CO
NOx
PM2.5
EC
2015
Crankcase: gliders
548%
160%
1478%
-4%
4979%
2020
550%
15%
1506%
0%
5316%
2030
554%
-19%
1521%
0%
5446%
2045
557%
-26%
1525%
-2%
5456%
2015
Crankcase: all HD
diesel regulatory
classes
6%
1%
18%
0%
139%
2020
18%
0%
22%
0%
500%
2030
29%
-1%
19%
0%
833%
2045
24%
-1%
15%
-1%
667%
2015
All exhaust
processes: all HD
diesel regulatory
classes
0.4%
0.0%
0.1%
0%
1%
2020
1.5%
0.0%
0.1%
0%
3%
2030
2.7%
0.0%
0.2%
0%
12%
2045
2.7%
0.0%
0.2%
0%
18%
Table K-l shows that Elemental Carbon (EC) is the pollutant most impacted by the way we
account for the pre-2007 crankcase ratios for gliders. In current and historic years, the impact is
less than 3%, but for future years, MOVES is overestimating EC emissions by up to 18% in CY
2045.
Table K-2 displays the EC/PM fraction between the MOVES3 and Sensitivity Case for different
groups of processes and regulatory classes. As shown, MOVES3 estimates much higher EC/PM
fractions for crankcase emissions from gliders (66% to 71%) than is expected in the sensitivity case
(1%). When aggregated to all HD regulatory classes and exhaust emission processes, this causes
the EC/PM fraction to be slightly higher in MOVES3 than in the sensitivity case (1% higher in
2015 to 6% higher in 2045).
289
-------
Table L-2. EC/PM Frations from MOVES3 and Sensitivity Case Applying pre-2007 Crankcase Ratios to Glider
Emissions by Process and Regulatory Classes
Year
Process and
Regulatory
Class
EC/PM
MOVES 3
Sensitivity Case
2015
HD diesel
crankcase: non-
gliders
2%
2%
2020
2%
2%
2030
5%
5%
2045
9%
9%
2015
HD diesel
crankcase:
gliders
66%
1%
2020
67%
1%
2030
69%
1%
2045
71%
1%
2015
HD diesel
crankcase: all
regulatory
classes
4%
2%
2020
11%
2%
2030
33%
4%
2045
40%
5%
2015
HD diesel all
exhaust
processes: all
regulatory
classes
61%
60%
2020
58%
57%
2030
48%
43%
2045
37%
31%
290
-------
8 References
1 USEPA (2020). Speciation of Total Organic Gas and Particulate Matter Emissions from Onroad Vehicles in
MOVES3. EPA-420-R-20-021. Office of Transportation and Air Quality. US Environmental Protection Agency. Ann
Arbor, MI. November 2020. https://www.epa.gov/moves/moves-technical-reports.
2 USEPA (2020). Air Toxic Emissions from Onroad 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 (2015). Greenhouse Gas and Energy Consumption Rates for Onroad Vehicles: Updates for MOVES2014.
EPA-420-R-15-003. Assessment and Standards Division. Office of Transportation and Air Quality. US Environmental
Protection Agency. Ann Arbor, MI. October, 2015. https://www.epa.gov/moves/moves-technical-reports.
4 USEPA (2012). Use of Data from "Development of Emission Rates for the MOVES Model, " Sierra Research, March
3, 2010. Assessment and Standards Division. Office of Transportation and Air Quality. Ann Arbor, MI. April 2012.5
USEPA (2020). Evaporative Emissions from Onroad Vehicles in MOVES3. EPA-420-R-20-012. Office of
Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI. November 2020.
https://www.epa.gov/moves/moves-technical-reports.
6 US EPA. Heavy-Duty Highway Compression-Ignition Engines and Urban Buses—Exhaust Emission Standards.
http://www.epa.gov/otaq/standards/heavv-dutv/hdci-exhaust.htm. Accessed February 12, 2015.
7 40 CFR§ 86.091(2).
8 USEPA (2020). Exhaust Emission Rates for Light-Duty Onroad Vehicles in MOVES3. EPA-420-R-20-019. Office of
Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI. November 2020.
https://www.epa.gov/moves/moves-technical-reports.
9 USEPA (2020). Population and Activity of Onroad Vehicles in MOVES3. EPA-420-R-20-023. Office of
Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI. November 2020.
https://www.epa.gov/moves/moves-technical-reports.
10 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.
11 USEPA (2015). Exhaust Emission Rates for Heavy-Duty Onroad 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.
12 USEPA (2016), Memorandum to the 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" Docket
No. EPA-HQ-OAR-2016. July, 2016.13 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/2020-08/documents/epa_peer_review_handbook_4th_edition.pdf.
14 USEPA (2017). Exhaust Emission Rates for Heavy-Duty Onroad Vehicles in MOVES20IX - Draft Report. Draft
report and peer-review documents. Record ID 328830. EPA Science Inventory. September 2017.
https://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=328830.
15 USEPA (2019). Exhaust Emission Rates of Heavy-Duty Onroad Vehicles in MOVES_CTI_NPRM - Draft Report.
Draft report and peer-review documents. Record ID 347135. EPA Science Inventory. December 2019.
https://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=347135.
291
-------
16 USEPA (2020). Exhaust Emission Rates of Heavy-Duty Onroad Vehicles in MOVES3 - Draft Report. Draft report
and peer-review documents. Record ID 347138. EPA Science Inventory. July 2020.
https://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=347138.
17. 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/ .
18 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.
(http://www.epa.gov/ttn/chief/conference/eil5/sessionl/jack.pdf).
19 McClement, Dennis. Reformatting On-Road In-Use Heavy-Duty Emissions Test Data. Sierra Research, Sacramento,
CA. April 2008.
20 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.
21 Gautam, Mridul, Nigel N. Clark, Gregory Thompson, Daniel K. Carder, and Donald W. Lyons. Development of In-
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.
22 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.
23 "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.
24 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-economy-data/manufacturer-run-use-testing-
program-data-heavy-duty-diesel-3.
25 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
26 Data Collection of Drayage Trucks in Houston-Galveston Port Area Draft Report. EP-C-06-080. May 27, 2011
27 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.
28 Rakha, Hesham and Ivana Lucic. Variable Power Vehicle Dynamics Model for Estimating Truck Accelerations. Page
6. http://filebox.vt.edu/users/hrakha/Publications/Variable%20Power%20Truck%20Acceleration%20-
%20Ver%202.0.pdf
29 National Renewable Energy Laboratory. Development of LNG-Poweredheavy-Duty Trucks in Commercial Hauling.
NREL/SR-540-25154, Golden, CO, December 1998.
30 Goodyear. "Factors Affecting Truck Fuel Economy - Section 9" Page 5.
http://www.goodvear.com/truck/pdf/radialretserv/Retread S9 V.pdf.
31 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.qut.edu.au/archive/00002359/01/RS&ETechForum2003_Ramsay&Bunker_2.pdf
292
-------
32 Society of Automotive Engineers. Commercial Truck and Bus SAE Recommended Procedure for Vehicle
Performance Prediction and Charting. SAE J2188. Revised October 2003.
33 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.
34 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.
35 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
36 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. http://data.mira.co.uk/Newsletter/MIRA-Newsletter2.pdf.
37 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 https://www.epa.gov/sites/production/files/2016-
06/documents/02-hd-emission-vehicle-test-programs.pdf.
38 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.
39 USEPA (2002). Update of Heavy-Duty Emission Levels (Model Years 1988-2004) for Use in MOBILE6. EPA-420-
R-02-018. July 2002. http://www.epa.gov/otaq/models/mobile6/r02018.pdf. (Table 17)
40 USEPA Office of Transportation and Air Quality. Heavy Duty Diesel Engine Consent Decree Low NOx Rebuild
Program Summary. Washington, D.C. http://www.epa.gov/compliance/resources/cases/civil/caa/diesel/progress.pdf.
41 USEPA (2014). Tier 3 Vehicle Emission and Fuel Standards Program. Regulatory Impact Analysis. EPA-420-R-14-
004. February 2014. http://www.epa.gov/otaq/tier3.htm
42 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.
43 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. Available at
www.crcao.org.
44 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. (http://www.epa.gov/otaq/models/ngm/420p05001.pdf). February
2005.
45 Kinsey, J. S., et al. (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/j.jaerosci.2005.03.007.
46 USEPA (2009). Development of Emission Rates for Heavy-Duty Vehicles in the Motor Vehicle Emissions Simulator
(Draft MOVES2009). EPA-420-P-09-005. Office of Transportation and Air Quality. US Environmental Protection
Agency. Ann Arbor, MI. August, 2009. http://www.epa.gov/otaq/models/moves/techdocs/420p09005.pdf.
47 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-
293
-------
005, M6.HDE.004. Assessment and Standards Division, Ann Arbor, MI.
(http://www.epa.gov/otaq/models/mobile6/r02005.pdf).
48 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.
http://www.crcao.org/reports/recentstudies2009/ACES%20Phase%201/ACES%20Phasel%20Final%20Report%2015J
UN2009.pdf
49 US EPA. 2012. Black Carbon Report to Congress. EPA-450/R-12-001. March 2012.
http://www.epa.gov/blackcarbon/2012report/fullreport.pdf.
50 Quiros, D. C., et al. (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.
51 Dixit, P., et al. (2017). Differences between emissions measured in urban driving and certification testing of heavy-
duty diesel engines. Atmospheric Environment, 166, 276-285. DOI: http://dx.doi.Org/10.1016/j.atmosenv.2017.06.037.
52 Thiruvengadam, A., et al. (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.
53 40 CFR § 86.1816-18. Table 1- Fully Phased-in Tier 3 HDV Exhaust Emission Standards.
54 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.
55 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.
56 40 CFR § 1065.260
57 40 CFR § 1065.250
58 USEP A (2012). Updates to the Greenhouse Gas and Energy Consumption Rates in MOVES2010a. EP A-420-R-12-
025. Office of Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI. August, 2012.
http://www.epa.gov/otaq/models/moves/documents/420rl2025.pdf.
59 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.
http://www.epa.gov/otaq/climate/documents/420fl4040a.pdf.
60 USEPA (2011). Greenhouse Gas Emission Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty
Engines and Vehicles (76 FR 57106, September 15, 2011)
61 40 Code of Federal Register Volume 76 at 57216 and 57236, September 15, 2011.
62 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
63 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)
64 USEPA (2014). Emission Adjustments for Temperature, Humidity, Air Conditioning, and Inspection and
Maintenance for On-road Vehicles in MOVES2014. EPA-420-R-14-012. Assessment and Standards Division. Office
of Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, MI. 2014.
http://www.epa.gov/otaq/models/moves/moves-reports.htm.
294
-------
65 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.
66 40 CFR Part 86, Appendix I (f) (2)
67 EMFAC 2000 Section 6.7 START CORRECTION FACTORS. 4/20/2000. On-Road Emissions Model Methodology
Documentation, http://www.arb.ca.gov/msei/onroad/doctable_test.htm
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. http://www.epa.gov/otaq/models/moves/moves-
reports.htm.
72 Keel-Blackmon, K., S. Curran and M. V. Lapsa (2016). Summary of OEM Idling Recommendations from Vehicle
OwnersManuals. 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.
73McCormick, 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):19928. Revised
May 3, 2000.
74Lim, 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.
75 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.
76Lambert, 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.
77Gautam, 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.
78National 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.
79Tang, 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.
80Brodrick, 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.
295
-------
81Storey, 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.
82 Keel-Blackmon, K., S. Curran and M. V. Lapsa (2016). Summary of OEM Idling Recommendations from Vehicle
OwnersManuals. 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.
83 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
84 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.
85 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.
86 Farzaneh, M., et al. (2014). TEXAS-SPECIFIC DRIVE CYCIES AND IDIE 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.
87 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.
88 California Environmental Protection Agency. Heavy-Duty Engines and Vehicles Executive Orders. Motor Vehicle
and Engine Certification Program. Air Resources Board. Available at:
http://www.arb.ca.gOv/msprog/onroad/cert/cert.php#3
89 US EPA, Navistar Inc. Heavy-Duty Engine Recall. EPA-420-F-13-038, Ann Arbor, MI: June 2013,
https://www3.epa.gov/otaq/cert/eng-recall/420fl3038.pdf.
90 Clark, N. and M. Gautam (2007). HEAVY-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.
91 US EPA, Engine Certification Data. On-Highway Heavy Duty - Diesel and Gasoline (2012).
https://www3.epa.gOv/otaq/certdata.htm#oh
92 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.
93 Khan, A. S., et al. (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.
94 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.
95 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.
96 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.
296
-------
97 Storey, J. M., et al. (2003). Particulate matter and aldehyde emissions from idling heavy-duty diesel trucks. SAE
Technical Paper.
98 See 40 CFR 1037.106(g).
99 US EPA, Nonroad Compression-Ignition Engines: Exhaust Emission Standards, EPA-420-B-16-022, March 2015.
https://www.epa.gov/sites/production/files/2016-03/documents/420bl6022.pdf
100 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.
101 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/ZyPDF.cgi?Dockey=P100MUVI.PDF
102 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.regulations.gov/document?D=EPA-HQ-OAR-2014-0827-2417
103 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://www.epa.gov/moves/mobile-source-observation-database-msod.
104 EPA. 1999. Tier 2 Vehicle & Gasoline Sulfur Program Final Rule, http://www.epa.gov/otaq/standards/light-
dutv/tier2stds.htm
105 USEPA (2017). Light-Duty Vehicles, Light-Duty Trucks, and Medium-Duty Passenger Vehicles: Tier 2 Exhaust
Emission Standards and Lmplementation 5cAec/u7eEPA-420-B-17-028. Office of Transportation and Air Quality. US
Environmental Protection Agency. September 2017.
https://nepis.epa.gOv/Exe/ZyPDF.cgi/P 100SMQA.PDF?Dockey=P 100SMQA.PDF.
106 Development of Emission Rates for Heavy-Duty Vehicles in the Motor Vehicle Emissions Simulator MOVES2010
(131 pp, EPA-420-B-12-049, August 2012)
107 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" http://www.epa.gov/otaq/standards/heavv-dutv/hdsi-
exhaust.htm108 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.
109 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.
110 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.
111 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=P 1000A01.pdf.
112 Toro, C., J. Warila, D. Sonntag, D. Choi and M. Beardsley (2019). Updates to "high-power" emission rates and
start deterioration for light-duty vehiclesMOVES 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.
297
-------
113 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 Environmental Protection Agency.
Ann Arbor, MI. October, 2015. https://www.epa.gov/moves/moves-technical-reports.
114 40 CFR Part 86 Appendix 1(f) (1)
115 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.
116 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/2003-01-1900).
117 "Central Business District (CBD)," Emission Test Cycles. DieselNet. Last Updated: September 2000, Accessed: 10
July 2012. (http://www.dieselnet.com/standards/cycles/cbd.php).
118 DieselNet, (http://www.dieselnet.com/standards/cycles/cbd.php).
119 Hesterberg, T.; Lapin, C.; Bunn, W.; 2008. "A Comparison of Emissions from Vehicles Fueled with Diesel or
Compressed Natural Gas." Environ. Sci. Techno! 42(17): 6437-6445. (http://pubs.acs.org/doi/abs/10.1021/esQ71718i).
120 Melendez, M.; Taylor, J.: Zuboy, J. et al. Emission Testing of Washington Metropolitan Area Transit Authority
(WMATA) 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).
121 LeTavec, C., Uihlein, J., Vertin, K., Chatterjee, 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).
122 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).
123 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/2003-01-0300).
124 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).
125 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).
126 McCormick, 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).
127 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
128 "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.
129 US EPA OTAQ, 2012. (http://iaspub.epa.gov/otaqpub/pubsearch.jsp).
130 40 CFR 86.105-94. "Clean-fuel fleet emission standards for heavy-duty engines." Code of Federal Regulations.
298
-------
131 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/pubsearch.jsp).
132 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/j.energy.2013.09.040.
133 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/10.1080/10962247.2013.800170.
134 Jaaskelainen, H. Crankcase Ventilation. DieselNet Technology Guide. www.DieselNet.com. Copyright © Ecopoint
Inc. Revision 2012.12.
135 4 0 CFR 86.004-11. "Control of Emissions from New and In-Use Highway Vehicles and Engines." Code of Federal
Regulations.
136 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.
137 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
138 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.
139 40 CFR 86.007-11. "Control of Emissions from New and In-Use Highway Vehicles and Engines." Code of Federal
Regulations.
140 Khalek, I. A.: Bougher, T. L; Merrit, P. M.; Phase 1 of the Advanced Collaborative Emissions Study. CRC Report:
ACES Phase 1, June 2009.
141 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. Available at:
http://www.worldnci.com/files/Documents/WVA_FinalReport_200600808.pdf
142 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.
143 Hare, C. T.; Baines, T. M.; Characterization of Diesel Crankcase Emissions. Society of Automotive Engineers, Off-
Highway Vehicle Meeting and Exhibition. MECA, Milwaukee. 1977.
144 Kalayci, Veli. "Spiracle™ Crankcase Filtration Systems: Technical Article" . Donaldson Company, Inc. January
2011. Available at: www.donaldson.com/en/engine/support/datalibrary/053490.pdf
145 Gerhardt, M. J., et al. (2020). Crankcase Emissions for MY2007+ Heavy-Duty Diesel Trucks. MOVES Review
Workgroup, Ann Arbor, MI. October 15, 2020. https://www.epa.gov/moves/moves-model-review-work-group.
146 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.
147 Seinfeld, J. H. and S. N. Pandis (2012). Atmospheric chemistry and physics: from air pollution to climate change.
John Wiley & Sons.
148 Dunlea, E. J., et al. (2007). Evaluation of nitrogen dioxide chemiluminescence monitors in a polluted urban
environment. Atmos. Chem. Phys., 7 (10), 2691-2704. DOI: 10.5194/acp-7-2691-2007.
299
-------
149 Kurtenbach, R., et al. (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.
150 Wang, Y. J., A. DenBleyker, E. McDonald-Buller, D. Allen and K. M. Zhang (2011). Modeling the chemical
evolution of nitrogen oxides near roadways. Atmospheric Environment, 45 (1), 43-52. DOI:
http://dx.doi.Org/10.1016/j.atmosenv.2010.09.050.
151 USEPA (2009). Development of Emission Rates for Heavy-Duty Vehicles in the Motor Vehicle Emissions Simulator
(Draft MOVES2009). EPA-420-P-09-005. Office of Transportation and Air Quality. US Environmental Protection
Agency. Ann Arbor, MI. August, 2009. http://www.epa.gov/otaq/models/moves/techdocs/420p09005.pdf.
152 U.S. Census Bureau. Vehicle Inventory and Use Survey. Service Sector Statistics Division, Transportation
Characteristics Branch, http://www.census.gov/svsd/www/vius/products.html.
153 Zhou, Lei. Revision of Heavy Heavy-Duty Diesel Truck Emission Factors and Speed Correction Factors. California
Air Resources Board, Sacramento. October 2006.
154 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.
www.epa.state.il.us/air/publications/obd-report-final.pdf
155 Manufacturers of Emission Controls Association. Technology Details - Catalytic Converters - SCR System.
http://www.meca.org/technology/technology-details?id=5
156 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.
157 Darlington, T., Dennis Kahlbaum and Gregory Thompson. On-Road NOxEmission Rates from 1994-2003 Heavy-
Duty Diesel Trucks. SAE 2008-01-1299. Society of Automotive Engineers, April 2008.
158 Preble, C. V., et al. (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.
159 Bishop, G. A., et al. (2015). On-road Heavy-duty Vehicle Emissions Monitoring System. Environ Sci Technol, 49
(3), 1639-1645. DOI: 10.1021/es505534e.
160 ARB (2015). Evaluation of Particulate Matter Filters in On-Road Heavy-Duty Diesel Vehicle Applications.
California Air Resources Board. May 8, 2015. http://www.arb.ca.gov/msprog/onrdiesel/documents/DPFEval.pdf.
161 40 C.F.R. §86.004-2. Revised as of Aug. 8, 2014.
162 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.regulations.gov/document?D=EPA-HQ-OAR-2014-0827-2417
300
------- |