Exhaust Emission Rates for Heavy-Duty
Onroad Vehicles in MOVES5

£%	United States

Environmental Protect
Agency


-------
Exhaust Emission Rates for Heavy-Duty

This technical report does not necessarily represent final EPA decisions
or positions. It is intended to present technical analysis of issues using
data that are currently available. The purpose in the release of such
reports is to facilitate the exchange of technical information and to
inform the public of technical developments.

Onroad Vehicles in MOVES5

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

NOTICE

4>EPA

United States
Environmental Protection
Agency

EPA-420-R-24-015
November 2024


-------
TABLE OF CONTENTS

TABLE OF CONTENTS	i

TABLE OF FIGURES	iii

TABLE OF TABLES	viii

1	Introduction	1

1.1	Pollutant	2

1.2	Emission Process	2

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	12

2	Heavy-Duty Diesel Exhaust Emissions	13

2.1	Running Exhaust Emissions	13

2.1.1	Nitrogen Oxides (NOx)	13

2.1.2	Particulate Matter (PM25)	59

2.1.3	Total Hydrocarbons (THC) and Carbon Monoxide (CO)	79

2.1.4	Energy	91

2.1.5	Evaluation of Fleet-average Running Rates with Real-World Measurements	102

2.2	Start Exhaust Emissions	102

2.2.1	THC, CO, and NOx	103

2.2.2	Particulate Matter (PM25)	112

2.2.3	Adjusting Start Rates for Soak Time	113

2.2.4	Start Energy Rates	124

2.3	Extended Idling Exhaust Emissions	126

2.3.1	1950-2006 Model Years	127

2.3.2	2007-2026 Model Years	130

2.3.3	2027-2060 Model Years	139

2.3.4	Model Year Trends	139

2.3.5	Extended Idle Energy Rates	142

2.4	Auxiliary Power Unit Exhaust	143

2.5	Glider Vehicle Emissions	148

3	Heavy-Duty Gasoline Exhaust Emissions	150

3.1	Running Exhaust Emissions	150

3.1.1	THC, CO, and NOx	150

3.1.2	Particulate Matter (PM25)	169

3.1.3	Energy	177

3.2	Start Emissions	182

3.2.1	THC, CO, and NOx	183

3.2.2	Particulate Matter (PM25)	197

3.2.3	Soak Time Adjustments	198

3.2.4	Start Energy Rates	199

i


-------
4	Heavy-Duty Compressed Natural Gas Exhaust Emissions	200

4.1	Running Exhaust Emission Rates	201

4.1.1	1950-2009 Model Years	201

4.1.2	2010-2060 Model Years	211

4.1.3	Model Year Trends	212

4.2	Start Exhaust Emission Rates	217

4.3	Extended Idle Exhaust Emission Rates	217

5	Heavy-Duty Ammonia Emissions	217

5.1	Heavy-Duty Diesel	218

5.2	Heavy-Duty Gasoline	222

5.2.1	1950-1980 Model Years	222

5.2.2	1981-2060 Model Years	223

5.3	Heavy-Duty Compressed Natural Gas	224

5.4	Summary	227

6	Heavy-Duty Crankcase Exhaust Emissions	229

6.1	Modeling Crankcase Emissions in MOVES	229

6.1.1	Heavy-Duty Diesel Crankcase Emissions	230

6.1.2	LHD2b3 Crankcase Emissions	231

6.1.3	LHD45, MHD and HHD Crankcase Emissions	232

6.1.4	Glider Crankcase Emissions	248

6.2	Heavy-Duty Gasoline and CNG Crankcase Emissions	248

7	Nitrogen Oxide Composition	249

7.1	Heavy-Duty Diesel	250

7.2	Heavy-Duty Gasoline	253

7.3	Heavy-Duty Compressed Natural Gas	253

8	Appendices	255

References	314

ii


-------
TABLE OF FIGURES

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	23

Figure 2-2 NOx Emissions by Operating Mode from MHD, HHD, and Urban Bus Regulatory Classes for Model

Year 2002	 24

Figure 2-3. HHD NOx emission rates for OpModelD 37 for Model Years 1991-1998 Adjusted for Low-NOx

Rebuilds by Vehicle Age	26

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	30

Figure 2-5 Distribution of NOx Family Emission Limit (FEL) for Model Year 2010 -2015 Heavy-Duty Diesel

Engine Families, as Reported in the Certification Database	31

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

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	33

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	34

Figure 2-9 NOx emission rates for the MY 2010-2013 and MY 2014-2016 vehicles in the LHD 0.20 NOx FEL

Group	35

Figure 2-10 NOx emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the MHD 0.20 NOx FEL

Group	35

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	36

Figure 2-12 Production Volume Contribution of Heavy-Duty Diesel Engine Families by NOx Family Emission

Limit Group for Each Regulatory Class	37

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	38

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	38

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	39

Figure 2-16 Percent change in in-use emission rates for 2010 standard (0.2 g/hp-hr) compliant HHD
vehicles, compared to the percent change in the 2010 duty-cycle standard across MOVES operating modes

	43

Figure 2-17 Duty-cycle-based running NOx emissions, ERdUtyjn_use, for HHD diesel for the HD2027 standards

	46

Figure 2-18 Base NOx rates and off-cycle NOx standard compliant emission rates for HHD diesel	50

Figure 2-19 Comparison of Running NOx emission rates for diesel-fueled HHD compliant with the duty-cycle

and off-cycle standards	51

Figure 2-20 Estimated zero-mile NOx emission rates for HHD diesel vehicles due to the HD2027 duty-cycle

and off-cycle standards	52

Figure 2-21 Estimated zero-mile NOx emission rates for MHD diesel vehicles due to the HD2027 duty-cycle
and off-cycle standards	52


-------
Figure 2-22 Estimated zero-mile NOx emission rates for LHD45 diesel vehicles due to the HD2027 duty-

cycle and off-cycle standards	53

Figure 2-23 Base Heavy-duty Diesel NOx Running Emission Rates (g/mile) by Age for Model Year 2015 by

Regulatory Class Estimated using Nationally Representative Operating Mode Distribution	56

Figure 2-24 Base running emission rates for NOx from age 0-3 diesel heavy-duty vehicles averaged over a

nationally representative operating mode distribution	58

Figure 2-25 Base running emission rates for NOx from age 0-3 gasoline heavy-duty vehicles averaged over a

nationally representative operating mode distribution for model years 2007 and later	59

Figure 2-26 MHD Diesel PM2.5 Emission Rates for MY 2006 (age 0-3) by Operating Mode	68

Figure 2-27 MHD Diesel PM2.5 Emission Rates for MY 2007 (age 0-3) by Operating Mode	68

Figure 2-28 PM2 5 Emissions by Operating Mode for LHD Model Year Groups 2010-2013 and 2014-2016... 72
Figure 2-29 PM2.5 Emissions by Operating Mode for MHD Model Year Groups 2010-2013 and 2014-2015.72
Figure 2-30 PM2.5 Emissions by Operating Mode for HHD Model Year Groups 2010-2013 and 2014-2015.. 73
Figure 2-31 Heavy-duty Diesel PM2.5 Emission Rates (g/mile) by Age Group and Regulatory Class for Model

Year 2006 using Nationally Representative Operating Mode Distribution	76

Figure 2-32 Heavy-duty Diesel PM2 5 Emission Rates (g/mile) by Age Group and Regulatory Class for Model

Year 2015 using Nationally Representative Operating Mode Distribution	77

Figure 2-33. Base running emission rates for PM2.5from age 0-3 diesel heavy-duty vehicles averaged over a

nationally representative operating mode distribution	78

Figure 2-34 Base running emission rates for PM2.5from age 0-3 diesel heavy-duty vehicles for MY 2007 and

later averaged over a nationally representative operating mode distribution	78

Figure 2-35 Heavy Heavy-duty (HHD) Diesel PM2.5 Emission Rates by Elemental Carbon (EC) and Non-
Elemental Carbon (nonEC) Fraction for the 0-3 Age Group by Model Year using Nationally Representative

Operating Mode Distributions	79

Figure 2-36 THC Emission Rates [g/hr] by Operating Mode for Model Year 2002 and Age Group 0-3. Error

Bars Represent the 95 Percent Confidence Interval of the Mean	82

Figure 2-37 CO Emission Rates [g/hr] by Operating Mode for Model Year 2002 and Age Group 0-3. Error

Bars Represent the 95 Percent Confidence Interval of the Mean	82

Figure 2-38 Average HHD THC Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013

and the 0.35 and 0.5 NOx FEL for MY 2010-2015. Error Bars are 95% Confidence Intervals of the Mean	84

Figure 2-39 Average HHD CO Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013 and

the 0.35 and 0.5 NOx FEL for MY 2010-2015. Error Bars are 95% Confidence Intervals of the Mean	84

Figure 2-40 THC Emissions by Operating Mode from HHD Trucks for Model Year 2013. Error Bars represent

the 95 percent confidence interval of the Mean	85

Figure 2-41 CO Emissions by Operating Mode from HHD Trucks for Model Year 2013. Error Bars represent

the 95 percent confidence interval of the Mean	85

Figure 2-42 THC emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the HHD 0.20 NOx FEL

Group	86

Figure 2-43 CO emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the HHD 0.20 NOx FEL

Group	86

Figure 2-44 Heavy-duty Diesel THC Running Emission Rates (g/mile) by Age for Model Year 2015 by

Regulatory Class Estimated using Nationally Representative Operating Mode Distribution	89

Figure 2-45 Heavy-duty Diesel CO Running Emission Rates (g/mile) by Age for Model Year 2015 by
Regulatory Class Estimated using Nationally Representative Operating Mode Distribution	89

iv


-------
Figure 2-46 Base running emission rates for THCfrom age 0-3 diesel heavy-duty vehicles averaged over a

nationally representative operating mode distribution	90

Figure 2-47 Base running emission rates for THC from age 0-3 diesel heavy-duty vehicles for MY 2007 and

later averaged over a nationally representative operating mode distribution	91

Figure 2-48 Base running emission rates for CO from age 0-3 diesel heavy-duty vehicles averaged over a

nationally representative operating mode distribution	91

Figure 2-49 Diesel running exhaust energy rates for MHD, HHD, and Urban Buses for 1950-2009 model

years. Error bars represent the 95 percent confidence interval of the mean	93

Figure 2-50 Average HHD C02 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	94

Figure 2-51. Base running emission rates for C02from age 0-3 diesel heavy-duty vehicles averaged over a

nationally representative operating mode distribution	101

Figure 2-52. Fuel economy for age 0-3 diesel heavy-duty vehicles averaged over a nationally representative

operating mode distribution	101

Figure 2-53 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)	104

Figure 2-54: Relationship between HD Diesel Engine Rated Power and FTP Cycle Work	106

Figure 2-55 Calculated relationship between the HHD NOx 12-hour cold-start and the composite FTP NOx

standards	109

Figure 2-56 Heavy-duty Diesel THC Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory

Class and Model Year	110

Figure 2-57 Heavy-duty Diesel CO Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class

and Model Year	Ill

Figure 2-58 Heavy-duty Diesel NOx Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory

Class and Model Year	Ill

Figure 2-59 Heavy-duty Diesel PM25 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	113

Figure 2-60. 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 Report	114

Figure 2-61 National Renewable Energy Laboratory's Heavy-Duty Vocational Transient Cycle	116

Figure 2-62 MY 2015 Heavy-Duty Vehicle NOx Emissions by Soak Time	117

Figure 2-63 MY 2015 Heavy-Duty Vehicle THC Emissions by Soak Time	117

Figure 2-64 MY 2015 Heavy-Duty Vehicle CO Emissions by Soak Time	118

Figure 2-65 MY 2015 Heavy-Duty Vehicle PM25 Emissions by Soak Time	118

Figure 2-66 Onroad Soak Drive Route	119

Figure 2-67 MY 2016 Heavy-Duty Vehicle NOx Emissions by Soak Time	119

Figure 2-68 MY 2016 Heavy-Duty Vehicle THC Emissions by Soak Time	120

Figure 2-69 MY 2016 Heavy-Duty Vehicle CO Emissions by Soak Time	120

Figure 2-70 Soak Emission Ratios from a MY 2015 HD Day-Cab and a MY 2016 HD Van	121

Figure 2-71 Duty-cycle-based NOx start emissions for HHD Diesel comparing MY2010-2026 and MY2027+

rates	122

Figure 2-72 Heavy-Duty Diesel Cold Start C02 Rates (Operating Mode 108) by Model Year and Regulatory
Class	125

v


-------
Figure 2-73. C02 Emission Rates from the TTI and ARB Programs by Engine Model Year, and Average

Emission Rate (line) based on all the data	133

Figure 2-74. CO Emission Rates from the TTI and ARB Programs by Engine Model Year, and Average

Emission Rate (line) Based on All the Data	134

Figure 2-75. 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)	135

Figure 2-76. 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)	136

Figure 2-77. 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)	137

Figure 2-78. Extended Idle NOx Emission Rates for HHD and MHD Diesel Vehicles by Model Year	140

Figure 2-79. Extended Idle THC Emission Rates for HHD and MHD Diesel Vehicles by Model Year	141

Figure 2-80 Extended Idle CO Emission Rates for HHD and MHD Diesel Vehicles by Model Year	141

Figure 2-81. Extended Idle PM2.5 Emission Rates HHD and MHD Diesel Vehicles by Model Year	142

Figure 2-82. Extended Idle Energy Emission Rates for HHD and MHD Diesel Trucks	143

Figure 2-83. Comparison of the running exhaust emission rates (0-3 age group) of HHD (regClassID 47) vs.

Gliders (regClassID 49) for selected pollutants (NOx, ECPM, NonECPM) and model year groups	149

Figure 3-1 Brake-Specific Certification Emission Rates by Model Year for Heavy-Duty Gasoline Engines... 151
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	152

Figure 3-3. Emission Rates by Operating Mode and Age Group for MY 1998-2007 Vehicles in Regulatory

Class LHD2b3 and LHD45	154

Figure 3-4. Emission Rates for MY 1990-1997 at age 0-3 years for Regulatory Class MHD and HHD	155

Figure 3-5. Emission Rates by Operating Mode and Age Group for MY 1998-2007 Vehicles in Regulatory

Class MHD and HHD Gasoline Vehicles	157

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	160

Figure 3-7. Base running emission rates for THCfrom age 0-3 gasoline heavy-duty vehicles averaged over a

nationally representative operating mode distribution	167

Figure 3-8. Base running emission rates for CO from age 0-3 gasoline heavy-duty vehicles averaged over a

nationally representative operating mode distribution	168

Figure 3-9. Base running emission rates for NOx from age 0-3 gasoline heavy-duty vehicles averaged over a

nationally representative operating mode distribution	169

Figure 3-10. Base running emission rates for PM2.5 from age 0-3 gasoline heavy-duty vehicles averaged over

a nationally representative operating mode distribution	176

Figure 3-11 Base running emission rates for PM2.5 from age 0-3 gasoline heavy-duty vehicles for MY2007

and later averaged over a nationally representative operating mode distribution	176

Figure 3-12 Heavy Duty Gasoline Running Exhaust PM2.5 Emission Rates by Elemental Carbon and Non-
Elemental Carbon (nonEC) Fractions for the 0-3 Age Group by Model Year and Regulatory Class using

Nationally Representative Operating Mode Distributions	177

Figure 3-13. Gasoline Running Exhaust Energy Rates for MHD (1960-2009) and HHD (1960-2009)	178

Figure 3-14. Base running emission rates for C02 from age 0-3 gasoline heavy-duty vehicles averaged over a
nationally representative operating mode distribution	181

vi


-------
Figure 3-15. Fuel economy for age 0-3 gasoline heavy-duty vehicles averaged over a nationally

representative operating mode distribution	182

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)	186

Figure 3-17. Cold-Start Emission Rates for Heavy-Duty Gasoline Trucks, with 95 Percent Confidence

Intervals	188

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	195

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	195

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	196

Figure 3-21 Heavy-duty Gasoline Cold-Start Rates (opModelD 108) vs. Vehicle Age for Select Model Years

LHD45, MHD and HHD are equivalent	196

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	198

Figure 3-23 HD Gasoline Start Emission Ratio Compared to Recent Data	199

Figure 3-24 Heavy-Duty Gasoline Cold Start C02 Emission Rates (g/start) by Model Year and Regulatory

Class (OpmodelD 108)	200

Figure 4-1 Driving Schedule Trace of the Central Business District (CBD) Cycle	202

Figure 4-2 Driving Schedule Trace of the Washington Metropolitan Area Transit Authority (WMATA) Cycle

	203

Figure 4-3 Operating Mode Distribution for the CBD Cycle	207

Figure 4-4 Operating Mode Distribution for the WMATA Cycle	207

Figure 4-5. Base running emission rates for THC from age 0-3 CNG heavy-duty vehicles averaged over a

nationally representative operating mode distribution	213

Figure 4-6. Base running emission rates for CO from age 0-3 CNG heavy-duty vehicles averaged over a

nationally representative operating mode distribution	214

Figure 4-7. Base running emission rates for NOx from age 0-3 CNG heavy-duty vehicles averaged over a

nationally representative operating mode distribution	215

Figure 4-8. Base running emission rates for PM2.5 from age 0-3 CNG heavy-duty vehicles averaged over a

nationally representative operating mode distribution	216

Figure 4-9. Base running emission rates for C02 from age 0-3 CNG heavy-duty vehicles averaged over a

nationally representative operating mode distribution	217

Figure 5-1. NH3 emission rates (g/hour) by operating mode for regulatory class LHD2b3 and HHD and

Model Years 2017 for all ages	222

Figure 5-2. NH3 emission rates (g/hour) by operating mode for regulatory class LHD2b3 and LHD45 and

Model Years 2017 for ages 0-3	224

Figure 6-1 Speed trace of the NVFEL testing program along with each phase	240

vii


-------
TABLE OF TABLES

Table 1-1 Emission Processes for Onroad Heavy-Duty Vehicles	4

Table 1-2 Regulatory Classes for Heavy-Duty Vehicles	6

Table 1-3 Fixed Mass Factor, fscaie (metric tons)	10

Table 1-4 Operating Mode Definition for Running Exhaust for Heavy-Duty Vehicles	11

Table 1-5 Operating Modes for Start Emissions (as a function of soak time)	12

Table 1-6 MOVES Age Group Definitions	12

Table 2-1 Model Year Groups for NOx Analysis Based on Emissions Standards	14

Table 2-2 Numbers of Diesel Vehicles from the ROVER, WVU MEMS, HDIUT, and Houston Drayage

Programs by Model Year Group	16

Table 2-3 Accessory Use as a Function of Speed and Load Ranges, Coded by Power Level	18

Table 2-4 Estimates of Accessory Load in kW by Engine Power Level	19

Table 2-5 Driveline Efficiencies Found in the Literature	20

Table 2-6 NOx Family Emission Limit (FEL) based Groups for LHD, MHD, HHD, and Urban Bus Classes in the

HDIUT Data	30

Table 2-7 Number of 2010 and Later Model Year HDIUT Vehicles by NOx FEL Group and Model Year Groups1

Used for Emission Rate Analysis	32

Table 2-8 Phase-in Assumptions for Tier 3 NOx Standards for Light Heavy-Duty 2b3 Diesel Vehicles	40

Table 2-9 Heavy-duty Compression Ignition Duty-Cycle NOx Standards for the HD2027 Rule	41

Table 2-10 Rduty Ratios Calculated for the HD2027 Standards	41

Table 2-11 Calculation of Rin_use by MOVES Operation Mode	44

Table 2-12 Calculated Off-Cycle NOx Standards based on the HD2027 Standards	47

Table 2-13 Calculation of Optional Idle NOx /C02 Standard (g/kg)	47

Table 2-14 Calculation of the Off-cycle NOx Standard Compliant Emission Rate for HHD Diesel Vehicles for

the HD2027 Standards	49

Table 2-15 NOx T&M adjustment factor (percent) by regulatory class	54

Table 2-16 Summary of Methods for Heavy-Duty Diesel NOx Emission Rate Development for Each

Regulatory Class and Model Year Group	56

Table 2-17 Model Year Groups Used for Analysis Based on the PM Emissions Standard	60

Table 2-18 Vehicle and Test Counts by Regulatory Class and Model Year Group	62

Table 2-19 Vehicle Test Counts by Test Cycle	62

Table 2-20 Vehicle and Test Counts by Heavy-Duty Class and Model Year	63

Table 2-21 Urban Bus PM Standards in Comparison to Heavy-Duty Highway Compression Engine Standards

	65

Table 2-22 Average Certification Results for Model Years 2003-2007	66

Table 2-23. HDIU Vehicles with Valid PM25 Measurements By Regulatory Class, Model Year Group and NOx

FEL Group	69

Table 2-24. Range of Valid Vehicle Measurements by Operating Mode by Regulatory Class and Model Year

Group	70

Table 2-25 Tampering and Mal-maintenance Adjustment Factors for PM2.5 over the Useful Life2	75

Table 2-26 Numbers of Vehicles by Model Year Group, Regulatory Class, and Age Group	80

Table 2-27 Age Groups for which THC and CO Emission Rates are Populated Directly Based on the Data... 81
Table 2-28 Tampering and Mal-maintenance Effects for THC and CO over the Useful Life	88


-------
Table 2-29 HDIU Vehicles with Valid C02 Measurements By Regulatory Class, Model Year Group and NOx

FEL Group	94

Table 2-30 Estimated Reductions in Diesel Engine Energy Consumption Rates from the HD GHG Phase 1

Program	96

Table 2-31 Estimated Reductions in Diesel and CNG Engine Energy Consumption Rates due to the HD GHG

Phase 2 Program61	97

Table 2-32 Phase 3 fleet-average C02 emission percent reductions	98

Table 2-33 Phase 3 regulatory groupings mapped to MOVES vehicle types	98

Table 2-34 Estimated Total Vehicle Reductions in Energy Consumption Rates for LHD2b3 Diesel Vehicles

due to the HD GHG Phase 1 Program	99

Table 2-35 Estimated Total Vehicle Reductions in Fleet Average Energy Consumption Rates for LHD2b3 due

to the HD GHG Phase 2 Program	100

Table 2-36. Comparison of MOVES Emissions with Remote Sensing and Tunnel Measurements (g/kg-fuel)

	102

Table 2-37 Average Start Emissions Increases (g/start) for pre-2010 Model Year Light Heavy-Duty Diesel

Vehicles for Regulatory Class LHD2b3 and LHD45 (regClassID 41 and 42)	103

Table 2-38. Cold-start Emissions Increases (g/start) in Grams on the 2007 Cummins ISB	104

Table 2-39. MOVES Inputs for Pre-2010 HHD and MHD Diesel Start Emissions (grams/start) for Regulatory

Class 46, 47, and 48. No Differentiation by Model Year or Age	105

Table 2-40 Engine Data Analyzed to Estimate the Cold Start Emission Rates for HD Diesel Engines	105

Table 2-41: Cold Start Emissions for MY 2010-2026 Heavy Heavy-Duty Diesel Engines	107

Table 2-42 Cold Start Emissions for MY 2010-2026 Medium Heavy-Duty Diesel Engines	107

Table 2-43 Cold Start Emissions for MY 2010-2026 Light Heavy-Duty Diesel Engines (LHD45 and LHD2b3)

	107

Table 2-44 Calculation of NOx 12-hour Cold Starts from the CARB Stage 1 HHD Engine from the Cold and

Hot FTP Cycle	108

Table 2-45 HHD Cold Start Emissions for Baseline and HD2027 Standards	109

Table 2-46 Average PM25 emissions (grams) from MY 2004 HHD diesel engine tested on the FTP Cycle.. 112

Table 2-47: Cold Start PM25 Emission Rates for Heavy-Duty Diesel Emissions for 2010+ MY	113

Table 2-48. Heavy-Duty diesel THC, CO, and NOx Start Emissions (g/start) by Operating Mode for 2009 and

Earlier Model Year Vehicles	115

Table 2-49. Particulate Matter Start Emission Rates (g/start) by Operating Mode (soak fraction) for all HD

Diesel vehicles through MY 2009	115

Table 2-50 HD Diesel Engine Soak Ratios for MY 2010 and Newer	121

Table 2-51 PM25 Start Emission Rates (g/start) by Regulatory Class and Operating Mode (soak fraction) for

all MY 2010 and newer HD Diesel Vehicles	123

Table 2-52 THC Diesel Start Ambient Temperature Adjustment by Operating Mode	124

Table 2-53 Fraction of Energy Consumed at Start of intermediate Soak Lengths compared to the Energy

Consumed at a Full Cold Start (operating mode 108)	125

Table 2-54. Relative contribution of total energy consumption from each pollutant process by regulatory

class for heavy-duty diesel vehicles in calendar year 2016	126

Table 2-55. Pre-2007 Extended idle emission rates (g/hour) in MOVES by pollutant for MHD and HHD.... 130

Table 2-56. HHD Diesel Tractors Used to Update the MY 2007-2026 Extended Idle Emission Rates	131

Table 2-57. Ambient Test Conditions for the TTI Extended Idle Tests	132

ix


-------
Table 2-58. Baseline elemental carbon to PM2.5 fraction assumed for extended idling	137

Table 2-59. Extended Idle Emission Rates for 2007 and Later Model Year Heavy-Duty Vehicles	139

Table 2-60 Calculation of HHD and MHD Extended Idle NOx g/hr Emission Rates for MY2027+	139

Table 2-61: Projected APU Use during Extended Idling for Combination Long-Haul Tractor-Trailers	143

Table 2-62. APU Engines and Studies Used in This Analysis	144

Table 2-63. In-Use APU Emission Rates	145

Table 2-64. Average APU Emission Rates from non-DPF APU IDs 1, 2, and 3 according to Cold, Hot, and Mild

Ambient Conditions	146

Table 2-65. Average Elemental Carbon/Total Carbon Ratio for APU ID 1 without and with a Diesel

Particulate Filter (DPF)	146

Table 2-66. Fleet-Average Non-DPF Equipped APU Emission Rates in MOVES	146

Table 2-67: Nonroad (8
-------
Table 3-21 Multiplicative Age Effect Used for Start Emissions for Gasoline LHD2b3 Vehicles for 2009-2017

Model Years Adopted from the Deterioration Effects for Light-Duty Trucks	190

Table 3-22 Multiplicative Age Effect Used for Start Emissions for Gasoline LHD2b3 Vehicles for 2022-2060

Model Years	191

Table 3-23 Engine Data Analyzed to Revise the Cold Start Emission Rates for HD Gasoline Engines	193

Table 3-24 Cold Start Emissions for MY 2008 and Later Heavy-Duty Gasoline Engines	193

Table 3-25 Summary of Cold Start Emission Rates for Heavy-Duty Gasoline Vehicles	194

Table 3-26 Relative Contribution of Total Energy Consumption from Each Pollutant Process by Regulatory

Class for Heavy-Duty Gasoline Vehicles in Calendar Year 2016	200

Table 4-1. Summary of External Emissions Testing Programs by Driving Cycle and Number of Unique

Measurements and their Corresponding Model Years	204

Table 4-2 Model Year Group Based Certification Emission Rate for Heavy-Duty CNG Engine Families	206

Table 4-3 Ratios Applied to MHD Gasoline Rates to Compute CNG Rates	210

Table 4-4 Age-based Deterioration Factor for PM2.5 Emission Rates for HHD and Urban Bus CNG Vehicles in

Model Year 2010+	212

Table 5-1. Fleet-average fuel-based NH3 emission rates (± 95% Confidence Intervals) from heavy-duty

vehicles reported from recent studies	219

Table 5-2. Fuel-based NH3 emission rates (± 95% Confidence Intervals) from heavy-duty vehicles by

aftertreatment and engine model year measured at the Caldecott Tunnel by Preble et al. (2019)	220

Table 5-3 Literature used to inform heavy-duty CNG emission rates in MOVES	226

Table 6-1 Speciation of Chained Crankcase Pollutants from the Primary Pollutants	230

Table 6-2 Literature Review on the Contribution of Crankcase Emissions to Diesel Exhaust (Tailpipe +

Crankcase)	231

Table 6-3 LHD45, MHD, and HHD 1960-2006 Diesel Crankcase Ratios for HC, CO, NOx, and PM2.5	232

Table 6-4. LHD45, MHD, and HHD Exhaust and Crankcase Ratios for 1960-2006 Diesel by Pollutant, Process,

and Model Year Group for PM2.5 Species	233

Table 6-5 The MY 2007 crankcase exhaust rates (CCACESavg) from ACES Phase 1 FTP cycle used for

running and extended idling (g/hr)	235

Table 6-6: The MY 2007-2009 MOVES exhaust base rates (g/hr) for running (weighted by operating mode

activity) and extended idling	235

Table 6-7 Fraction of Engines with Open Crankcase Systems by Vehicle Regulatory Class and Model Year236
Table 6-8 Crankcase/Tailpipe Ratios for Model Year 2007-2009 Engine-certified Vehicles by Heavy-duty

Diesel Regulatory Class	238

Table 6-9. MOVES Exhaust and Crankcase Ratios for Model Year 2007-2009 HHD Diesel by Pollutant and

Process for PM2 5 Species	239

Table 6-10: Testing information for the NVFEL test program	240

Table 6-11 NVFEL Vehicle average emission rates by testing phase and truck	241

Table 6-12. Assignment of MOVES Operating Modes to Test Phase from the NVFEL data, and Phase
weighting determined from the Operating Mode activity from a National draft MOVES run for MY 2015

HHD Vehicles in the Short- and Long-haul Single-unit and Combination Trucks Source Types	242

Table 6-13 Average Crankcase Emission Rates for MOVES Processes from NVFEL data	244

Table 6-14 The NOx FEL 0.2 MOVES Exhaust Base Rates for Running (Weighted by Operating Mode Activity)

and Extended Idling (g/hr)	244

Table 6-15 Crankcase/Tailpipe Ratios for Model Year 2010-2026 Heavy-Duty Diesel Regulatory Classes.. 244

xi


-------
Table 6-16. MOVES Exhaust and Crankcase Ratios for 2010-2026 MHD and HHD Diesel by Pollutant,

Process, and Model Year Group for PM2.5 Species	247

Table 6-17 Crankcase to Tailpipe Exhaust Emission Ratio for Heavy-Duty Gasoline and CNG Vehicles for

THC, CO, NOx, and PM2.5	248

Table 6-18 MOVES Exhaust and Crankcase Ratios for Heavy-Duty Gasoline and CNG Vehicles by Pollutant,

Process, Model Year Group, and Fuel Type, and Source Type for PM2.5 Species	249

Table 7-1. N02/N0X ratios (± 95% Confidence Intervals, if available) from heavy-duty diesel vehicles

reported from recent studies	251

Table 7-2. Fleet-average and diesel specific HONO/NOx Ratios	252

Table 7-3 NOx and HONO Fractions for Heavy-Duty Diesel Vehicles	253

Table 7-4 NOx and HONO Fractions for Light-Duty (Source Type 21, 31, 32) and Heavy-Duty Gasoline

Vehicles (Source Type 41, 42, 43, 51, 52, 53, 54, 61)	253

Table 7-5 NOx and HONO Fractions CNG Heavy-Duty Vehicles	254

xii


-------
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, accounting for EPA
rulemakings. Heavy-duty vehicles in MOVES are defined as any vehicle with a Gross Vehicle Weight Rating
(GVWR) above 8,500 lbs.

In particular, we incorporated the Greenhouse Gas Emissions Standards for Heavy-Duty Vehicles - Phase 31
(HD GHG Phase 3) rule in MOVES5. Phase 3 sets greenhouse gas emissions standards for heavy-duty (Class
4 and above) vehicles that phase in beginning in model year 2027. We expect compliance with Phase 3 to
be achieved primarily through electrification of the HD fleet, even though the Phase 3 rule does not
mandate HD zero-emission vehicle (ZEV) sales. We document the impact of Phase 3 on HD ICE energy
consumption rates in Section 2.1.4.3. Because it is only a GHG rule, Phase 3 does not impact criteria
pollutant emissions. MOVES modeling of the Phase 3 rule is further documented in the GHG and Energy
Report,2 Emission Adjustments Report,3 and Population and Activity Report.4

In April 2024, EPA also finalized the Multi-Pollutant Emissions Standards for Model Years 2027 and Later
Light-Duty and Medium-Duty Vehicles (LMDV2027)5 standards. We also incorporated this rule into
MOVES5. The details of our modeling of this rule are available in the GHG and Energy Report, LD Exhaust
Report, and Emission Adjustments Report.

Heavy-duty vehicles in MOVES are defined as any vehicle with a Gross Vehicle Weight Rating (GVWR)
above 8,500 lbs. Emission rates for THC, CO, NOx, PM2.5 and NH3 are stored in the "EmissionRateByAge"
table in the MOVES database according to the following:

•	Pollutant

•	Emission process

•	Fuel type

•	Regulatory class

•	Model year group

•	Operating mode

•	Vehicle age

Energy emission rates are stored in the "EmissionRate" table, which is similar to the "EmissionRateByAge"
table, except emission rates are not differentiated by vehicle age. The MOVES framework and additional
details regarding the "EmissionRateByAge" and "EmissionRate" table are discussed in the report
documenting the rates for light-duty vehicles.12

In Section 1.1, we provide more background on the factors used to estimate heavy-duty exhaust emissions
in the "EmissionRateByAge" and "EmissionRate" tables. Sections 2 through 4 document the tailpipe
exhaust emission rates for heavy-duty diesel, heavy-duty gasoline, and heavy-duty compressed natural gas

1


-------
(CNG) vehicles. Section 5 documents the ammonia (NH3) running exhaust emission rates for heavy-duty
diesel, heavy-duty gasoline, and heavy-duty compressed natural gas fueled vehicles, including the updates
made to MOVES5. Section 6 documents the crankcase emission rates used for each fuel type of heavy-duty
vehicles. Section 7 documents the methods used to estimate nitric oxide (NO), nitrogen dioxide (N02), and
nitrous acid (HN02 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 particulate matter with
mean aerodynamic diameter less than 2.5 microns (PM25), ammonia (NH3) and energy consumption (in
units of kJ).

Total hydrocarbons (THC) is the measurement of hydrocarbons from a flame ionization detector.6 From
THC emissions, MOVES generates other estimates of hydrocarbon and organic gas missions, including non-
methane hydrocarbons (NMHC), non-methane organic gases (NMOG), 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 Speciation6 and
Toxics7 MOVES Reports.

MOVES reports PM25 emissions in terms of elemental carbon (EC) and the remaining non-elemental carbon
PM (nonECPM), as well as PM10 emissions. 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 using speciation profiles as documented in the Speciation Report.6

Ammonia emission rates are documented in Section 5. From NOx, MOVES estimates NO, N02, and HONO
emissions as documented in Section 7. Nitrous Oxide (N20) emissions from heavy-duty vehicles are
documented in the MOVES Greenhouse Gas and Energy Report. MOVES estimates C02 emissions from the
energy rates documented in this report, using conversion factors, which are also documented in the
Greenhouse Gas and Energy Report.

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:

The running exhaust process occurs as the vehicle is operating on the road either under load or in idle
mode, and also includes "off-network idle. The running process is delineated by 23 operating modes as
discussed in Section 1.6.

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 which is a reasonable approximation since cold
start emissions are typically much higher than hot-start emission.

g

Operationally — defined cold start emission rate (——=

Equation 1-2

= (emissions from drive cycle; with cold start

— emissions from drive cycle; with hot start and vehicle in warm condition)

The extended idle exhaust process in MOVES occurs during periods of hotelling, when long-haul trucks are
used during rest periods, such as when a vehicle is parked for the night and left idling. Extended idle in
generally defined to cover idling periods for longer than one hour. Extended idle can result in different
emissions than incidental idle that occurs during running operation because the engine may be operated at
a higher engine speed and the exhaust aftertreatment system may be too cool to operate at its full
efficiency.

Auxiliary power exhaust are emissions that come from diesel-powered generators that power the truck's
accessory loads, sometimes are used in place of the main engine during periods of hotelling. In MOVES,
extended idle and APU emissions also modeled for CNG and electric (EV) combination long-haul trucks, as
well as diesel. MOVES does not allow combination long-haul trucks to use the gasoline fuel type.
Documentation of the extended idle and auxiliary power exhaust emissions for heavy-duty diesel trucks
are in Sections 2.3 and 2.4 while documentation of CNG rates are in Section 4.3 and EV energy
consumption rates are documented in the Greenhouse Gas and Energy Consumption Rates Technical
Report.

MOVES also includes the capability to model energy consumption for trucks that are plugged into a
facility's electricity to run their accessories, known as "shore power." Shore power energy consumption
rates are documented in the Greenhouse Gas and Energy Consumption Technical Report.

Crankcase exhaust emissions (for running, start, and extended idle) include combustion products and oil
that are vented from the engine crankcase to the atmosphere. Crankcase emissions are estimated for THC,
CO, NOx, PM2.5 emissions and their chained pollutants as discussed in Section 6. We do not estimate energy
or C02 emissions because crankcase emissions are a small contribution to the total C02 emissions.
Historically, crankcase emissions have been significant sources of THC and PM2.5 emissions from heavy-duty
diesel engines. Crankcase emission rates for all four pollutants for all heavy-duty source types and fuels are
discussed in Section 6.

3


-------
Estimation of evaporative hydrocarbon emissions from heavy-duty gasoline vehicles is described in the
evaporative report.8 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.8 Brake and tire wear emission rates from heavy-duty vehicles are discussed in the
Brake and Tire Wear Report.9

Table 1-1 Emission Processes for Onroad Heavy-Duty Vehicles





Covered

processID



in this



processName

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



Refueling Displacement Vapor



18

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 source types, with the exception of long-haul combination trucks which
cannot be modeled as gasoline trucks. Note that, for most pollutants, the emissions from the heavy-duty
sector predominantly come from diesel vehicles and the majority of the data analyzed were from diesel
vehicles. MOVES also models E85 for light-duty vehicles, but this fuel type is not available to be modeled
for heavy-duty vehicles. And MOVES models energy consumption from battery electric and fuel cell electric
vehicles as explained in the GHG and Energy report.

1.4	Regulatory Class

The MOVES regulatory classes group vehicles that have similar emission rates. The MOVES heavy-duty
regulatory classes are largely determined based on gross vehicle weight rating (GVWR) classifications,
because the heavy-duty emission standards are based on GVWR as shown in Table 1-2.

There are additional criteria that define the heavy-duty regulatory classes in MOVES. Urban Bus vehicles
are distinguished from other heavy heavy-duty vehicles (GVWR >33,000 lbs.) because they have tighter PM
emission standards for the 1994 through 2006 model years.10 Urban bus is a regulatory class that is defined
by its intended use as well as the GVWR. EPA regulations define urban buses as "heavy heavy-duty diesel-
powered passenger-carrying vehicles with a load capacity of fifteen or more passengers and intended

4


-------
primarily for intra-city operation."11 In MOVES, we also assign all CNG transit buses and some electric
transit buses to the Urban Bus regulatory class.

In MOVES, gliders (regClassID 49) are defined as heavy heavy-duty diesel trucks with an old powertrain,
combined with a new chassis and cab assembly. Currently in MOVES, gliders are limited to the diesel long-
haul and short-haul combination truck source types. As discussed in Section 2.5, the emissions are
equivalent to MY 2000 HHD diesel vehicles.

We classify diesel light-heavy-duty Class 3 engine-certified vehicles in model year 2017 and later years as
LHD45 vehicles. The emission rates for LHD2b3 vehicles are based on the assumption that all vehicles are
chassis-certified and subject to the Tier 3 standards (discussed in Section 2.1.1.5.5), LMDV2027 standards
(see the LD Exhaust Report), and the Heavy-duty Phase 1 and 2 greenhouse gas emission standards
(discussed in Section 2.1.4.3.2). Because Class 3 engine-certified vehicles are instead subject to the same
emission standards as Class 4 and 5 engine-certified vehicles, we reclassified these vehicles as LHD45
vehicles. Model year 2017 is selected because this is the first model year the emission rates are different
between LHD2b3 and LHD45. The reclassification of diesel LHD2b3 as diesel LHD45 vehicles is documented
in the Population and Activity report.

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. As shown in Table 1-2, vehicles of a regulatory class may be mapped to multiple source types.
Likewise, each source type used to model heavy-duty vehicles includes population from several different
regulatory classes. For example, single unit short-haul trucks (sourceTypelD 52), include vehicles from the
following regulatory classes: LHD2b3, LHD45, MHD, and HHD.

5


-------
Table 1-2 Regulatory Classes for Heavy-Duty Vehicles

Regulatory Class
Description

regClassName

regClassID

Gross Vehicle
Weight Rating
(GVWR) [lb.]

Existing Source Types in
default database
(sourceTypelD)

Light Heavy-Duty
Class 2b and 3
trucks

LHD2b3

41

8,501-
14,000*

Passenger Trucks (31), and
Light-Commercial Trucks (32),
School Buses (43), and Single
Unit Trucks (51, 52, 53, 54)

Light Heavy-Duty
Class 4 and 5
Trucks

LHD45

42

14,001 -
19,500*

Buses (41, 42, 43) and Single
Unit Trucks (51, 52, 53, 54)

Medium Heavy-
Duty (Class 6 and
7 Trucks)

MHD

46

19,501-
33,000

Buses (41,42,43), Single Unit
Trucks (51, 52, 53, 54), and
Combination Trucks (61, 62)

Heavy Heavy-Duty
(Class 8 Trucks)

HHD

47

> 33,000

Buses (41, 42, 43), Single Unit
Trucks (51, 52, 53, 54), and
Combination Trucks (61, 62)

Urban Bus

Urban Bus11

48

> 33,000

Transit Bus (42)

Gliders (Class 8
Trucks)

Glider Vehicles

49

> 33,000

Combination Trucks (61, 62)

*Model year 2017-and-later engine-certified Class 3 (GVWR 10,001-14000 lbs) trucks (only present

within source types 52, 53, and 54) are classified as LHD45 (regclassID 42).

1.5	Model Year Groups

MOVES model year groupings are designed to represent major changes in emission rates due to changing
vehicle and aftertreatment technologies introduced in response to EPA emission standards. Model year
groups in MOVES can represent a single model year (e.g., 2007), or a range of model years (e.g., 2030-
2060). The emission rates discussed in the following sections are discussed in terms of model year ranges.
The model year groups cover all the model years between 1950 through 2060. When data are limited or
unavailable for model year groups, we make assumptions about the impact of emission standards and
vehicle aging to estimate those emission rates.

1.6	Operating Modes

Emission rates in MOVES are stored by regulatory class, fuel type, model year, and operating mode. To
calculate emissions from each process, MOVES sums the product of the emission rate for each operating
mode by the time spent in each operating mode.

For example, the activity basis for running process is source hours operating (SHO). The running process is
divided into 23 operating modes (as shown in Table 1-4). Using Equation 1-3, the total running emissions is
calculated by summing the product of the emission rates with the fraction of time spent in each operating
mode (the operating mode distribution). This is multiplied by the total hours (SHO) spent in this emission
process.

6


-------
Running Emissions

23

Z	Equation 1-3

(Operating Mode Distributiont x Emission Ratet)

i=1

MOVES performs the calculations shown in Equation 1-3, at a detailed level that accounts for each factor
that impacts the emission rates (e.g., model year, vehicle age, fuel type, regulatory class) and the operating
mode distribution (source type, roadtype, average speed (which varies across hour of the day)). Then, the
emissions can be aggregated to different levels (e.g., by sourcetype). Similar equations can be constructed
for other process, the equation for starts is shown in Equation 1-4, where the starts are classified into eight
operating modes (Table 1-5).

Start Emissions

= Starts

8	Equation 1-4

x ^ (Operating Mode Distributioni x Emission Ratet)

i=1

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,12 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.

VSp Avt + Bvt + Cvt + m ¦ Vt(at + g ¦ sin6t)	Equation !-S

f	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],
ot= instantaneous vehicle acceleration [m/s2]
g = the acceleration due to gravity [9.8 m/s2]
sin 0t = the (fractional) road grade at time t

7


-------
For heavy-duty vehicles, we classify running exhaust using Scaled Tractive Power (STP) as shown in
Equation 1-6 using road-load coefficients. STP is equivalent to VSP, except the power for all vehicles within
the same regulatory class and model year are scaled using a fixed mass factor, rather than the individual
weight of the vehicle. The fscaie is used to bring the numerical range of tractive power from heavy-duty
vehicles into the same numerical range as the VSP values when assigning operating modes. When
developing emission rates for MOVES, operating modes are assigned to individual vehicles using both the
individual truck mass, m, and the common fscaie value used for all heavy-duty vehicles from the same
regulatory class, source type and model year group. Because a common fscaievalue is used, individual
vehicles assigned to the same STP-defined operating mode bin are producing the same absolute tractive
power, regardless of differences in their individual source masses.

Avt + Bvf + Cv? + m ¦ vt(at + a ¦ sindt)	_ . .

SfPt =	-	-	-	tv r	_	—	Equation 1-6

fscaie

Where:

STPt = t\r\e scaled tractive power at time t [scaled kW or skW]

fscaie = fixed mass factor (see Table 1-3)

Other variables as previously defined in Equation 1-5

When MOVES estimates STP and assigns operating mode distributions for the heavy-duty source types,
Equation 1-6 uses the average source type mass (m) for each regulatory class, source type, and model year
group in the numerator and uses the common fscaie value for the regulatory class and model year group
which was also used in the emission rate analysis. At County and Default National Scale, MOVES uses the
instantaneous speed (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.

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.

8


-------
For onroad tests, measuring power from the ECU is generally more accurate than estimating power from
road-load coefficients. Unlike a generic road-load equation where vehicle characteristics, such as
aerodynamic drag and rolling resistance are assumed, the ECU measures engine speed and calculates
torque directly during the test, avoiding the need to capture the impact of wind speed and wind direction,
as well as weight and the road grade throughout the entire test cycle and route. Wind can have a
significant impact on power needs, and the payload of heavy-duty vehicles can be greater than the vehicle
weights itself, while also varying significantly over the test. Thus, for onroad tests, we generally use power
calculated from the ECU measurements, because the vehicle and environmental characteristics determine
the axle power (Section 2.1.1.3).

The use of STP instead of VSP is preferable for modeling heavy-duty vehicles emissions because heavy-duty
vehicle emissions are strongly correlated with power output. Heavy-duty vehicles are regulated using
engine dynamometer tests, and emissions standards are in units of grams per brake-horsepower-hour
(g/bhp-hr). Additionally, each heavy-duty regulatory class contains a wide variety of truck sizes, truck
weight, power ratings, and in-use payloads. Using STP, we can scale the heavy-duty emission rates to
different power outputs that were not measured in our emission rate database. The sample of trucks we
used to develop emission rates for each regulatory class has a limited number of trucks and loads
compared to the in-use fleet, which may not be representative of the average vehicle weight and power
output of the in-use fleet. The use of VSP would require the sample of vehicles to match the average
vehicle weights and load to accurately estimate average in-use emission rates. By using STP, MOVES scales
the measured emissions to match the estimated weight and power output of the modeled in-use fleet.

Thefscaie 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/sco;e (17.1 metric tons) was used, which provided
emissions within each operating mode bin for the largest heavy-duty truck vehicles.

However, in our analysis of in-use data from recent trucks, we found that an fSCaie of 17.1 metric tons
limited most of the real-world activity to low and medium power bins within a speed-bin, especially for the
light-heavy and medium-heavy-duty regulatory classes. In MOVES3, we revised the/sco/e values for MY 2010
and later to provide more resolution in the/sco/e by regulatory class. Derivation of the new/sco/e values is
described in Appendix G.

9


-------
Table 1-3 Fixed Mass Factor, fscaie (metric tons)

Regulatory Class (regClassID)

MY 1950-2009

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/Sca/e 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/Sca/e 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/Sca/e in the denominator, as is the case for all heavy-duty
vehicles, the unit is scaled kW and there is no normalization to a physical quantity.

The operating modes bins for running exhaust are shown in Table 1-4.

10


-------
Table 1-4 Operating Mode Definition for Running Exhaust for Heavy-Duty Vehicles

OpModelD

Operating Mode
Description

Scaled Tractive Power
(STPt, skW)

Vehicle Speed
(i/t, mph)

Vehicle Acceleration
including grade
(mph/sec)1

0

Deceleration/Braking2





ot +gsin(dt) 13 -2.0 OR
[ot +g sin(dt) < -1.0
AND

Ot-1 +g sin(Ot-i) < -1.0
AND

Ot-2 +g sin(dt-2) < -1-0)

1

Idle



vt < 1.0



11

Coast

STPt< 0

113 vt < 25



12

Cruise/Acceleration

0 13 STPt< 3

113 vt < 25



13

Cruise/Acceleration

3 13 STPt< 6

113 vt < 25



14

Cruise/Acceleration

6 13 STPt< 9

113 vt < 25



15

Cruise/Acceleration

9 13 STPt< 12

113 vt < 25



16

Cruise/Acceleration

12 13 STPt

113 vt < 25



21

Coast

STPt< 0

25 13 vt < 50



22

Cruise/Acceleration

0 13 STPt< 3

25 13 vt < 50



23

Cruise/Acceleration

3 13 STPt< 6

25 13 vt < 50



24

Cruise/Acceleration

6 13 STPt< 9

25 13 vt < 50



25

Cruise/Acceleration

9 13 STPt< 12

25 13 vt < 50



27

Cruise/Acceleration

12 13 STPt< 18

25 13 vt < 50



28

Cruise/Acceleration

18 13 STPt< 24

25 13 vt < 50



29

Cruise/Acceleration

24 13 STPt< 30

25 13 vt < 50



30

Cruise/Acceleration

30 13 STPt

25 13 vt < 50



33

Cruise/Acceleration

STPt< 6

50 13 vt



35

Cruise/Acceleration

6 13 STPt< 12

50 13 vt



37

Cruise/Acceleration

12 13 STPt<18

50 13 vt



38

Cruise/Acceleration

18 13 STPt< 24

50 13 vt



39

Cruise/Acceleration

24 13 STPt< 30

50 13 vt



40

Cruise/Acceleration

30 13 STPt

50 13 vt



1 The units o

vehicle acceleration for determining the braking mode are in units of mph/sec. MOVES

converts the acceleration in meters/sec2 to mph/sec using 0.4470 meter*hours = 1 mile*second.
2 The deceleration/braking definition will overlap with some of the other operating modes. In these
cases, the deceleration/braking categorization takes precedence over other definitions.

Start emission rates are also distinguished according to operating modes in MOVES, as shown in Table 1-5.
MOVES uses eight operating modes to classify starts according to the amount of time a vehicle was parked
prior to vehicle start (soak time). These range from a hot start (opMode 101) where the vehicle has been
soaking for less than 6 minutes, to a cold start (opMode 108) where the vehicle has been soaking for more
than 12 hours.

11


-------
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

ageGroupID

Lower bound
(years)

Upper
bound
(years)

3

0

3

405

4

5

607

6

7

809

8

9

1014

10

14

1519

15

19

2099

20

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

12


-------
currently does not allow explicit modeling of tampered vehicles - we hope to address this in future
versions of MOVES as more data on tampering become available.

Start emissions are generally a small contributor to total exhaust emissions from heavy-duty diesel
vehicles. No T&M effects are currently applied to the diesel start emissions.

For gasoline heavy-duty vehicles, the age effects are estimated directly from the a sample of emissions
data, or are adopted from light-duty deterioration as discussed in Section 3.1.1.1. These effects are applied
to both running and start exhaust emissions.

Not all emission rates are distinguished by vehicle age. Rates with no age dependence in MOVES are
stored in the "EmissionRate" table. This table includes energy consumption rates for all processes as well
as THC, CO, NOx, PM2.5, and ammonia (NH3) emission rates for extended idle and auxiliary power units
(APU), nitrous oxide (N20) rates for start and running emissions, and tire and brake wear emission rates.
We calculate T&M effects for the diesel extended idle emission rates without regard to vehicle age as
discussed in Section 2.3.2.3. This report documents the THC, CO, NOx, and PM2.5 emissions from extended
idle and APU usage, and heavy-duty ammonia emissions for all emission processes. However, heavy-duty
nitrous oxide and tire and brake wear9 emission rates are documented in separate MOVES reports.

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 Decree13, 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.

13


-------
Table 2-1 Model Year Groups for N0X Analysis Based on Emissions Standards

Model vear group

Standard (g/bhp-
hrl

NTE limit (g/bho-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

2010-2026

0.2

2027+

0.035 for LHD,
0.05 for
MHD/HHD

The off-cycle NTE standards was replaced by a
moving average window (MAW) approach in the
HD2027 rule.

Notes:

1	NMHC+NOx Standard

2	Assumes Phase-in of NOx standard

2.1.1.1 Data Sources

In modeling NOx emissions from HHD, MHD, LHD, and urban buses, we relied on the following data
sources:

ROVER. This dataset includes measurements collected during onroad operation using the ROVER system, a
portable emissions measurement system (PEMS) developed by the EPA. The measurements were
conducted by the U.S. Army Aberdeen Test Center on behalf of U.S. EPA.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.15 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).16,17,18 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

14


-------
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.19 It requires
each manufacturer of heavy-duty highway diesel engines to assess the in-use exhaust emissions from their
engines using onboard, portable emissions measurement systems (PEMS) during typical operation while on
the road. The PEMS unit must meet the requirements of 40 CFR 1065 subpart J. The in-use testing program
began with a mandatory two-year pilot program for gaseous emissions in calendar years 2005 and 2006.
The fully enforceable program began in calendar year 2007 and is ongoing. The vehicles selected for
participation in the program are within the engine's useful life, and generally, five unique vehicles are
selected for a given engine family. This dataset includes results for HHD, MHD, and LHD vehicles. The
HDIUT data are publicly available on EPA's website.20

The data available for use in MOVES2014 were collected during calendar years 2005 through 2010 and
represent engines manufactured in model years 2003 to 2009. For MOVES3 and later versions, we
evaluated data from engines selected for testing in calendar years 2010 through 2018. These engines cover
model years 2010 to 2016. The MY 2010+ data set included 40 unique engine families and 372 vehicles.
There are about 10 million seconds of quality-assured second-by-second data covering about 85,000 miles
of instrumented travel. The operational conditions include a wide range of driving speeds, transient and
steady-state conditions, engine loads, and exhaust temperature conditions that have implications for
emissions control efficacy, particularly for NOx.21 For the HHD class, out of a total 159 vehicles selected for
testing during 2010-2016, 109 were line-haul, 46 were delivery, and the remaining were marked as "Other"
in the metadata. Since the HDIUT data is measured and submitted by the manufacturer and the test
vehicles are required to be free of any tampering or mal-maintenance, we can safely assume that they
represent zero-mile vehicles for the purpose of assigning base rates and applying the tampering and mal-
maintenance effects. We have expanded the characterization of the MY 2010+ HDIUT data set by
separating the 0.2 NOx FEL group for heavy-duty diesel vehicles into two model year groups (2010-2013
and 2014 and later) as described in 2.1.1.5.

Houston Drayage Data. In coordination with the Texas Commission on Environmental Quality, the
Houston-Galveston Area Council, and the Port of Houston Authority, EPA conducted a study collecting
emissions data from trucks in drayage service using portable emission measurement systems (PEMS) from
December 2009 to March 2010.22 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.

15


-------
Table 2-2 Numbers of Diesel Vehicles from the ROVER, WVU MEMS, HDIUT, and Houston Drayage Programs by Model

Year Group

Data Source

MYG

Regulatory Class

HHD

MHD

LHD1

BUS

ROVER and
Consent Decree
Testing

1991-1997

19

-

-

2

1998

12

-

-

-

1999-2002

78

30

-

25

2003-2006

91

32

-

19

HDIUT

2003-2006

40

25

15

-

2007-2009

68

71

24

-

2010-2016

194

74

94

10

Houston Drayage

1991-1997

8

-

-

-

1998

1

-

-

-

1999-2002

10

-

-

-

2003-2006

8

-

-

-

1 LHD45 and LHD2b3 vehicles were grouped together and classified
as LHD vehicles for analysis of rates. The only difference in the
LHD45 and LHD2b3 emission rates is the application of production
volume weighting and T&M effects summarized in Table 2-16.

For the pre-2010 emission rates, we used a combination of the ROVER, Consent Decree Testing and the
HDIUT testing to estimate the emission rates. The Houston Drayage Study was used as comparison study
only. Additional details are provided in Section 2.1.1.4. The HDIUT data are used exclusively for the 2010-
2026 model year emission rates as discussed in Section 2.1.1.5. These 2010-2026 rates were also used as
the "base rates" to develop the MY2027+ emission rates based on the duty-cycle standard as explained in
Section 2.1.1.6.1.

2.1.1.2 Measurement Accuracy and Quality Assurance

PEMS devices continue to make improvements that affect measurement accuracy, such as sensor
response, sample conditioning, and noise reduction. The data sets represent the accuracy of the
instruments at the time of measurement. In compliance determinations, when determining whether the
tested vehicle meets the in-use emissions standard or not, an "accuracy margin for portable in-use
equipment" (commonly referred to as measurement allowance) is added onto the standard increasing the
vehicle compliance margin. The accuracy margins vary by model year and type of measurement method
and are described in 40 CFR 86.1912. This is done to prevent measurements that are biased-high from
affecting the compliance decision. However, since the true value for each second of data is unknown and
errors could be biased either high or low, the in-use emission rates used in MOVES from each of these data
sets are not adjusted to reflect the measurement allowance.

16


-------
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. C02 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 C02(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 C02 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 C02 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 ®eng and torque reng from the ECU were used to determine
engine power Peng, as shown in Equation 2-1.

Peng = <»eng*eng	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

17


-------
(to run electrical accessories), and air compressor. We identified where the accessories were
predominately used on a vehicle speed versus vehicle load map to properly allocate the loads. For
example, the cooling fan will be on at low vehicle speed where the forced vehicle cooling is low and at high
vehicle loads where the engine requires additional cooling. The air compressor is used mostly during
braking operations; therefore, it will have minimal load requirements at high vehicle speeds. Table 2-3
identifies the predominant accessory use within each of the vehicle speed and engine load map areas.

At this point, we translated the vehicle speed and engine load map into engine power levels. The engine
power levels were aggregated into low (green), medium (yellow) and high (red) as identified in Table 2-3.
Low power means the lowest third, medium is the middle third, and high is the highest third of the
engine's rated power. For example, for an engine rated at 450 hp, the low power category would include
operation between 0 and 150 hp, medium between 150 and 300 hp, and high between 300 and 450 hp. So,
for example, when vehicle operation is in middle of the engine load map and vehicle speed is low or mid
speed, the engine power level is in medium (yellow) band and the active accessory loads are as listed in the
respective cells. However, for the same engine load map operation (mid) with vehicle speed at high, the
engine power level is high (red) and active accessory loads are as listed in the cell. Some accessory loads,
such as cooling fan, are absent from cells with the same engine power level (identified by color) based on
the reasons given in the previous paragraph.

Table 2-3 Accessory Use as a Function of Speed and Load Ranges, Coded by Power Level





Vehicle Speed (mph)





Low (0-25)

Mid (25-50)

High (50+)



Low

Cooling Fan
Air Cond.
Engine Access.
Alternator
Air Compress

Air Cond.
Engine Access.
Alternator
Air Compress

Air Cond.
Engine Access.
Alternator

Q_

ro

"O

ro
O

	i


-------
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 P/0Ss,acc listed
below in Table 2-4 are subtracted from the engine power determined from Equation 2-1 to get net engine
power available at the engine flywheel.

For pre-2010 model years, LHD losses were set to zero. The losses for MY 2010+ LHD vehicles were
estimated by adjusting the MHD vehicle losses as such: (1) removed the loss for air compressor; (2) no
change to air condition loss; (3) scaled the losses for cooling fan, alternator, and engine accessories by 5/7
(where 5 and 7 are rough estimates of the average engine displacement in liters for LHD and MHD engines,
respectively). Based on these adjustments, the LHD losses are estimated to be approximately 60 percent of
MHD losses for the low power band and 70 percent for the mid and high power bands. We acknowledge
this calculation relies on a number of assumptions, but we believe it is a step forward from having LHD
losses equal to zero for all model years (as was the case in MOVES2014).

Table 2-4 Estimates of Accessory Load in kW by Engine Power Level

Engine Power Level
(of rated power)

HHD

MHD

LHD1

(Dre-2010)

LHD1
(2010+)

Urban
Bus

Low (0 - 1/3rd)

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

1 In MOVES2014, the accessory load losses for LHD were assumed to be
zero for all model years. In MOVES3 and later models, MY 2010+ LHD
data (Table 2-2) is analyzed with non-zero accessory load losses.
However, for pre-2010 MY LHD, we continue to assume zero accessory
load loss.

We then adjusted for the driveline efficiency, accounting for efficiency losses in the wheel bearings,
differential, driveshaft, and transmission. The efficiency values were determined through literature
searches. Driveline efficiency rjdnveUne varies with engine speed, vehicle speed, and vehicle power
requirements. Using sources available in the literature, 24-25-26-27-2&-29-30-31-32 We estimated an average value
for driveline efficiency. Table 2-5 summarizes our findings.

19


-------
Table 2-5 Drivel in e 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 Pox/e. MOVES uses the Pi0SSACC from Table
2-4 for each regulatory class and engine power-level (high, medium, or low).

Paxte ~ Vdriveline (Peng ~ Ploss,acc)	Equation 2-2

Finally, we scaled the axle power using Equation 2-3, and the STP-scaling factors/sco;e presented in Table
1-3 for every second of data.

p

STP = axle	Equation 2-3

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 1950-2009 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 Houston
Drayage Program data.33,34 When we extended calculations to 40 ages, we duplicated the emission rates
for MY 1960 for model years back to 1950.

20


-------
In MOVES2014, the emission rates were considered for an update using the HDIUT or Houston Drayage
Program data if:

•	MOVES2010 rates were not based on actual data, and

•	The comparison to the new independent data shows that more than a half of MOVES2010
emission rates are outside the boundary of the 95 percent confidence intervals of the
independent data.

Using this methodology, the emission rates for HHD MY 2003-2006 vehicles and LHD2b3 and LHD45 MY
2003-2006 vehicles were updated using the MY 2003-2006 HDIUT data.A The Houston Drayage Program
was not used to estimate any of the emission rates in MOVES.

Table 2-16 outlines the data sets used to estimate emission rates for each model year group and regulatory
class. ROVER and Consent Decree testing data was used to estimate the following regulatory class and
model year groups combinations: HHD 1991-1997, 1998, 1999-2002, 2003-2006, MHD 1999-2002, 2003-
2006B, and Urban Bus 1991-1997, 1999-2002, 2003-2006. The HDIUT was used to estimate emissions for
HHD 2007-2009 and LHD 2003-2006.

Emissions in each data set were reported in grams per second. To calculate MOVES heavy-duty exhaust
emission rates, we first averaged all the 1-Hz NOx emissions by vehicle and operating mode because we did
not believe the amount of driving done by each truck was necessarily representative. Then, the emission
rates were again averaged by regulatory class and model year group. For trucks MY 2009 and older, these
data sets were assumed to be representative and each vehicle received the same weighting. Equation 2-4
summarizes how we calculated the mean emission rate for each stratification group (i.e., model year
group, regulatory class, and operating mode bin).

rij = the number of 1-Hz data points (for a given operating mode bin) for each vehicle j,
fiveh = the total number of vehicles,

rpjj = the emission rate of pollutant p for vehicle j at second /',

fp = 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.

A This analysis is described in more detail in the MOVES2014 heavy-duty exhaust report. However, since we have
updated the/sCa/e values used to assign STP bins, the MOVES3 rates are no longer directly comparable.33
B For 2003-2006 MHD, the emission rates are different than the 2003-2006 HHD emissions for operating modes 0,1,
11, 21, and 33. For the other operating modes, the emission rates are equivalent to the HHD emission rates.

nveh

Equation 2-4

Where:

21


-------
The data included in the emissions analysis does not cover all operating modes or regulatory classes and
model year combinations needed for MOVES. In the following sections, we discuss the methods used to fill
missing operating mode bins and missing regulatory class and model year combinations. In addition, we
also estimate the impact of low- NOx rebuilds which were not included in the sampled vehicles. To do so,
we rely on the heavy-duty diesel emission standards, as well as engineering knowledge and test data of
emission control technologies that were implemented or forecasted to be implemented to meet the
standards.

2.1.1.4.1	Light Heavy-Duty Class 2b3 and Classes 4&S Trucks

As described in Section 1.5, the LHD regulatory classes were redefined for MOVES3, and the/SCate value for
2009-and-earlier LHD2b3 and LHD45 regulatory classes is now 2.06 metric tons. c This differs from
MOVES2014 where the value for these vehicles was 17.1 metric tons. Thus, it was not possible to carry
over the emission rates from MOVES2014. Instead, in MOVES3 and later versions, the 2009-and-earlier
LHD2b3 and LHD45 regulatory classes were assigned the same emission rates as the LHD<10K in
MOVES2014.

2.1.1.4.2	High-Power Operating Modes

As described in Section 1.6,/sco;e values for MHD and HHD trucks were not changed for model years 2009-
and-earlier. Thus, for MHD and HHD trucks, the maximum operating mode (opModelD = 40) represents a
tractive power greater than 513 kW (STP= 30 skW x 17.1). This value exceeds the capacity of most HHD
vehicles; MHD vehicles and buses exert even less power. Nearly all of the HHD activity occurs in modes 0,
1, 11-16, 21-28, and 33-38, with activity for buses and MHD vehicles usually occurring over an even smaller
range.

To estimate emission rates in the modes beyond the ranges of available data, for each model year group
we linearly extrapolated the rates from the highest operating mode in each speed range where significant
data were collected. In most cases, this mode was mode 16 for the lowest speed range, mode 27 or 28 for
the middle speed range, and mode 37 or 38 for the highest speed range. For each of these operating
modes, work-specific emissions factors (g/kW-hr) were calculated using the midpoint STP (Table 1-4).

Then, these emissions factors were multiplied by the midpoint STP of the higher operating modes (e.g.,
modes 39 and 40 for speed>50 mph) to input emission rates for the modes lacking data. For the highest
bins in each speed range, a "midpoint" STP of 33 skW (564.3 kW) was used. Equation 2-5 displays an
example calculation of the emission rate for opModelD 40, using a mean emission rate from opModelD 37,
for a given regulatory class and model year group.

Emission RateopModeID 40 = Emission RateopModeID 37 x ( „T°VM°de'D )	Equation 2-5

\->l ropModelD 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 extrapolated 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

- This is consistent with the 2.06 fSCaie used to develop LHD2b3 and LHD45 emission rates in MOVES2010.

22


-------
more of the high speed, high power emission rates were extrapolated, which explains the strictly linear
trend in the emission rates between operating mode 27 -30 and 37-40.

4000

3500

_3000

-C

0,2500 -
+¦»
re

i—

x

92000

01

s 1500

1000

500	*

*

' 1 *

	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1

0 1 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

23


-------
6000 -|

5000

4000

ttfl
Qi

(V

g 3000

1999-2002

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 these 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 2-16 provides a summary of the assumptions used to estimate emission rates for regulatory class-
model year groups with missing data.

24


-------
2,1,1,4.4 Defeat Device and Low-NO x 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.13 Reflashes
reduce NOx emissions from these engines by reconfiguring certain engine calibrations, such as fuel
injection timing.

Since defeat devices were in effect mostly during highway or steady cruising operation, we assumed that
NOx emissions were elevated for only the top two speed ranges in the running exhaust operating modes
(>25mph). To modify the relevant emission rates to represent reflash programs, we first used emission
rates from model year 1999 (the first model year with not-to-exceed emission limits, see Table 1-2) to
calculate baseline ratios of the emission rates for operating modes 27 and 37 to the rate for opMode 16 (as
discussed in the beginning of Section 2.1.1.4). We then multiplied the MY 1999 ratios by the emission rates
in operating mode 16 for model years 1991 through 1998, to get estimated "reflashed" emission rates for
operating modes 27 and 37. This step is described in Equation 2-7 and Equation 2-8. To estimate
"reflashed" rates in the remaining operating modes, we multiplied the reflashed rates by ratios of the
remaining operating modes to mode 27 for MY 1991-1998, as shown in Equation 2-9 and Equation 2-10.
The second step preserves the relationship among operating mode 21 and operating modes 21-30, and
among operating mode 37 and operating modes 31-40, that existed in the original 1991-1998 data.

_ _	(h999,n\	Equation 2-7

rreflash;91-98,27 = r91-98,16 I ~ 1 )

\r1999,16/

Operating modes 21-30

(f91-98,omx\	Equation 2-8

^"reflash,91-98,OMx ^"reflash,91-98,27 I -	I

\ '91-98,27 /

rreflash,91-98,37 ~ r91-98,16 I -

\'1999,16

Operating modes 31-40

(?i999,37\	Equation 2-9

/j"Vi-98.0MA-\ Equation 2-10
rreflash,MY1991—1998,0Mx ~ rreflash;91-98,37 I -

'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 2008.36 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

25


-------
been reflashed. The reflashed HHD 1991-1998 MY emission rates were also applied to the 1991-1998 MHD
diesel emission rates. Note that there are no tampering and mal-maintenance aging effects in MOVES for
pre-2010 NOx emission rates. Thus, the only change in the NOx emission rates by vehicle age for these
model years is due to reflashing.



2540



2520



2500



2480







2460

X

2440

6
z

2420



2400



2380



2360



2340











•o-o.. \
•o-o.. \

•0"G.\















> • • • %























—•—s

.lodel Year 1991-1993







...O- f

flodel Year 1994-1997







—•—J

flodel Year 1998





































0

10	15	20

Age (years)

25

30

Figure 2-3. HHD NOx emission rates for OpModelD 37 for Model Years 1991-1998 Adjusted for Low-NOx Rebuilds by

Vehicle Age

2.1.1.4.5	2007-2009 Model Year HHD, MHD, and Urban Bus

The 2007 Heavy-Duty Rule108 required the use of ultra-low sulfur diesel fuel; this fuel enabled diesel
engines with diesel particulate filters to reach the 0.01 g/bhp-hr PM standard beginning in 2007. In
addition, the 2007 Heavy-Duty Rule108 established much tighter NOx emission standards (0.2 g/bhp-hr).
While the NOx standard went into effect for MY 2007 at 0.2 g/bhp-hr, it was phased in over a three-year
period ending in MY 2010. Rather than phasing in the aftertreatment technology needed to meet the new
standard, most manufacturers chose to meet a 1.2 g/bhp-hr 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.34

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
standard in vehicles such as the Dodge Ram. This technology allows for the storage of NOx during fuel-lean
operation and conversion of stored NOx into N2 and H20 during brief periods of fuel-rich operation. In

26


-------
addition, to meet particulate standards in MY 2007 and later, heavy-duty vehicles are equipped with diesel
particulate filters (DPF). The DPF must be regenerated at regular intervals to remove and combust
accumulated PM to relieve backpressure and ensure proper engine operation. This step requires high
exhaust temperatures. However, these conditions adversely affect the LNT's NOx storage ability, resulting
in elevated NOx emissions.

In order to determine the fraction of time that DPFs spend in PM regeneration mode, in 2007, EPA
acquired a truck equipped with a LNT and a DPF and performed local onroad measurements using portable
instrumentation and chassis dynamometer tests. We distinguished regimes of PM regeneration from
normal operation based on operating characteristics, such as exhaust temperature, air-fuel ratio, and ECU
signals. During the testing conducted onroad with onboard emission measurement and on the chassis
dynamometer, we observed a PM regeneration frequency of approximately 10 percent of the operating
time.

Emissions from this vehicle were not directly used to calculate MOVES emission rates, because only one
vehicle was tested. Rather, to estimate average emissions of LHD2b3 vehicles with LNT, we calculated a
ratio to the MOVES20100 MY 2003-2006 NOx emission rates. During DPF regeneration, we assumed that
the LNT did not reduce emissions from 2003-2006 levels. During all other times, we assumed that
emissions were reduced by the percent reduction in the certification standards from 2003-2006 levels (i.e.,
90 percent). These assumptions result in an estimated NOx reduction of 81 percent for LNT equipped
trucks from 2003-2006 to 2007-2009, as shown in Equation 2-11.

LNT LHD2b3 NOx emissions
Baseline Emissions

(LNT normal emissions

= (normal op. frequency) x

Baseline Emissions J	Equation 2-11

(Baseline Emissions\

+ (DPF reg. frequency) x 	-	

\Baseline EmissionsJ

= (0.90) X (0.10) + (0.10) X (1) = 0.19

Where Baseline Emissions = MOVES2010 MY 2003-2006 NOx emission rates for LHD2b3

We weighted the average rates for the LHD2b3 regulatory class (regClassID 41) for model years 2007-2009
assuming that LNT-equipped trucks account for 25 percent of the LHD2b3 diesel vehicles. We assume that
the remaining 75 percent of MY 2007-2009 LHD2b3 diesel trucks will not have LNT aftertreatment and will
exhibit a 50% NOx reduction from the 2003-2006 model year emission rates as was assumed for the HHD
and MHD described in the previous section. Overall, these assumptions result in a 58 percent reduction in
NOx emission rates in the model year 2007-2009 emission rates from the MOVES2010 MY 2003-2006 NOx
emission rates as shown in Equation 2-12.

D In MOVES2014, we updated the diesel NOx emission rates for 2003-2006 based on the HDIUT program.

27


-------
2007 — 2009 LHD2b3 NOx emissions
Baseline emissions

(LNT NOx emissions\

= (LNT market share) —	;			

\Baseline Emissions J
+ (non	Equation 2-12

/2007 — 2009 emission standards\

— LNT market share) 	-	

V	Baseline Emissions	J

= (0.25) x (0.19) + (0.75) x (0.5) = 0.4225
Where Baseline Emissions = MOVES2010 MY 2003-2006 NOx emission rates for LHD2b3

In the absence of other data, we applied the LHD2b3 emission rates to LHD45 vehicles. Newer data shows
that LNT is not being used in LHD45 vehicles, however, we have not updated this assumption because the
LHD results compare well to the HDIUT data.33

2.1.1.5 2010-2026 Model Years

In MOVES3 and later versions, the MY 2010-2026 emission rates for HHD, MHD, Urban Bus, and LHD45,
and LHD2b3 are based on analysis of the HDIUT data and model-year specific production volume weighting
for each model year from 2010 through 2018. The rates for HHD, MHD, and the two LHD classes use data
from vehicles with HHD, MHD, and LHD engines, respectively. The NOx emission rates are projected to
remain constant for MY 2018 and later vehicles for regulatory classes HHD, MHD, Urban Buses, and LHD45.
The LHD2b3 trucks are projected to have a further decrease in NOx emissions through the implementation
of the Tier 3 program as discussed in Section 2.1.1.5.5.

In calculating the mean emission rates for MY 2010-2026 vehicles, we made several additions to the
method presented for the pre-2010 model years described in Section 2.1.1.4:

1.	For a given regulatory class (HHD, MHD, LHD), we grouped the vehicles in the HDIUT data set into
three NOx family emission limit (FEL)e groups as detailed in Section 2.1.1.5.1 below.

2.	Within the NOx FEL group and regulatory classes, we grouped vehicles into model year groups as
detailed in Section 2.1.1.5.2.

3.	We calculated the operating mode-based average emission rate for each vehicle by regulatory
class, NOx FEL group, and model year group (Equation 2-13). Then, we calculated the operating
mode-based average emission rate for all vehicles in the NOx FEL group and model year group
(Equation 2-14).

4.	We weighted the operating mode-based average emission rates for each of the NOx FEL groups,
and model year group within each regulatory class by the model year specific production volumes
of the engines in the NOx FEL group to the total production volume of the regulatory class
(Equation 2-15). Thus, we created operating mode-based average emission rates for each model
year and regulatory class.

E A Family Emission Limit is the maximum emission level established by a manufacturer for the certification of an
engine family.

28


-------
ERpol,OM,C,FEL,MYG,veh ~

Equation 2-13

ERpol,OM,C,FEL,MYG ~

S^Ccount

^Iveh ERpolfOM,C,FEL,MYG,veh
V6hrmirit

count

Equation 2-14

ERpol,OM,C,MYG,MY ~

',C,MYG,MY ~

pol,OM,C,FEL,MYG *

Equation 2-15

Where:

C = Regulatory class (LHD, MHD, HHD, and Urban Bus)

ERx,y,z = Emission rate in mass/time. The subscripts show the categorization.
FEL = NOx FEL group of engine (0.20 g/bhp-hr, 0.35 g/bhp-hr, and 0.50 g/bhp-hr)
MYG = Model year group (2010-2013, 2014-2016)

MY = Model year

OM = Running exhaust emissions operating mode
pol = Pollutant (NOx, THC, CO)

PVc,my,fel	= Production volume by class, model year, and FEL group

sec; secCOUnt = a second of data (for a given i/eh and OM); number of seconds in that

category

veh; vehCOUnt = 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 intervals from the different FEL groups
were weighted together using the production volumes, similar to how the means were calculated.

29


-------
0.10 -

0.08 -

< 0.06 -
3B

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 NO* 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.



¦ HDIUT, MY 2013

—





—























i

























X

1









i 1

1









T





i i

1











11











.ill





30


-------
NOx FEL(g/bhp-hr)

Figure 2-5 Distribution ofNOx 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.

31


-------
Table 2-7 Number of 2010 and Later Model Year HDILJT Vehicles by N0X FEL Group and Model Year Groups1 Used for

Emission Rate Analysis



NOx FEL Group

Total

02

0.35

05

MY 2010-
2013

MY 2014-
2016

MY 2010-
2016

MY 2010-
2016

LHD2

52

27

0

15

94

MHD

19

23

23

9

74

HHD3

78

50

31

35

194

Total

149

100

54

59

362

^HC, CO, and C02 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	LHD data for the 0.2 FEL group includes MY up to MY 2016. MHD and HHD data
for the 0.2 FEL group only include up to MY 2015

3	The HHD 0.35 group contains five MY 2009 vehicles, which have similar
NOx emission rates as the other vehicles in this group.

The average NOx emission rates for the 0.2 NOx FEL Group for model year 2010-2013, compared to the
0.35 and 0.5 NOx FEL Groups are shown below for LHD (Figure 2-6), MHD (Figure 2-7) and HHD (Figure 2-8).
Because the 0.35 and 0.5 NOx FEL groups do not have a model year distinction like the 0.2 NOx FEL group,
the figures display all the available MY 2010 and later vehicles for the 0.35 and 0.5 NOx FEL groups.

As shown, the NOx emission rates in the 0.2 NOx FEL Group and model year 2010-2013 are consistently
lower than those in the 0.35 and 0.50 NOx FEL groups, with a few minor exceptions. In addition, for many
operating mode and regulatory class combinations, the NOx emission rates in the 0.2 NOx FEL Group are
statistically significantly lower than the NOx emission rates in the 0.35 and 0.5 NOx FEL Groups. However,
no such trend exists in the NOx emission rates for the 0.35 and the 0.5 NOx FEL groups. In general, the NOx
emission rates in the 0.35 and the 0.5 NOx FEL groups are not statistically different than one another by
operating mode for the MHD and HHD regulatory classes.

32


-------
Figure 2-6 Average LHD N0X Emission Rates by Operating Mode for the 0.2 N0X FELfor MY 2010-2013 and the 0.5 N0X
FELfor MY 2010-2015. Error Bars are 95% Confidence Intervals of the Mean

li MHD FEL 0.20, MY 2010-2013!, N=23

¦	MHD FEL 0.35, MY 2010-2013(, N=23

¦	MHD FEL 0.50, MY 2010-2013- N=9

.1.1

li

il

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES OpMode

Figure 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

33


-------

-------
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 aii pollutants, and to account for the potential
differences in real-world NOx emissions performance due to the updated engine and aftertreatment
systems.

MY 2010-2013
MY 2014-2016





L

li I.

Il

Jl

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES OpMode

Figure 2-9 NOx emission rates for the MY 2010-2013 and MY 2014-2016 vehicles in the LHD 0.20 NOx FEL Group

0.030
0.025

^ 0.020

cu
+-•
ru

^ 0.015
c
o
—

o.oio

E

LLI

0.005
0.000

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES OpMode

Figure 2-10 NOx emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the MHD 0.20 NOx FEL Group

MY 2010-2013
MY 2014-2015

.1 il

1

i

35


-------
0.08

0.07

~ 0.06

™ 0.05
x
O

2 0.04
0.03
0.02
0.01
0.00

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES OpMode

Figure 2-11 NO NOx emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the HHD 0.20 NOx FEL Group
2.1.1.5.3 Weighting by Production Volume

We collected production volume data by the same regulatory classes (LHD, MHD, HHD, Urban Bus) and
NOx FEL groups (0.20, 0.35, 0.50) that we used for the emission rate analysis. We then combined the NOx
FEL group-based rates (averaged within the model year groups described above) with the production
volume data (distinct for each model year) to create emission rates unique to each model year. We did this
for each model year from 2010 through 2018 (the last model year for which we have production volume
data). For MY 2019-2026, we used the same production volume weighting as MY 2018. The per-model-
year production volume weighting, by regulatory class and NOx FEL groups, is shown in Figure 2-12. This
method allows us to better represent the technology adoption landscape in the heavy-duty domain. For
example, for HHD, model years 2010 through 2013 had engines with NOx FEL in the 0.50 group (0.35
g/bhp-hr < NOx FEL < 0.50 g/bhp-hr), but starting with model year 2014, there are no engines in the 0.50
group. Compared to engines in the 0.20 group and 0.35 group, engines in the 0.50 group predominantly
use a different emissions control strategy to reduce tailpipe NOx emissions. Thus, our approach using the
NOx FEL groups and per-model-year production volume correctly captures the prevalence and influence
(on emissions) of different technologies in the fleet. Production volume percent contributions of the three
NOx FEL groups sum to 100 percent of the production volume for each regulatory class and model year.



¦ MY 2010-2013













¦ MY 2014-2015



































.









































tT





I,

I

















ii







ii











.. Ii il







ii 1









ii II II

















36


-------
Figure 2-12 Production Volume Contribution of Heavy-Duty Diesel Engine Families by N0X Family Emission Limit Group

for Each Regulatory Class

2.1.1.5.4 Operating Modes and fSCaie Values

For the updates to THC, CO, N0X, PM2.5, and energy rates for MY 2010-2026 HD vehicles, we used new/sco/e
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/sco/e values on OpMode coverage using the example
of NOx emission rates for the vehicles in the NOx FEL=0.20 group for LHD, MHD, and HHD regulatory
classes, respectively/ Note that the absolute mass/time OpMode-based emissions rates between the two
series based on different fscaie cannot be compared. The main benefit of the new/sco/e values is that all the
operating modes are populated even if the trends may not be perfectly monotonically increasing for each
pollutant in a regulatory class. The comparison is similar for the 0.35 and 0.50 NOx FEL groups. Note the
final rates input into MOVES are estimated as production volume weighted rates from each of the NOx FEL
groups.

F These plots were generated using model year 2010-2015 HDIUT data that was available at the time of the analysis.
The final emission rates were developed using the complete HDIUT data that includes both MY 2010-2015 and MY
2016-2018.

37


-------
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 -

¦J2



U)



'

0.030 -

O



z





0.020 -



0.010 -



0.000 -

m MHDD_0.20_fs=7.00
BMHDD 0.20 fs=17.1

B IB I

1 11 12 13 14 15 16

ie_

ill

21 22 23 24 25 27 28 29 30
MOVES OpMode

M.I

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

38


-------
M

x

O

0.080
0.070
0.060 -|
0.050
0.040 -
0.030 -
0.020
0.010 H

0.000

li

¦ H H D DO. 20_f s=10.00
BHHDD 0.20 fs=17.1

II

0|1 11 12 13 14 15 16 121 22 23 24 25 27 28 29 30 ¦ 33 35 37 38 39 40

MOVES OpMode

Figure 2-15 Effect of MOVES2014 (17.1) and MOVES3 (10.00) 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 HD Incorporation of Tier 3 Standards for LHD2b3

In addition to regulating light-duty vehicles, the Tier 3 vehicle emission standard37 also applies to chassis-
certified light heavy-duty diesel vehicles, which includes all of regclass 41 (LHD2b3). In addition, we
updated the T&M assumptions in MOVES4 for regclass 41 by reflecting the 50,000 miles warranty
requirement for LHD vehicles, and the 150,000 useful life mileage for Tier 3 LHD2b3 vehicles (see Section
B.l). For these LHD2b3 diesel vehicles, we estimate reductions in the NOx zero-mile emission rates
attributable to the Tier 3 standards beginning in 2018. We did not estimate Tier 3 reductions in the THC,
CO, and PM2.5 emission rates as discussed in subsequent sections.

For diesel vehicles in the LHD2b3 regulatory class, we estimate that the Tier 3 NOx standard results in a
different percent reduction of start and running emissions. We applied a greater portion of the reduction
to running emissions and a smaller reduction to start emissions. These reductions were phased-in over the
same schedule as for gasoline vehicles, as detailed in Table 2-8. The derivation of the phase-in assumptions
is discussed in the MOVES2014 heavy-duty exhaust report.34

39


-------
Table 2-8 Phase-in Assumptions for Tier 3 N0X Standards for Light Heavy-Duty 2b3 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

1 These reductions are based on comparison of Tier 3 standards against Tier 2
standards.

In generating the reduced rates for running operation, the starting point (or pre-Tier 3 baseline) are the
LHD2b3 rates for MY2017. The ending point, representing full Tier 3 control, was model year 2022. The MY
2022 rates were calculated by multiplying the rates for MY2017 by 0.3855. This fraction reflects the
percent reduction in running emission rates for MY2022 as shown in Table 2-8.

In addition to tightening emission standards, the Tier 3 regulations require an increase in the regulatory
useful life from 120,000 miles to 150,000 miles. We used the Tampering and Mal-maintenance (T&M)
methodology presented in Appendix B to estimate the emissions impact of the lengthened useful life, and
to estimate different age effects for Tier 3 LHD2b3 vehicles as shown in Table B-3. For the phase-in model
years, we calculated a weighted average of the Model Year 2017 and Model Year 2022 emission rates using
the Tier 3 phase-in, as shown in Equation 2-16.

LHD2b3 ERmy = (LHD2b3 ER20,7 ) X (1 - Tier 3 PhaseinMy)	Equation 2-16

+ (LHD2b3 ER2022) x (Tier 3 PhaseinMy)

Where:

LHD2b3 ERmy = the LHD2b3 diesel emission rate for each process (start/running), age,
operating mode, and model year between 2018 and 2021.

Equation 2-16 is also used to estimate the impact of the lengthened Tier 3 useful life standard on NOx,
PM2.5, THC, and CO emissions. Tier 3 also allows fleet-wide averaging with electric vehicles as explained in
the MOVES adjustments report. While this averaging is allowed for LHD2b3s, it is not modelled in MOVES.
The reduction in NOx grams per mile from the Tier 3 rulemaking across model years is displayed in Figure
2-25.

2.1.1.6 2027-2060 Model Years

In MOVES4, we updated the MY 2027 and later emission rates for the LHD45, MHD, HHD and Urban Bus
regulatory classes to account for the impact from the HD2027 standards. The HD2027 standards include
duty-cycle standards, off-cycle standards and changes to warranty and useful life requirements. This
section describes the methodology to derive the new zero-mile NOx emission rates for running emissions.

40


-------
Temperature adjustments for running and extended idle N0X emissions are described in the MOVES
adjustments report.

We first estimated the effects of the duty-cycle standards and the off-cycle standards separately, as
discussed in Sections 2.1.1.6.1 and 2.1.1.6.2, respectively. Then, we estimated the combined effect of both
the duty-cycle standards and off-cycle standards on the zero-mile emission rates used in MOVES as
discussed in 2.1.1.6.3. The effects of aging are described in Section 2.1.1.7.

2.1.1.6.1 Emission Rates Based on Duty-Cycle Standards

The HD2027 NOx heavy-duty compression ignition duty-cycle exhaust emission standards for MY2027+ are
shown in Table 2-9. The duty-cycle standards include three separate tests: Federal Test Procedure (FTP),
Supplemental Emission Test - Ramped Modal Cycle (SET-RMC) and low load cycle (LLC). We used the 2027+
standards for the FTP and (SET-RMC) to estimate the effect of the duty-cycle standards on MOVES NOx
emission rates. Because we do not have sufficient (LLC) test data on existing heavy-duty diesel vehicles to
develop the modeling inputs specific for the LLC standard in MOVES, we used the FTP standard to model
the impact of the standards on low-power operation.

Table 2-9 Heavy-duty Compression Ignition Duty-Cycle NOx Standards for the HD2027 Rule

Regulatory
Classes

FTP

(g/hp-hr)

SET-RMC
(g/hp-hr)

LLC (g/hp-hr)

LHD

0.035

0.035

0.05

MHD

0.05

0.05

0.065

HHD

0.05

0.05

0.065

_ HD2027 FTP AND SET RMC STANDARD	EQUATION 2-17

Rduty ~	MY2010 STANDARD

Equation 2-17 through Equation 2-22 were used to incorporate the effects of more stringent FTP and SET-
RMC engine duty-cycle emission standards on MOVES running exhaust NOx emission rates for model years
subject to the standards. The 2010 0.2 g/hp-hr NOx emission standard38 is the most recent previous heavy-
duty NOx emission standard. The term Rduty is the ratio between the new emission standards and the MY
2010 FTP and SET-RMC duty-cycle standards (0.2 g/hp-hr).

HD2027 FTP and SET RMC standard	c , , 7

d _		Equation 2-17

duty	MY2010 standard

Rduty ranges between 17 and 25 percent as shown in Table 2-10.

Table 2-10 Rduty Ratios Calculated for the HD2027 Standards

Regulatory Classes

FTP and SET standard
(g/hp-hr)

Rduty

LHD

0.035

17.5%

MHD, HHD

0.05

25%

41


-------
To estimate the effect of the new engine dynamometer duty-cycle standards on in-use emissions, we
looked at this relationship for the previous change in NOx standards. To evaluate the in-use effectiveness of
the 2010 standard, we compared the in-use NOx emission rates from vehicles that were certified to the
pre-2010 heavy-duty NOx standard with the 2010 standard. Equation 2-18 defines Rin use as the ratio
between the percent change observed in-use from vehicles compliant with the 2010 NOx standard relative
to vehicles compliant with the pre-2010 standard, and the percent change in the 2010 standard FTP
standards relative to the pre-2010 standard.3 In other words, Rin use is the ratio between the relative
change in in-use emissions compared to the relative change in the FTP duty-cycle emission standard. This
ratio is operating mode specific and the ratio computed for HHD is applied to MHD and LHD45.

% Change in the in-use emission rates from 2010 compliant vehicles	„

R. =	_	_		Equation 2-18

in-use	% Change in the 2010 FTP standard

The percent change in in-use emission rates from vehicles certified to the 2010 standard (the numerator in
Equation 2-18) was estimated using Equation 2-19. For each operating mode, the MOVES emission rates
for HHD vehicles certified to the 2010 0.2 g/hp-hr standard (the numerator) were calculated from 93 MY
2010-2015 HHD vehicles with a certified engine family emission limit (FEL) below the 0.2 g/hp-hr NOx
emissions level and tested as part of the Heavy-Duty In-Use Testing program.20 The MOVES emission rates
for HHD vehicles certified to the 2004-2006 standard (the denominator) are based on 91 MY 2003-2006
trucks from two in-use datasets: ROVER data collected by the US EPA and the Heavy-Duty Diesel Consent
Decree data collected by West Virginia University. Because data was sparse for LHD and MHD, the HHD
data ratios were also used for LHD and MHD.

% Change in the in-use emission rates from 2010 compliant vehicles

Emission rate from HHD 2010 compliant vehicles	Equation 2-19

HHD MY 2006 MOVES emission rate

The percent change from Equation 2-19 was calculated separately for each MOVES operating mode to
evaluate the effectiveness of the 2010 standard across different ranges of in-use operating conditions, as
shown in Figure 2-16 and Table 2-11. The emission reduction is larger for the higher speed and higher load
MOVES operating modes, with the largest decrease observed for speeds above 50 mph (operating modes
33 through 35). The lowest effectiveness of the standards is observed for low speed and several low power
operating modes (operating mode 1, 11, and 21), with an exception of the deceleration bin (operating
mode 0).

Equation 2-20 was used to estimate the percent reduction between the 2010 standard and the 2004-2006
emission standard. This term is also the denominator of Equation 2-18. The 2004-2006 NOx emission
standard was estimated as 1.68 g/hp-hr, assuming that NOx emissions consist of 70 percent of the

G In 2004-2006, the NMHC+NOx emission standard was 2.4 g/hp-hr; the 0.2 g/hp-hr NOx standard began to be phased-
in starting in 2007, with a full-phase in 2010.

42


-------
combined NMHC (Non-Methane Hydrocarbons) plus NOx standard, consistent with the assumption used in
MOVES339. The percent change associated with the standard change is plotted as a horizontal line in Figure
2-16 to compare to the in-use emission rate reductions.

0.2

Percent Change in the 2010 FTP standard = — . — 1 = —88.1%	Equation 2-20

i/O 70 X iLi • x! )

0

-10 -
-20 -
-30 -
-40
-50 -
-60 -
-70
*80 -
-90 -
-100

Speed<25

¦ I ¦ ¦ ¦

I I ¦ i

D//° Change in the NOx FTP standard

In-use emission reductions from 0.2 bhp-
hr certified vehicles

2550

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES Operating Mode

Figure 2-16 Percent change in in-use emission rates for 2010 standard (0.2 g/hp-hr) compliant HHD vehicles,
compared to the percent change in the 2010 duty-cycle standard across MOVES operating modes

Table 2-11 displays the Rin use as calculated by Equation 2-18. For operating modes with Rin use values
greater than one, the observed in-use emissions are less than would be expected due to the change in FTP
emission standard. Operating modes with Rin_use values less than one see less impact than implied by the
change in the FTP emissions standards.

43


-------
Table 2-11 Calculation of Ri„ use by MOVES Operation Mode

MOVES
OpMode

HHD MOVES
MY 2006 NOx
emission rates
(g/hr)

NOx Emission rate
from 2010 compliant
HHD vehicles (g/hr)a

Percent change in in-use
NOx emission rates from
2010 compliant vehicles
(%)

Rin_use c

0

0.038

0.0031

-91.8

1.04

1

0.015

0.0063

-57.9

0.66

11

0.015

0.0106

-28.8

0.33

12

0.058

0.0210

-63.5

0.72

13

0.093

0.0339

-63.7

0.72

14

0.127

0.0453

-64.4

0.73

15

0.145

0.0565

-60.9

0.69

16

0.188

0.0734

-60.9

0.69

21

0.010

0.0066

-32.1

0.36

22

0.064

0.0163

-74.4

0.84

23

0.093

0.0243

-73.9

0.84

24

0.132

0.0359

-72.7

0.83

25

0.165

0.0485

-70.6

0.80

27

0.225

0.0633

-71.8

0.82

28

0.244

0.0558

-77.2

0.88

29

0.314





0.88b

30

0.384





0.88

33

0.051

0.0062

-87.8

1.00

35

0.148

0.0130

-91.2

1.04

37

0.226

0.0323

-85.7

0.97

38

0.268

0.0306

-88.6

1.01

39

0.345





1.01

40

0.422





1.01

aThe HHD rates in this table are based on fscaie of 17.1 metric tons to be consistent with the
fscale of the MY 2006 HHD emission rates in MOVES3 and later versions. Note that the fscaie for
model year 2010 and later is 10 for HHD, 7 for MHD, and 5 for LHD45 and LHD2b3.

b For operating modes lacking data, we used the same Rm use for the closest operating mode.
c Rin_use is calculated based HHD data, but also applied to LHD45 and MHD.

Equation 2-21 is used to estimate the percentage reduction to NOx running emissions from the change in
the duty-cycle standard for each operating mode.

Rdutyjn.use = 0 ~ Rduty) X Rin.use	E^Uatio" 2'21

Where:

Rduty in use= the percent emission reductions in the in-use running NOx emissions
estimated from changing the FTP duty-cycle standard.

44


-------
Equation 2-22 is used to estimate the age 0-3 heavy-duty diesel NOx running emission rates from the
changes in the duty-cycle standards. The same calculations were applied to estimate rates for all the
heavy-duty diesel regulatory classes.H

ERduty_in_use — (l Rduty_in_use) x ER.MOVES_baseline	Equation 2 22

Where:

ERduty in use= the MOVES running NOx emission rates for the control scenarios based on
reduction in the duty-cycle standard

Rduty_in_use= the percent emission reductions in running NOx emissions estimated from
changing to FTP duty-cycle standard

ERMOVES_baseiine= the MOVES baseline running NOx emission rates for each regulatory
class

The estimated HHD MOVES running NOx emission rates for the HD2027 standards, estimated based solely
on the duty-cycle standards in Table 2-9, are shown in Figure 2-17.

H We assumed that the Rm use values calculated by MOVES operating mode can be applied to the MOVES rates that
are derived using a different fscaie- The change in fscaie does not change the definition of operating modes that are not
defined by Scaled Tractive Power, STP (deceleration operating mode 0, and idle operating mode 1), or operating
modes with negative STP values (operating mode 11 and 21), defined in Table 1-4. Changing the fscaie values does
change the definition of vehicle operation in the other operating modes. However, the Rin_use values are relatively
constant for the positive power operating modes within each speed range as observed in Table 2-11. In calculating
the results shown in Figure 2-16, we deemed it was not necessary to attempt to account for the fscaie differences
when applying the Rin_use values.

45


-------
300

,250

| 200

£2 150

8

100

50

I Without HD2027
HD2027 Duty-Cycle Standard Compliant Rates for MY2027+

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES Operating Mode

Figure 2-17 Duty-cycle-based running NOx emissions, ERduty_in_use, for HHD dieselfor the HD2027standards

2.1.1.6.2 Emission Rates Based on Off-cycle Standards

In this section, we document the methods used to estimate MOVES NOx running emission rates based on
the HD2027 off-cycle standards for heavy-duty diesel vehicles. Table 2-12 presents the calculated off-cycle
standards used to develop MOVES inputs. The off-cycle standards include requirements for operating
conditions in two bins: idle bin (less than 6 percent of maximum power) and non-idle load bin. In
developing inputs to MOVES, we assumed manufacturers will comply with the optional idle standard in all
off-cycle idle operation. We then developed the off-cycle standards using the procedures as described
below. These calculations are detailed here for HHD, but we used the same approach for MHD and LHD45.

In modeling the HD2027 off-cycle standards in MOVES, we incorporated the impact of the compliance
margin EPA used in setting the off-cycle standards. The in-use compliance margin of approximately 40% of
the standards, accounted for a number of in-use factors, including test procedure variability, production
and engine variability, fuel and diesel exhaust fluid (DEF) quality, aftertreatment aging due to severe-
service operation, and aftertreatment packaging to name a few. When manufactures design engines, they
must include similar margin to ensure each of their engines meet the standards, which results in the
average fleet emissions being below the standards.

46


-------
Table 2-12 Calculated Off-Cycle N0X Standards based on the HD2027 Standards

Regulatory Class

Off-cycle
Bin

Off-Cycle NOx Standards (g/hr for idling, g/hp-
hr for non-idle operation)

LHD

Idle

9 for MY2027-2028; 13.5 for MY2029+A

Non-idle

0.052 for MY2027-2028, 0.051 for MY2029+B

MHD

Idle

9 for MY2027-2028; 13.5 for MY2029+A

Non-idle

0.065 for MY2027-2028; 0.064 for MY2029+B

HHD

Idle

9 for MY2027-2028; 13.5 for MY2029+A

Non-idle

0.065 for MY2027-2028; 0.064 for MY2029+B

A The optional idle standards were 10 g/hr in the HD2027 final standards. However,

MOVES uses the values in the table to specifically account for the compliance
margin to the off-cycle standards.

B The non-idle standards in the table are different from the HD2027 final standards
values (0.058 g/hp-hr and 0.073 g/hp-hr for LHD and MHD/HHD/Urban bus,
respectively). They are now based on the modeling of off-cycle standards
compliance margin in MOVES4.

In order to apply these off-cycle standards to MOVES operating modes, we first converted the standards to
a fuel consumption (i.e., C02 emissions) basis. We calculated the optional idle NOx g/hr standard in units of
NOx g/C02 kg using Equation 2-23 and the resulting values are displayed in Table 2-13.

NOv	/R\

Optional Idle	standard -—

2	Vkg/	Equation 2-23

= ^Optional Idle N0X standard (j^-)Jy/|ldle C02

Where Idle C02 (j~j= the MOVES average C02 (kg/hr) emission rate for HHD diesel vehicles for MOVES
idle (operating mode 1). We assume the C02 rates are unaffected by the HD2027 standard.

Table 2-13 Calculation of Optional Idle N0x/C02 Standard (g/kg)

Optional Idle NOx standard
(g/hr)

Average HHD Idle (Operating
Mode=l) C02 emission rate (kg/hr)

Optional Idle NOx/C02
standard (g/kg)

9 for MY2027-2028; 13.5 for
MY2029+

7.68

1.17 for MY2027-2028;
1.13 for MY2029+

Next, we converted the reference off-cycle NOx standards into units of gram per hour (g/hr) for each
MOVES operating mode. We refer to g/hr rates as the off-cycle standard compliant emission rates, which
are shown in Table 2-14 in Columns (F) for HHD vehicles for the final standards.

In Table 2-14, Column (B) lists the MOVES C02 emission rate for model year 2027 HHD diesel vehicles for
each MOVES operating mode. Column (C) lists the mean power for each operating mode bin as calculated
from the Heavy-Duty In-Use Testing data, which is the same data set that was used to derive the MOVES

47


-------
C02 emission rates (see Section 2.1.4.3).20 The percent load, Column (D), is calculated for each operating
mode bin using Equation 2-24.

Mean Power 0pMode=i

Percent Load0pMode=i= 77	5	 P	Equation 2-24

p	Mean PowerOpMode=40

Where:

i= one of the 23 MOVES running exhaust operating modes from 0 to 40

Mean Power0pMode= the mean power for each of the MOVES operating modes shown

in Column C.

Mean PowerOVMode=A0= assumed maximum power bin = 470 hp for HHD diesel vehicles.

We then assigned each MOVES operating mode into a power classification (Column E) based on the
percent load (Column D) as defined in the rule, where percent load less than 6 percent of maximum power
is defined as idle and above 6 percent of maximum power is non-idle load.

48


-------
Table 2-14 Calculation of the Off-cycle NO* Standard Compliant Emission Rate for HHD Diesel Vehicles for the HD2027

Standards

A

B

C

D

E

F

G

MOVES

operating

mode

MOVES
MY 2027
HHD C02
emission
rate
(kg/hr)

Mean
power
(hp)

Percent load

Power classification

MY 2027-
2028 off-
cycle
compliant
emission
rate (g/hr)

MY 2029+
off-cycle
compliant
emission
rate (g/hr)

0

4.92

6.04

1.3%

Idle

5.76

5.57

1

7.68

8.10

1.7%

Idle

9.00

8.70

11

13.42

1.04

0.2%

Idle

15.72

15.19

12

21.69

28.90

6.2%

Non-idle

1.88

1.85

13

37.31

75.16

16.1%

Non-idle

4.89

4.81

14

52.20

121.18

26.0%

Non-idle

7.88

7.76

15

68.68

166.98

35.8%

Non-idle

10.85

10.69

16

110.42

282.24

60.5%

Non-idle

18.35

18.06

21

13.92

-1.61

-0.3%

Idle

16.31

15.76

22

32.99

34.43

7.4%

Non-idle

2.24

2.20

23

44.71

77.71

16.7%

Non-idle

5.05

4.97

24

59.82

121.62

26.1%

Non-idle

7.91

7.78

25

77.03

167.82

36.0%

Non-idle

10.91

10.74

27

102.53

230.56

49.4%

Non-idle

14.99

14.76

28

142.09

327.41

70.2%

Non-idle

21.28

20.95

29

181.90

403.76

86.5%

Non-idle

26.24

25.84

30

212.63

470.01

100.7%

Non-idle

30.55

30.08

33

28.36

34.76

7.4%

Non-idle

2.26

2.22

35

71.87

145.03

31.1%

Non-idle

9.43

9.28

37

106.93

227.23

48.7%

Non-idle

14.77

14.54

38

148.35

323.23

69.3%

Non-idle

21.01

20.69

39

183.17

396.00

84.9%

Non-idle

25.74

25.34

40

196.42

466.62

100.0%

Non-idle

30.33

29.86

Finally, the off-cycle N0X standard compliant emission rate in Column (F) was calculated based on the
power classification and the stringency of the off-cycle standard. For operating modes classified as idle, we
multiplied the MOVES C02 emission rate in Column (B) by the NOx/CC>2 off-cycle idle standard calculated in
Table 2-13 using Equation 2-25.

49


-------
Idle Emission Rate

= MOVES MY 2027 HHD C02 Emission Rate

NOx	/

x Voluntary Idle	inusestandard -—

C02	Vkg.

Equation 2-25

For the operating modes classified as non-idle load, we multiplied the off-cycle (g/hp-hr) standard in Table
2-12 for the corresponding power classification by the mean power (Column C), as shown in Equation 2-26.

Non — idle Emission Rate = Mean Power(hp) x In_use standard

Equation 2-26

Vhp ¦ hiV

The estimated off-cycle NOx standard compliant emission rates for heavy heavy-duty diesel vehicles are
shown in Figure 2-18. Similarly, we applied Equation 2-24 through Equation 2-26 to estimate the off-cycle
standard compliant emission rates for the other MOVES regulatory classes using corresponding C02 rates
and the mean power for those vehicles.

3 300


E

LU

8

z
w

LU

>
O

200

150

100

50

0

I Without HD2027

HD2027 Off-Cycle Standard Compliant Rates for MY2027-2028
HD2027 Off-Cycle Standard Compliant Rates for MY2029+























































































































































































1.

1,











ll











1 ¦

1.

I,



0 1 11 12 1314151621 2223242527282930333537383940

MOVES Operating Mode
Figure 2-18 Base NOx rates and off-cycle NOx standard compliant emission rates for HHD diesel

2.1.1.6.3 Emission Rates Based on Combination of Duty-Cycle and Off-Cycle Standards
In this section, we document the methods used to develop MOVES NOx emission rates for heavy-duty
diesel vehicles that reflect the effects of both duty-cycle standards and off-cycle standards. As an example,
Figure 2-19 shows the HHD duty-cycle and off-cycle standards for MY 2027+ across MOVES operating
modes. The duty-cycle standard is estimated to constrain emissions in four operating modes (operating
modes 0, 21, 35 and 37), while the off-cycle standard constrains emissions in the remaining operating
modes.

50


-------
	Duty Cycle Compliant Rates for MY2027+

	Off-cycle Compliant Rates for MY2027-2028

	Off-cycle Compliant Rates for MY2029+

.c
3
0)
ro
o:

E

LU
X

O
z
w

LU
>
O

120
100
80
60
40
20
0





1-25 mph





25-50 mph



>50 mph













































































































































/





























0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES Operating Mode

Figure 2-19 Comparison of Running NOx emission rates for diesel-fueled HHD compliant with the duty-cycle and off-

cycle standards

Because manufacturers will need to simultaneously comply with both the duty-cycle and off-cycle
standards, we estimated the final MOVES NOx emission rate for each operating mode as the lower of the
two rates generated from the duty-cycle and the off-cycle standards (e.g., the emission rate based on the
off-cycle standards is selected for operating mode 12, but the emission rate based on the duty-cycle
standards is selected for operating mode 35). Figure 2-20, Figure 2-21 and Figure 2-22 present the
estimated emission rates for HHD, MHD, and LHD45 diesel vehicles that meet both the final duty-cycle and
off-cycle standards, respectively.

51


-------
300

-250

ra
QL

o 200

x 150

o

> 100

o

50
0

i Without HD2027	

HD2027 Standards Rates for MY2027-2028
HD2027 Standards Rates for MY2029+



1.11111

._L

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES Operating Mode

Figure 2-20 Estimated zero-mile NOx emission rates for HHD diesel vehicles due to the HD2027 duty-cycle and off-cycle

standards

120

sz

S100

aj
ra

O 80

60
40
20
0

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

lllll

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES Operating Mode

Figure 2-21 Estimated zero-mile NOx emission rates for MHD diesel vehicles due to the HD2027 duty-cycle and off-

cycle standards

52


-------
120

IE"

3100

a>
as
cr

o 80

w
w

'E

111

5 60

~Z.

to

£ 40

o

5

20
0

Figure 2-22 Estimated zero-mile N0X emission rates for LHD45 diesel vehicles due to the HD2027 duty-cycle and off-

cycle standards

2.1.1.7 Tampering and Mal-maintenance

MOVES accounts for the fleet-average increase in emissions with vehicle age. For heavy-duty diesel
vehicles, the increase with age is modeled in the development of the emission rates by applying tampering
& mal-maintenance (T&M) adjustment factors to age zero rates. We assume that the T&M effects are the
dominant cause of emission increases with age in heavy-duty diesel vehicles. No effort has been made to
explicitly model emission increases due to normal wear and tear (e.g., catalyst degradation, sensor
deterioration from properly maintained vehicles) from heavy-duty diesel vehicles.

Table 2-15 shows the percent increases due to T&M adjustment factors for aggregate NOx emissions by
regulatory class and model year group. As described in Appendix B, the T&M adjustment factors in Table
2-15 are calculated by combining information regarding the assumed frequency rate of an equipment
failure at the useful life of the engine with the estimated emission impact of the equipment failure.

¦	Without HD2027

¦	HD2027 Standards Rates for MY2027-2028
HD2027 Standards Rates for MY2029+



































































































































































































1













¦





















M	

1.

..1.

1.

1.

9_





i

I.

1.

1.

_





1

i

i.

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

MOVES Operating Mode

53


-------
Table 2-15 N0X T&M adjustment factor (percent) by regulatory class

Model vears

LHD2b3 trucks (%)

LHD45 trucks (%)

MHD. HHD. Urban Bus (%)









1994-1997

0

0

0

1998-2002

0

0

0

2003-2006

0

0

0

2007-2009

3.8

3.84

0

2010-2012

55.9

77.9

77.9

2013-2026

58.6

58.6

58.6

2027-2028

58.6

329.7

324.0 (MHD),
410.5 (HHD),
433.7 (Urban Bus)

2029-2060

58.6

331.6

325.6 (MHD),
416.3 (HHD),
440.6 (Urban Bus)

We also apply an age effect; that is, we assume that emissions begin to deteriorate at the age vehicles pass
the warranty requirements and increase to the full T&M adjustment factors at the age the vehicles reach
the useful life mileage requirement (Figure B-l). Warranty periods vary by model year and regulatory class
as shown in Table B-l and Table B-2.

The emission rates described in Sections 2.1.1.4 through Section 2.1.1.6 are assumed to represent new or
"zero-mile" emission rates. The 2007-2009 LHD emission rates are derived from the 2003-2006 HDIUT data
as discussed in Section 2.1.1.4.6. The MY 2010 and later heavy-duty emission rates are also based on the
HDIUT data which tested newer and well-maintained vehicles. We then account for the emission increases
with age by multiplying the MY 2007-2009 LHD emission rates, and MY-2010-and-later emission rates by a
function of the corresponding T&M adjustment factors and age effects (Equation 8-3). The MOVES
emission rates for regulatory classes with the same zero-mile emission rates (Table 2-16) are different due
to the T&M NOx effects (Table 2-15) and phase-in of T&M effects by age (Table B-4and Table B-5).

The T&M adjustments for NOx are zero for some of the model year groups because these vehicles lack the
heavy-duty advanced aftertreatment systems that we assume are most affected by tampering and mal-
maintenance, as discussed more in Appendix B.

The LHD vehicles have different T&M NOx increases than HHD, MHD, and Urban Bus vehicles due to the
penetration of lean NOx trap (LNT) aftertreatment. For MY 2007-2009 we assumed that there was a 25
percent penetration of LNT-equipped vehicles within both LHD2b3 and LHD45 regulatory classes, with the
remaining 75% having no NOx aftertreatment equipment. Subsequent certification data shows that LNT
was not actually used in LHD45 vehicles, however, we have not updated this assumption, or the T&M
adjustment factors, for MY 2007-2009 because the resulting LHD45 2007-2009 emission rates compare
well to the HDIUT data for 2007-2009 vehicles.33 For MY 2010-2012, we assumed a 25 percent penetration
of LNT and 75 percent of selective catalytic reduction (SCR) NOx aftertreatment system within the LHD2b3
regulatory class. We assume that all heavier heavy-duty vehicles (LHD45, MHD, HHD, Urban bus) have 100

54


-------
percent penetration of SCR systems starting in model year 2010. For model years 2027+, the NOx standard
is much more stringent and we assume that aftertreatment systems are much more effective in reducing
emissions. As detailed in Section B.7, we also assume the converse—that aftertreatment failure in a
MY2027 vehicle would bring NOx emissions to the same level as estimated for a MY 2010 vehicle's engine-
out emissions without any reductions from aftertreatment system.

The T&M values for model year 2010 and later vehicles also account for implementation of OBD. For
LHD2b3 trucks, OBD systems were assumed to be fully implemented in MY 2010. For the other HD
regulatory classes (LHD45, MHD, HHD), we assumed there would be a phase-in period from MY 2010 to
2012 where one-third of those trucks were equipped with OBD systems. In MY 2013 and later, all trucks
have OBD systems. These OBD adoption rates have been incorporated into the tampering and mal-
maintenance emission increases in Table 2-15 based on the assumptions and calculations detailed in
Appendix B.

Figure 2-23 displays the NOx rates in gram per mile by age and regulatory class for model year 2015
vehicles. Due to faster mileage accumulation, the HHD trucks tend to reach their useful life quicker than
other heavy-duty vehicles, with the maximum emission rates by the 4-5 age group. In contrast Urban
Buses do not reach the useful life until the age 10-14 group. The gram per mile (g/mile) emission rates
were calculated outside the model by using operating mode distributions and average speeds for each
regulatory class estimated from a national MOVES run. The emission rates showed here do not include any
adjustments to the rates due to fuel effects or humidity that are applied during MOVES run, however they
do incorporate differences due to activity.1

Despite the significant T&M adjustment factors for MY 2010+ heavy-duty diesel vehicles, the fully aged NOx
emissions of MY 2010+ vehicles are significantly lower than the emission rates of MY 2009 and earlier
vehicles. As shown in Figure 2-25, the age 0-3 MY 2009 NOx g/mile emission rates are over 100% higher
than the MY 2010 vehicle emission rates, while the T&M adjustment factors are less than 100% (Table
2-15). Thus, MOVES estimates that, on average, a fully-aged MY 2010 heavy-duty truck has lower NOx
emissions than a new MY 2009 heavy-duty truck.

1 For example, the MY 2015 NOx emission rates by operating mode for LHD2b3 and LHD45 diesel vehicles are the
same, but Figure 2-17 shows different gram per mile emission rates due to vehicle and activity differences, including
heavier weights of LHD45 vehicles (source mass), and because the two regulatory classes are distributed differently
among the source types in MOVES, which have different operating mode distributions. See the MOVES Population
and Activity Report, for more information. The zero-mile HHD and Urban Bus NOxemission rates by operating mode
are equivalent. However, the emission rates in the age groups: 0-3, 4-5, 6-7, and 8-9 are different because they have
a different phase-in of the T&M effects. The difference between HHD and Urban Bus in grams per mile for ages 10+
are due to differences in the operating mode distributions and average speeds.

55


-------
Reg Class

41-LHD2b3

¦	42-LHD45
46-MHD67

¦	47-HHD8
48-Urban Bus

10-14

15-19

8-9
Age (years)

Figure 2-23 Base Heavy-duty Diesel NOx Running Emission Rates (g/mile) by Age for Model Year 2015 by Regulatory
Class Estimated using Nationally Representative Operating Mode Distribution

2.1.1.8 Summary and Model Year Trends

Table 2-16 summarizes the methods used to estimate emission rates for each regulatory class and model
year group combination. The emission rates in MOVES are based on the analysis of ROVER, Consent Decree
testing data, and HDIUT data. The HHD, MHD, LHD, and Urban Bus rates for MY 2010+ vehicles are based
on HDIUT data.

Table 2-16 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)

1950-
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

56


-------
Model

year

group

HHD

(regClass 47)

MHD

(regClass 46)

Urban Bus
(regClass 48)

LHD45
(regClass 42)

LHD2b3
(regClass 41)

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'5

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'5 with
MY specific
production
volume
weighting

MHD data
analysis'5 with
MY specific
production
volume
weighting

Same as HHD
except T&M
adjustment
factors specific to
Urban Bus

LHD data analysis'5
with MY specific
production
volume weighting;
T&M specific to
LHD45

LHD data analysis'5
with MY specific
production volume
weighting; T&M
specific to LHD2b3
& Tier 3 reductions
starting in MY 2018

2019-
2026

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
reduction phase-in
from MY 2019-2022

2027-
2060

Revised rates
based on
HD2027
standards

Revised rates
based on
HD2027
standards

Revised rates
based on HD2027
standards

Revised rates
based on HD2027
standards

Same as 2019-2026
rates

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 2-24, which shows NOx emission rates for 0-3 age group by model year and regulatory class
estimated in grams per mile (g/mile) using nationally representative operating mode distributions and
average speeds. The MOVES model estimates drastic reductions in NOx emissions from pre-1990
technologies to modern 2027 and later model year engines. Some of the differences in the emission rates
for the 0-3 age group are caused by different phase-in of T&M adjustment factors, which began to be

57


-------
applied in the first age group for HHD, MHD, and Urban Bus regulatory classes (Section 2.1.1.5.5). Because
the nationally representative operating modes are different across model years and regulatory classes,
some of the differences between model years and the regulatory classes are due to activity differences and
not the emission rates by operating mode.1

The figure also shows NOx emissions for "gliders", which are trucks with a new chassis and cab assembly
and equipped with a rebuilt engine typically without an exhaust aftertreatment system. These emissions
are of similar magnitude to emission rates from the model year 2000 HHD vehicles as discussed in Section
2.5. In MOVES, zero population of gliders is estimated starting in MY2020.

Figure 2-25 shows the model year trend for NOx in gram per mile for the 2007 and later model years. The
reduction in LHD2b3 emission rates due to the phase-in of the Tier 3 standards between 2018 and 2022 is
evident. All other heavy-duty emission rates by operating mode are unchanged from model year 2018 to
2026 until the HD2027 rule introduces additional reductions for MY2027+ as documented in Table 2-16.
The slight reductions observed in the gram per mile emission rates for all heavy-duty vehicles (including
gliders) in model years 2014 through 2027 are due to changes in the operating mode distribution from
lower weights, lower rolling resistance, and improved aerodynamics of trucks implemented in the Phase 1
and Phase 2 Heavy-duty Greenhouse Gas Regulations as documented in the Population and Activity
Report.

Reg Class

41-LHD2b3

42-LHD45

46-MHD67

47-HHD8

48-Urban	Bus

49-Gliders

1980	2000	2020	2040

Model Year

Figure 2-24 Base running emission rates for NOxfrom age 0-3 diesel heavy-duty vehicles averaged over a nationally

representative operating mode distribution.

58


-------
Model Year

Figure 2-25 Base running emission rates for NOxfrom age 0-3 gasoline heavy-duty vehicles averaged over a nationally
representative operating mode distribution for model years 2007 and later.

2.1.2 Particulate Matter (PM2.5)

PM2.5 refers to particulate matter with a mean aerodynamic diameter less than 2.5 microns. Particulate
matter is a complex mixture of particles that are composed of one or many of the following substances:
organic material, metals, elements, and ions, including sulfate. MOVES PM25 emission rates are typically
based on filter measurements, 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 PM2.5
emissions. Other, secondary reasons include wider availability of PM2.5data in the HDIUT data set (section
1) for MY 2010-2011 and the/sco;e updates to all HD regClasses for MY 2010 and later (see section 2.1.1.4.2
and Appendix G). Table 2-17 shows the model year group ranges and the applicable brake-specific
emissions standards.

59


-------
Table 2-17 Model Year Groups Used for Analysis Based on the PM Emissions Standard

Model Year Group Range

PM Standard (g/bhp-hr)

1950-1987

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 2.1.2.1 and Section 2.1.2.2 discuss the derivation of the MY 1950-2009 and 2010-2060 heavy-duty
diesel PM2.5 emission rates, respectively. The tampering and mal-maintenance factors applied to emissions
in both model year groups are discussed in Section 2.1.2.3.

2,1,2,1 1950-2009 Model Years

The PM2.sdata were analyzed in several steps to obtain MY 1950-2009 PM2.5 emission rates. First, STP
operating mode bins were calculated from the chassis dynamometer data from the CRC E55/59 testing
program described below. Second, continuous PM2.5 data measured by the Tapered Element Oscillating
Microbalance (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.

When we extended calculations to 40 ages, we duplicated the emission rates for MY 1960 for model years
back to 1950.

2,1,2,1,1 Data Sources

All of the data used to develop the MOVES PM2.5 emission rates for MY 1960-2009 were generated based
on the CRC E-55/59 research program.40 The following description in the "Compilation of Diesel Emissions
Speciation Data - Final Report"41 provides a good summary of the program:

The objective of the CRC E55/59 test program was to improve the understanding of the California heavy-
duty vehicle emissions inventory by obtaining emissions from a representative vehicle fleet, and to include
unregulated emissions measured for a subset of the tested fleet. The sponsors of this project include CARB,
EPA, Engine Manufacturers Association, DOE/NREL, and SCAQMD. The project consisted of four segments,
designated as Phases 1, 1.5, 2, and 3. Seventy-five vehicles were recruited in total for the program, and
recruitment covered the model year range of 1974 through 2004. The number and types of vehicles tested
in each phase are as follows:

•	Phase 1: 25 heavy heavy-duty (HHD) diesel trucks

•	Phase 1.5: 13 HHD diesel trucks

•	Phase 2: 10 HHD diesel trucks, 7 medium heavy-duty (MHD) diesel trucks,

2 MHD gasoline trucks

•	Phase 3: 9 MHD diesel, 8 HHD diesel, and 2 MHD gasoline

60


-------
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
2-18.

61


-------
Table 2-18 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 2-19.

Table 2-19 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).42 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.

62


-------
Note that this approach differs from the NOx emission rates analysis described in Section 2.1.1.3, since the
particulate data was collected on a chassis dynamometer from vehicles lacking electronic control units
(ECU). We have not formally compared the results of the two methods of calculating STP. However, on
average, we did find the operating-mode distributions to be similar between the two calculation methods
for a given vehicle type. For example, we found that the maximum STP in each speed range was
approximately the same.

2,1,2,1,3 Compute Normalized TEOM Readings

The TEOM readings were obtained for a subset of tests in the E-55/59 test program. Only 29 vehicles had a
full complement of 1-Hz TEOM measurements. However, the continuous particulate values were modeled
for the remaining vehicles by West Virginia University, and results were provided to EPA. In the end, a total
of 56 vehicles and 470 tests were used in the analysis out of a possible 75 vehicles. Vehicles and tests were
excluded if the total TEOM PM2 5 reading was negative or zero, or if corresponding full-cycle filter masses
were not available. Table 2-20 provides vehicle and test counts by vehicle class and model year. The MHD
(Class 6 and Class 7) trucks tested in the study included seven Class 6 and eleven Class 7 vehicles,
representing only a limited model years.

Table 2-20 Vehicle and Test Counts by Heavy-Duty Class and Model Year

Model Year

HDD Class 6/7(MHD)

HDD Class 8 (HHD)

No. Vehicles

No. Tests

No. Vehicles

No. Tests

1969

-

-

1

6

1974

1

10

-

-

1975

-

-

2

10

1978

-

-

1

5

1982

1

5

-

-

1983

1

10

1

6

1985

1

28

1

10

1986

1

3

1

4

1989



11

1

4

1990

1

12

1

3

1992

1

11

1

11

1993

1

11

1

3

1994

1

9

3

15

1995

2

24

3

13

1998

2

20

3

28

1999

-

-

3

43

2000

2

18

5

44

2001

1

5

2

21

2004

-

-

4

29

2005

-

-

1

6

Since the development of MOVES emission rates is cycle independent, all available cycles/tests which met
the above requirements were utilized. As a result, 488,881 seconds of TEOM data were used. The process
required that each individual second-by-second TEOM rate be normalized to its corresponding full-cycle
filter mass, available for each combination of vehicle and test. This step was necessary because individual

63


-------
TEOM measurements are highly uncertain and vary widely in terms of magnitude (extreme positive and
negative absolute readings can occur). Kinsey et al. (2006)43 demonstrated that time-integrated TEOM
measurements compare well with gravimetric filter measurements of diesel-generated particulate matter.
Equation 2-27 shows the normalization process for a particular one second TEOM measurement.

P Wfj|t6r 7

PMnormaylzed,ij = ¦ 'J PMteqmj,;	Equation 2-27

L rAJjeoivu

Where:

i = an individual 1-Hz measurement (g/sec),
j = an individual test on an individual vehicle,

PMteomjj = an individual TEOM measurement on vehicle j at second /',

PMfnter,j~ the total PM2.s filter mass on vehicle j,

PM n0rmaiiZed,u = an estimated continuous emission result (PM25) emission result on vehicle j
at second /'.

2.1.2.1.4	Compute Average Rates by MOVES Operating Mode

After normalization, the data were classified into the 23 operating modes defined in Section 1.6 by
regulatory class and model-year group. Mean average results, and standard deviation for PM2.5 emission
values were computed in terms of g/hour for each operating mode. In cases where the vehicle and TEOM
samples were sufficient for a given mode (based on the number of points within each operating mode bin),
these mean values were adopted as the MOVES emission rates for total PM2.5.

2.1.2.1.5	Populating Missing and High-Power Operating Modes

As detailed in Appendix E, a log-linear regression was performed on the existing PM2 5 data against STP to
fill in emission rates for missing operating mode bins. Similar to the NOx rates for MY 2009 and older
vehicles, emission rates were extrapolated for the highest STP operating modes.

2.1.2.1.6	LHD and Urban Bus Emission Rates

The PM2.5 emission rates for LHD and Urban Buses are based on the available TEOM data collected on MHD
and HHD vehicles. We believe this is reasonable because the certification standards in terms of brake
horsepower-hour (bhp-hr) are the same for LHD, MHD, and HHD regulatory classes.

The following steps were conducted to adjust the emissions estimated from the MHD and HHD regulatory
classes because the data were not analyzed for the/scofe. 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 MOVES2009.44

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 factors45, which accounts for the lower
power requirements per mile (bhp-hr/mile) of light heavy-duty trucks versus MHD trucks. This scaling
factor estimates VSP-based emissions rates for LHD vehicles. We approximated the STP-based MOVES

64


-------
emission rates for LHD vehicles using these VSP-based rates.J Equation 2-28 used to derive the PM2.5
emission rates for LHD regulatory class is shown below:

LHDPM2,s emission rate = 0.46 x MHD (VSPbased)PM2,5 emission rate Equation 2-28

Despite the uncertainty used in this approach, the representative PM2.5gram per mile emission rates show
reasonable trends (see Figure 2-33). LHD2b3 and LHD45 emission rates for pre-2010 model years are lower
than the MHD and HHD emission rates, but higher than the Urban Bus emission rates.

Urban Bus emission rates are generally assumed to be the same as the HHD emission rates. However, they
have different emission deterioration effects as discussed in Appendix B.l. Also, for some model year
groups, the Urban Buses are subject to a different emission standard, so we adjust the emission rates
before deterioration by applying a ratio of the EPA certification standards. Table 2-21 displays the model
years for which the Urban Bus regulatory class has different PM emission standards from other heavy-duty
compression-ignition engines. For these model years (1991-2006), the Urban Bus PM emission rates were
set equal to the HHD emission rates multiplied by the ratio in emission standards before applying
deterioration. The gram per mile emission rates for Urban Bus presented in Figure 2-33 show expected
trends, with lower emission rates than the other regulatory classes starting in MY 1991 through 2006.

Table 2-21 Urban Bus PM Standards in Comparison to Heavy-Duty Highway Compression Engine Standards

Engine
Model Year

Standard for Heaw-Dutv
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

a The 0.1 g/bhp-hr US EPA Urban Bus standard began with model year 1993.
In California, the 0.1 g/bhp-hr Urban Bus standard began in 1991. MOVES
assumes all Urban Buses met the stricter CA standard beginning in 1991.

2,1,2,1,7 Model Year 2007-2009 Vehicles (with Diesel Particulate Filters)

The heavy-duty diesel emission regulations were made considerably more stringent for PM2.5 emissions
starting in model year 2007 - even considering the phase-ins and average banking and trading program,
the emission standard fell by a factor of ten from 0.1 g/bhp-hr to 0.01 g/bhp-hr. This increase in regulatory
stringency required the use of diesel particulate filter (DPF) systems on heavy-duty diesels. As a result, the
PM2.5 emission performance of diesel vehicles has improved dramatically.

At the time of analysis (originally done for MOVES2014, but carried over to later versions of MOVES for the
2007-2009 vehicles), no continuous PM2.5 emissions data were available on the 2007-2009 model-year
vehicles. However, heavy and medium heavy-duty diesel PM2.5 data were available from the EPA engine

J When this approximation was conducted in MOVES2010, the/Sca/e of 2.06 matched the average mass of LHD vehicles
for the source types. This approximation needs to be revisited now that we have updated the mass of LHD vehicles in
MOVES to range from 3.5 to 7.8 metric tons.

65


-------
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.s emissions from model years 2003 through 2006
to model year 2007. This same relative improvement was applied to the existing, operating mode-based,
1998-2006 model year PM2.5 running emission rates to estimate in-use rates for MY 2007-2009 vehicles.

An analysis of the certification data is shown in Table 2-22 below. It suggests that the particulate trap
reduced PM2.5 emissions by a factor of 27.7. This factor is considerably higher than the relative change in
the certification standards (i.e., a factor of 10). The reason for the difference is that the new trap-equipped
vehicles certify at emission levels which are much lower than the standard, and thus, create a much larger
compliance margin than previous technologies could achieve.

As an additional check on the effectiveness of the trap technology, EPA conducted some limited in-house
testing of a Dodge Ram truck, and carefully reviewed the test results from the CRC Advanced Collaborative
Emission Study (ACES) phase-one program, designed to characterize emissions from diesel engines
meeting 2007 standards. The results from these studies demonstrated that the effectiveness of working
particulate traps is very high.46

Table 2-22 Average Certification Results for Model Years 2003-2007

Certification Model Year

Mean
(g/bho-hr)a

St. Dev.

n

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

a Average ratio from MYs 2003-2006 to MY 2007 is 27.7.

2.1.2.1.8 Elemental Carbon and Non-Elemental Carbon Emission Factors

Particulate matter from conventional (pre-2007) diesel engines is largely composed of elemental carbon.
Elemental carbon is often used synonymously with soot and black carbon. Black carbon is important
because of its negative health effects and its environmental impacts as a climate forcer.47 Elemental
carbon from vehicle exhaust is measured with filter-based measurements using thermal optical methods.
Continuous surrogate measures of elemental carbon can also be made with photoacoustic instruments.

MOVES models Total PM2.5 emissions by vehicle operating mode using elemental carbon (EC) and non-
elemental particulate matter carbon (NonECPM), as shown in Equation 2-29.

PM2 5 = EC + NonECPM	Equation 2-29

By having emission rates for EC and nonECPM for each operating mode, the MOVES design permits the EC
fraction of PM (EC/PM) to vary for each operating mode. In practice, the data used to develop EC and

66


-------
nonECPM emission rates does not support such fine resolution, and the EC/PM is the same across all the
running exhaust operating modes except 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.6

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.46 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 is
composed of several different operating cycles. Because the fraction is based upon a range of driving
conditions, we applied the constant 9.98 percent EC/PM fraction across all operating modes for the 2007+
diesel emissions rates, including the idle operating mode (opModelD 1).

Figure 2-26 and Figure 2-27 show the operating mode trend for PM2.5 with the EC and nonECPM fractions.
As with NOx, the highest operating modes in each speed range will rarely be attained due to the power
limitations of heavy-duty vehicles and the high fscaie used for these model years, but are included in the
figures for completeness. At high speeds (greater than 50 mph; operating modes > 30), the overall PM2.5
rates are lower than the other speed ranges. For pre-2007 model years (Figure 2-26), the PM2.5 rates are
dominated by EC (except for the idle operating mode, opModelD 1). With the introduction of DPFs in
model year 2007 (Figure 2-27), we model the large reductions in overall PM2.5 rates and the smaller
relative EC contribution to PM emissions. Figure 2-35 shows the PM2.5 gram per mile emission rates
separated into the elemental carbon and non-elemental carbon fractions for HHD vehicles across model
years.

67


-------
35

30

" 25
.c

S 20

15

c

TO
0J

10

I EC

InonEC

llxl

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

Operating mode

Figure 2-26 MHD Diesel PM2.s Emission Rates for MY 2006 (age 0-3) by Operating Mode

1.4

1.2

a

S. 0.8

0.6

c

TO
0J

0.4

0.2

I EC

InonEC

LI_J

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

Operating mode

Figure 2-27 MHD Diesel PM2.5 Emission Rates for MY 2007 (age 0-3) by Operating Mode
2.1.2.2 2010-2060 Model Years

The MY 2010+ HDIUT data set described in Section 2.1.2.2 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 /sco;e values developed for the
2010+ MY NOx analysis described in Section 2.1.1.4.2 and Appendix G.

68


-------
2,1,2,2,1 Estimating Base Rates

Compared to the NOx, THC, and CO data, the PM25 emissions data reported from the HDIUT program had
many missing, negative and zero values. When the PM2.5 emissions data are distributed over the NOx FEL
groups, the number of vehicles with valid measurements (non-zero or non-negative) by regulatory class
and model year group was very limited as shown in Table 2-23. The number of vehicles with valid
measurements for each model year group and regulatory class are significantly smaller than the total
number of vehicles tested, with the exception of the HHD 2014-2015 model year group.

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





Valid Measurements bv NO* FEL Group

Total

Total

Reg
Class

Model Year
Group



Vehicles
Tested

02

0.35

05

Valid

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

1 Due to an absence of MHD 2014-2015 vehicles certified to 0.35 and 0.5 NOx in the HDIUT,
we replicated the HDIUT MHD 2010-2013 vehicles certified to 0.35 and 0.5 NOx FEL to the
2014-2015 model year group.

When evaluating the data by operating mode, the data are even more sparse. Table 2-23 shows the total
number of vehicles tested by model year group and regulatory class that had valid (positive) PM2.5
emissions data in at least one operating mode. However, the number of vehicles with valid measurements
is less when the data is evaluated by operating mode. Table 2-24 shows the minimum and maximum range
of valid vehicle measurements by individual operating modes. The operating modes with the smallest
number of valid measurements tended to be for the high-power operating modes (opModelD 29, 30, 39
and 40). The highest number of valid measurements tend to be for the high speed (>50 mph) at moderate
power (opModelD 33, 35, 37, and 38).

We addressed the issue of data sparsity in the PM2.5 emissions data by not using NOx FEL based grouping
and production volume weighting. By aggregating the data across NOx FEL groups, there was sufficient data
to divide the data into the two model year groups used for the NOx analysis: MY 2010-2013 and 2014-
2015/2016. Note that for this HDIUT dataset, LHD includes MY 2016 vehicles, but MHD and HHD only
include up to MY 2015.

69


-------
Table 2-24. Range of Valid Vehicle Measurements by Operating Mode by Regulatory Class and Model Year Group

Resulatorv
Class

Model Year
Group

Total Valid Vehicle
Measurements (in
at least one
operating mode)

Minimum Valid
Vehicle

Measurements for
anv operating mode

Maximum Valid
Vehicle

Measurements for
anv operating mode

LHD

2010-2013

56

11

52

2014-2016

21

3

18

MHD

2010-2013

32

4

27

2014-2015

29

3

26

HHD

2010-2013

92

17

83

2014-2015

52

4

45

The operating mode-based PM2.5 emission rates were estimated using Equation 2-30 and Equation 2-31,
which are similar to Equation 2-13 and Equation 2-14 used for the development of NOx emission rates
except the FEL grouping factor is removed. Zero or negative emission rates are as 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

^ n	_ Esec ERp0i,oM,c,MYG,veh,sec	Equation 2-30

^^pol,OM,C,MYG,veh	~T~

S^Ccount

_ ^tveh ERpol,OM,C,veh	Equation 2-3!

tHpol,OM,C>MYG -	~~7

Wi l 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; vehcoum

= a vehicle (in the class); number of vehicles in that category

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.

70


-------
Similar to N0X, the PM2.5 rates for LHD2b3 and LHD45 are identical and based on the combined LHD class
vehicles in HDIUT, while the MHD rates are based on MHD class vehicles, and HHD and Urban Bus rates are
based on HHD class vehicles.

Because we did not use production volume weighting by NOx FEL, the PM2.5 emission rates are the same
for each model year within the two model year groups. The 2014-2015 model year group sample contains
a higher penetration of HHD vehicles certified to the 0.2 NOx FEL group, than the 2010-2013 model year
HHD group (Table 2-23) which is consistent with the trend in production volumes (Figure 2-12). The LHD
vehicles had a similar proportion of vehicles certified to the 0.2 and 0.35 NOx FELs within the 2010-2013
and 2014-2016 model year groups. The PM2.5 emission rates decrease for each of the regulatory classes
between the MY 2010-2013 and 2014 and later model years as observed in Figure 2-34.

There was no valid HDIUT PM data for 2014-2015 MY MHD vehicles within the 0.35 and 0.50 FEL groups.
Because there are 2014-2015 MY MHD vehicles certified to these higher levels in the production volume
data (Figure 2-12), we supplemented the 2014-2016 MY MHD data with data from the MY 2010-2013 MHD
vehicles in the 0.35 and 0.5 FEL groups. As such, the MHD vehicles have a similar proportion of vehicles
certified to the 0.35 and 0.5 NOx FEL Groups within the 2010-2013 and 2014-2015 model year groups.

Figure 2-28 through Figure 2-30 display the average PM2.5 emission rates by regulatory class, model year
group, and operating mode. The error bars are the 95% confidence intervals of the mean, calculated by
treating the mean emission rates from each vehicle within each operating mode 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).

L As discussed in Section 2.3.2.2, we believe the reduction in THC and PM2.5 with 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 N02, which also reduces PM2.5 tailpipe emissions.

71


-------
0.60
0.50
0.40

ijr

g 0.30

i 0.20 :

0.10 :
0.00



¦ MY 2010-2013











¦ MY 2014-2016













































T 11

II

j	 T 1

1



III

-L I



Tt II

i X

I i It III



1-. 1, L IL 1







:|, I

_
1

1

0 1

11 12 13 14 15 16

! 21 22 23 24 25 27 28 29 30

i 33 35 37 38 39 40

MOVES OpMode

Figure 2-28 PM2.5 Emissions by Operating Mode for LHD Model Year Groups 2010-2013 and 2014-2016



0.80 :



0.70 :



0.60 :

"orT

0.50 i

CuO

-

fc

0.40 :



Q.

0.30 :



0.20 :



0.10 i



0.00

Li

I MY 2010-2013
I MY 2014-2015

Uii

iL

111

0

11 12 13 14 15 16

33 35 37 38 39 40

21 22 23 24 25 27 28 29 30
MOVES OpMode

Figure 2-29 PM2.5 Emissions by Operating Mode for MHD Model Year Groups 2010-2013 and 2014-2015

72


-------
L .x ii

I MY 2010-2013
! MY 2014-2015

li !¦ li ii

0 1 11

12 13 14 15 16

21 22 23 24 25 27 28 29 30
MOVES OpMode

33 35 37 38 39 40

Figure 2-30 PM2.5 Emissions by Operating Mode for HHD Model Year Groups 2010-2013 and 2014-2015

As observed in Figure 2-28 through Figure 2-30, the PM2.s 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 NO* FEL groups is higher for the MHD than most of the other regulatory class
and model year combinations, but not all (Table 2-23).

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 2-34, the MOVES age 0-3 PM2.5 rate ranges from 3 to 7 mg/mile for MY
2010-2014 HHD and LHD vehicles with the MHD significantly higher at 26 mg/mile and from 2 to 7 mg/mile
for all HD MY 2014+ vehicles. Other studies have reported PM2.5 rates in the range of 1-7 mg/mi for MY
2010+ vehicles equipped with DPF and SCR and certified to NOx standard of 0.20 g/bhp-hr.48,49,50 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 years for 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.51 Because the age 0-3 LHD2b3 MOVES emissions rates are well below the Tier 3 standard, we
do not estimate a reduction in zero-mile PM2.5 rates with the phase-in of Tier 3 in MY 2018 and later
vehicles as we do for NOx emissions. For PM2.5, the MOVES LHD2b3 rates are in compliance with the 3
mg/mile standard as shown in Figure 2-34 However, we do incorporate different aging effects due to the
extended useful life in the Tier 3 program as discussed below in Section 2.1.2.3.2.

2.1.2.2.2 Elemental Carbon and Non-Elemental Carbon Emission Rates

The EC (9.98 percent) and non-EC (90.02 percent) fractions for MY 2010 and later are unchanged from MY
2007-2009 analysis described in section 2.1.2.1.8. As discussed in the Speciation Report6, the ACES Phase 1

73


-------
is deemed to be an appropriate source of the EC/PM fractions for 2010+ engines which continue to use
DPF technology along with other selective catalytic 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.6

2,1,2,2,3 DPF Regeneration Events

"As a DPF collects PM, the passage of exhaust gas through the pores of the filter element may be
progressively blocked, causing an increase in exhaust backpressure. Collected PM is combusted and
reduced to ash during filter regeneration, effectively unblocking the pores of the filer element and
decreasing exhaust backpressure."52

MOVES 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.M

2,1,2,3 Tampering and Mal-maintenance

Tampering refers to intentional disabling or modifying the vehicle engine, control systems, and/or exhaust
aftertreatment systems that results in increased emissions. Mal-maintenance refers to lack or improper
maintenance of the engine and aftertreatment, including neglecting to repair broken or mal-functioning
engine and aftertreatment parts, which also increases emissions. MOVES uses the same methodology to

M As discussed in the MOVES speciation report6, 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

74


-------
apply tampering and mal-maintenance (T&M) adjustment factors to both the pre-2010 and 2010 and later
PM2.5 emission rates.

We followed the same tampering and mal-maintenance methodology and analysis for PM2.5 as we did for
NOx, as described in Appendix B. We account for the emission increases with age by multiplying the zero-
mile emission rates by T&M adjustment factors and scaled aged effects (Equation 8-3). The MOVES T&M
adjustment factors on PM2.5 emissions over the fleet's useful life are shown in Table 2-25 and derived in
Appendix B.8. The value of 89 percent for 2010-2012 model years reflects the projected effect of heavy-
duty on-board diagnostic deterrence/early repair of T&M effects. It is an eleven percent improvement
from model years which do not have OBD (i.e., 2007-2009). The 67 percent value for 2013+ is driven by the
assumed full-implementation of the OBD in 2013 and later trucks, which assumes a 33 percent decrease in
T&M emission effects.

While we model a differential effect of DPFs on EC and nonECPM (see Section 2.1.2.2.2), the MOVES T&M
methodology applies the same adjustment to EC and nonECPM.

Table 2-25 Tampering and Mal-maintenance Adjustment Factors for PM2.5 over the Useful Life2

Model Year Group

Increase in PM2.5 Emissions (%)

Pre-1998

85

1998 - 2002

74

2003 - 2006

48

2007-2009

100

2010-20121

89 (HHD, MHD, LHD45, and Bus) 67 (LHD2b3)

2013+

67

1 LHD2b3 achieve ful

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-4 and Table B-5)

2,1,23,1 1950-2009 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.

75


-------
Figure 2-31 shows the estimated PM2.5 emission rates from MY 2006 heavy-duty diesel vehicles, accounting
for the effect of tampering and mal-maintenance. The different ages at which each regulatory class
reaches full useful life can be observed as the age when the emissions reach their maximum value. The
age 0-3 emission rates for the 2006 and earlier model year were extrapolated using the T&M adjustment
factors since the majority of these engines were older than three years when tested in the E-55/59
program.

0.012

0.009

v
E

ra
£
c

o
'

-------
0.012

10-14

15-19

Reg Class

41-LHD2b3

¦	42-LHD45
46-MHD67

¦	47-HHD8
48-Urban Bus

8-9
Age (years)

Figure 2-32 Heavy-duty Diesel PM2.s Emission Rates (g/mile) by Age Group and Regulatory Class for Model Year 2015

using Nationally Representative Operating Mode Distribution

2.1.2.3.3 2027-2060 Model Years

The T&M approach for model years 2027 through 2060 is similar to the approach for 2010 through 2026,
but since the HD2027 standards set longer warranty period and useful life requirements (as shown in ) for
diesel heavy-duty vehicles starting with the 2027 model year, we adjusted the T&M effects and the
resulting PM2.5 emission rates to account for those changes. This is discussed in more detail in Appendix B.

2.1.2.4 Model Year Trends

Figure 2-33 displays the PM2.5 rates by model year and regulatory class for 0-3 age group estimated in
grams per mile (g/mile) using nationally representative operating mode distributions and average speeds.
MOVES models a very large decrease in PM2.5 emission rates starting in model year 2007 (decrease on
order of ~10 to 40 times), when all regulatory classes are assumed to have implemented diesel particulate
filters, with the exception of gliders (Section 2.5). As discussed in Section 2.1.1.8, some of the variation
between regulatory classes is also due to differences in the application of T&M adjustment factors and
differences in the operating mode distributions and average speeds.

Figure 2-34 provides resolution to the model year changes in PM2.5 emission rates for the 2007 and later
model years. Further reductions in PM2.5 emissions are observed for each regulatory class between the
2007-2009 and the 2014 and later emission rates. The higher rates for MY 2010 and later MHD vehicles
stem directly from the HDIUT data as shown in Figure 2-29 and discussed in Section 2.1.2.2.1. The minor
variation in the gram per mile emission rates within the 2010-2013 and 2014+ model year groups by model
year and regulatory class are due to differences in operating mode distributions. Finally, there is a small
change in PM2.5 rates beginning in MY2027 due to warranty period provisions of the HD2027 rule.

77


-------
Reg Class

41-LHD2b3

42-LHD45

46-MHD67

47-HHD8

48-Urban	Bus

49-Gliders

1980	2000	2020	2040

Model Year

Figure 2-33. Base running emission rates for PM2.5from age 0-3 diesel heavy-duty vehicles averaged over a nationally

representative operating mode distribution.

0.03

2010	2020	2030	2040	2050

Model Year

Figure 2-34 Base running emission rates for PM2.sfrom age 0-3 diesel heavy-duty vehicles for MY 2007 and later
averaged over a nationally representative operating mode distribution.

Reg Class

41-LHD2b3

42-LHD45

46-MHD67

47-HHD8

48-Urban	Bus

78


-------
Figure 2-35 shows the PM2.5 emission rates separated into elemental carbon (EC) and non-elemental
carbon (nonEC) fractions for age 0-3 HHD diesel vehicles using nationally representative operating mode
distributions and average speeds. The EC fraction stays constant until model year 2007, when it is reduced
to less than ~10 percent due the implementation of diesel particulate filters.

0.6

£

B)

0.4'

re
Ct
c

o

E

HI

cvi 0.2

EC

NonEC

1990

2000

2010
Model Year

2020

2030

Figure 2-35 Heavy Heavy-duty (HHD) Diesel PM2.5 Emission Rates by Elemental Carbon (EC) and Non-Elemental Carbon
(nonEC) Fraction for the 0-3 Age Group by Model Year using Nationally Representative Operating Mode Distributions

2.1.3 Total Hydrocarbons (THC) and Carbon Monoxide (CO)

While diesel engine emissions of THC and CO are important, diesels are not the largest contributors to
mobile source THC and CO emission inventories. Diesel engines generally produces less THC and CO than
similar gasoline engines. 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 diesel 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 1950-2009 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 1950-1989, 1990-
2006, and 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 methodology12, where emission rates were calculated from 1-hz data produced from chassis
dynamometer testing. Data sources were all heavy-duty chassis test programs:

79


-------
1.	CRC E-55/5940: As mentioned earlier, this program represents the largest volume of heavy-
duty emissions data collected from chassis dynamometer tests. All tests were used, not
just those using the TEOM. Overall, 75 trucks were tested on a variety of drive cycles.
Model years ranged from 1969 to 2005, with testing conducted by West Virginia University
from 2001 to 2005.

2.	Northern Front Range Air Quality Study (NFRAQS)53: This study was performed by the
Colorado Institute for Fuels and High-Altitude Engine Research in 1997. Twenty-one HD
diesel vehicles from model years 1981 to 1995 selected to be representative of the in-use
fleet in the Northern Front Range of Colorado were tested over three different transient
drive cycles.

3.	New York Department of Environmental Conservation (NYSDEC)54: NYSDEC sponsored this
study to investigate the nature and extent of heavy-duty diesel vehicle emissions in the
New York Metropolitan Area. West Virginia University tested 25 heavy heavy-duty and 12
medium heavy-duty diesel trucks under transient and steady-state drive cycles.

4.	West Virginia University: Additional historical data collected on chassis dynamometers by
WVU is available in the EPA Mobile Source Observation Database.

The pre-2010 onroad data used for the NOx analysis was not used since THC and CO were not collected in
the MEMS program, and the ROVER program used the less accurate non-dispersive infrared (NDIR)
technology instead of flame-ionization detection (FID) to measure HC. To keep THC and CO definitions and
data sources consistent, we only used chassis test programs which measured THC using a FID exclusively
for the analysis. Time-series alignment was performed using a method similar to that used for light-duty
chassis test data.

Table 2-26 Numbers of Vehicles by Model Year Group, Regulatory Class, and Age Group

Model year group

Regulatory class

Age group

0-3

4-5

6-7

8-9

10-14

15-19

20+



HHD

58

19

16

9

16

6

7



MHD

9

6

5

4

12

15

6

1960-2002

LHD45

2





1









LHD2b3

6















Bus

26





1

3





2003-2006

HHD

6















HHD, MHD, LHD45,
LH2b3, Bus

No vehicles

:or this model year group. Rates for

2007-2009

this model year group are based on MY 2003-



2006 with 80 percent reduction.





2.13.12 1950-2006 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.

80


-------
Using the methods introduced in the NOx analysis, we averaged emissions by vehicle and operating mode.
We then averaged across all vehicles by model year group, age group, and operating mode. In populating
the emission rates in MOVES, we used the age group that had the most data in each regulatory class and
model year group combination. These age groups are shown in Table 2-27. We then used the T&M effects
discussed in Section 2.1.3.3 to extrapolate the emission rates for each age group. For missing operating
modes, we extrapolated using STP as was discussed for NOx in Section 2.1.1.4.2. For the 1950-2002 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 2-27 Age Groups for which THC and CO Emission Rates are Populated Directly Based on the Data

Regulatory class

Model yeargroupN

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 1950-2002 to the LHD 2003-2006 model year group. For 2003-2006
MHD and Urban Bus regulatory classes emission rates, we applied the HHD 2003-2006 emission rates.

Figure 2-36 and Figure 2-37 show the rates for MHD and HHD for MY 2002 for THC and CO, respectively,
based on the methods described above. The THC and CO mean emission rates increase with STP, though
there is much higher uncertainty than for the NOx rates (Figure 2-1). This pattern could be due to the
smaller data set or may reflect a less direct correlation of THC and CO to STP as is observed for the 2010
and later model year rates.

N When we extended calculations to 40 ages, we duplicated the emission rates for MY 1960 for model
years back to 1950.

81


-------
100

90
80
70
Jii 60

0)

E 50
u

ra 40
a

30
20
10
0

~ MHD
¦ HHD/Bus

~1	1	1	1	1	1	1~

~1	1	1	1	1	1	1	T"

~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 2-36 THC Emission Rates [g/hr] by Operating Mode for Model Year 2002 and Age Group 0-3. Error Bars
Represent the 95 Percent Confidence Interval of the Mean

600

500

400 -

300

o

u

2 200

100

~ MHD
¦ HHD/Bus

i >

« . S 5

-I	1	1	1	1	1	1-

ful

	1	1	1	1	1	1	1	r~



-i—i—i—i—i—i

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

Operating mode

Figure 2-37 CO Emission Rates [g/hr] by Operating Mode for Model Year 2002 and Age Group 0-3. Error Bars
Represent the 95 Percent Confidence Interval of the Mean

82


-------
2,1,3,1,3 2007-2009 Model Years

With the increased use of diesel oxidation catalysts (DOCs) in conjunction with DPFs, we assumed an 80
percent reduction in zero-mile emission rates for both THC and CO for 2007-2009 model years. The
derivation of the T&M effects for 2007-2009 model years are presented in

Table 2-28 and discussed in Appendix B.9. As shown in Figure 2-48, the CO emission rates developed using
this assumption are significantly lower than the model year 2010 and later emission rates that were
developed based on the HDIUT data and could be re-evaluated in future versions of MOVES.

2,1,3,2 2010-2060 Model Years

We used the MY 2010+ HDIUT data set, using the same vehicles as used for NOx and described in Section
2.1.1.2 and Table 2-2. The HDIUT dataset includes vehicles in the HHD, MHD, LHD45, LHD2b3 and Urban
Bus regulatory classes. The HDIUT emission measurements are made using instruments that conform to
the requirements described in 40 CFR Part 1065, which require the use of a flame ionization detector (FID)
for measuring total hydrocarbons (THC)55 and a non-dispersive infrared (NDIR) analyzer for carbon
monoxide (CO)56

The THC and CO emission rates have more uncertainty than the NOx emission rates, which suggests a less
direct correlation of THC and CO to STP. Nevertheless, we followed the analysis methodology used for MY
2010+ NOx rates as described in Sections 1.6 (calculation of STP and assignment of operating modes),
2.1.1.5 (calculation of mean emission rates), 2.1.1.5.1 (NOx FEL groups) and Appendix G (selection of/sco/e).
Figure 2-38 and Figure 2-39 display the HHD THC and CO emission rates estimated from the HDIUT data by
the NOx FEL Groups used to develop the MY 2010-2013 emission rates. Comparisons of the THC and CO
emission rates by NOx FEL Groups for the LHD and MHD regulatory classes are provided in Appendix H.
These comparisons show that there are significant differences among the emission rates in different NOx
FEL groups for THC emissions. The THC emission rates in the 0.2 and 0.35 NOx FEL group are lower than the
THC emission rates from the vehicles in the 0.5 NOx FEL groups for each regulatory class, with the
differences being the most significant for MHD (Appendix H.1.2) and especially HHD (Figure 2-38). For CO,
there is not a consistent trend among the different NOx FEL groups and regulatory classes. Regardless, we
have analyzed the CO emission rates using the NOx FEL groups for consistency.

83


-------
0.020

0.015 -

no

™ 0.010

u

X

0.005 -

0.000

ta HHD FEL 0.20, MY 2010-2013,!N=78

I

Ic HHD FEL 0.35, MY2010-2015,!N=31

I	I

¦ HHD FEL 0.50, MY2010-2015,! N=35

i .i

0 1, 11 12 13 14 15 16

i

i

iii

i i

j

1

33 35 37 38 39 40

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

Figure 2-38 Average HHD THC Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013 and the 0.35
and 0.5 NOx FEL for MY 2010-2015. Error Bars are 95% Confidence Intervals of the Mean

0.08 -

0.06 -

0.04 -

0.02 -

Lk

ii

I

I HHD FEL 0.20, MY 2010-2013, N=78
l HHD FEL 0.35, MY 2010-2015, N=31
I HHD FEL 0.50, MY 2010-2015, N=35

i

i.

i

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES OpMode

Figure 2-39 Average HHD CO Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013 and the 0.35
and 0.5 NOx FEL for MY 2010-2015. Error Bars are 95% Confidence Intervals of the Mean

Figure 2-40 and Figure 2-41 display the production-weighted average emission rates for THC and CO
emissions for model year 2013 HHD trucks. Production-weighted averages are calculated for each model
year between model year 2010 and 2018 using the production volumes displayed in Figure 2-12.

84


-------
0.0050

0.0040

«T 0.0030

.25
u

IE 0.0020

0.0010

0.0000



¦ HDIUT, MY 2013



























. , ,11

1



i I 1

1 II

i i



1

1

ii i i 1

II

0 1 i 11 12 13 14 15 16 ; 21 22 23 24 25 27 28 29 30 i 33 35 37 38 39 40

MOVES OpMode

Figure 2-40 THC Emissions by Operating Mode from HHD Trucks for Model Year 2013, Error Bars represent the 95

percent confidence interval of the Mean

0.12

0.09 -

hfl

•= 0.06

O

u

0.03 -

0.00



IHDiUT, MY 2013

i.

i



l

0 1 , 11 12 13 14 15 16 ; 21 22 23 24 25 27 28 29 30 i 33 35 37 38 39 40

MOVES OpMode

Figure 2-41 CO Emissions by Operating Mode from HHD Trucks for Model Year 2013. Error Bars represent the 95

percent confidence interval of the Mean

Figure 2-42 and Figure 2-43 display the comparison of the MY 2010-2013 and MY 2014-2015 groups within
the HHD 0.2 NOx 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.

85


-------
0.0040
0.0035

—	0.0030

i/i

no

—	0.0025
U

I- 0.0020
0.0015
0.0010
0.0005
0.0000

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES OpMode

Figure 2-42 THC emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the HHD 0.20 NOx FEL Group

¦ MY 2010-2013



¦ MY 2014-2015

























|









ii ii ii li ii ii

1 ii ii ii ii ii Ii

I



I 1

T I





i. .. ii Ii 1



MY 2010-2013
MY 2014-2015

i- i I' I1

l:

Ui

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES OpMode

Figure 2-43 CO emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the HHD 0.20 NOx FEL Group

LHD2b3 and LHD45 emission rates are based on the vehicles with "LHD" service class in the HDIUT data set,
which only contains engine-certified LHD45 vehicles. Urban Bus emission rates are based on HHD vehicles
in the HDIUT data set. MHD and HHD emission rates are based on the emission rates from those vehicle
classes. The HDIUT data set predominantly contains vehicles in the 0-3 age group with only a handful
vehicles in the 4-5 age group. Since the HDIUT data is measured and submitted by the manufacturer and
the test vehicles are required to be free of any tampering or mal-maintenance, we can safely assume that
they represent zero-mile vehicles for the purpose of assigning base rates and applying the tampering and
mal-maintenance effects.

A comparison of HDIUT-based THC and CO emission rates for MY 2010+ heavy-duty vehicles by regulatory
class are shown in Figure 2-47 and Figure 2-48, respectively. The THC rates, generally 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

86


-------
which are based on emission measurements. The variation in the 2010-2018 rates reflects the model year
variation in the production volume by NOx FEL group, and use of the different 0.2 FEL NOx model year
group between 2010-2013 and 2014 and later.

In the 2017 review of a draft version of this report, we received a comment that single-cell-NDIR-based CO
measurements suffer from severe drift that is not corrected by zero and span checks because the
calibration gases are dry, while vehicle tailpipe exhaust gases are not dry. Based on the HDIUT data, it is
not possible for us to determine if MY 2010+ CO emission rates are affected by the alleged drift in the CO
measurements. We looked at the CO emissions for each of 93 vehicles in the HHD 0.20 FEL group (from the
2010-2016 selection years) and confirmed the high average CO rate is not due to a few outliers. Further,
the CO emission rate for the MHD and LHD vehicles is significantly lower (see Figure 2-48). Based on the
available data and trends, we are unable to confirm whether or not the high CO emissions for the HHD
vehicles is real or an artifact of CO sensor drift. In Section 2.1.5, we demonstrated that the fleet-average
heavy-duty CO emission rate estimates from MOVES compare well with measurements from heavy-duty
exhaust plume capture and tunnel measurement campaigns conducted in 2015 and 2017, which increased
our confidence that the CO emission rates measured from HDIUT are reasonable. Thus, we decided to
accept the reported HDIUT CO emission rates as valid.

As discussed in 2.1.1.5.5, we did not reduce the LHD2b3 zero-mile THC and CO emission rates due to the
implementation of the Tier 3 standard. For LHD2b3 2010 and later vehicles, the MOVES emissions rates are
based on LHD45 vehicles measured in the HDIU program, as described above. The surrogate LHD45
emission rates, for THC, CO, and PM2.5emissions, imply that current levels on the FTP cycle are substantially
below the Tier 3 standards. For example, when MOVES rates are used to simulate FTP cycle for NMHC, the
result is a rate of approximately 0.05 grams per mile, while the simulated FTP estimate for CO is less than
1.0 gram/mile. However, we did account for the lengthened useful life standard required by the Tier 3
standard in the Tampering & Mal-maintenance standards as discussed in the next section.

The THC and CO emissions for well-maintained vehicle are treated as constant for all model years beyond
2018, but the tampering and mal-maintenance effects vary by model year as described in the next section.

2,1,3,3 Tampering and Mal-maintenance

For all model years, we applied tampering and mal-maintenance effects to adjust emissions from the
measured age to all age groups, lowering emissions for younger ages and raising them for older ages, using
the methodology described in Appendix B. We applied the tampering and mal-maintenance effects shown
below in Table 2-28 to CO and THC.

For MY2027+ LHD45, MHD, HHD vehicles, we adjusted the T&M effects in estimating the emission rates for
CO and THC to account for the longer warranty period and useful life requirements in HD2027 standards
(as shown in Table B-2).

87


-------
Table 2-28 Tampering and Mal-maintenance Effects for THC and CO over the Useful Life

Model years

Increase in THC and CO Emissions (%)

1994-2003

300

2003 - 2006

150

2007-2009

150

2010-20121

29 (HHD, MHD, LHD45, and Bus) 22 (LHD2b3)

2013-2026

22

2027+

22

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. Tables B-4 and B-5 estimate 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 Tables B-4 and B-5and applied the result to the zero-mile (or age 0) emissions
rate to estimate the emissions rate by age group using Equation 8-3. As shown in Table B-4, there are
different age effects for Tier 2 and Tier 3 LHD2b3 vehicles. We used Equation 2-27 to estimate a weighted
average of the THC and CO emission rates during the Tier 3 phase-in (Model year 2017-2022) as discussed
in Section 2.1.1.5.5.

Figure 2-44 and Figure 2-45 show THC and CO emission rates by age group for MY 2015. Due to our
projections of T&M effects, there are large increases as a function of age. Additional data collection would
be valuable to determine if real-world deterioration effects are consistent with those in MOVES.

88


-------
0.06

E

3


-------
per mile emission rates within the model year groups and between regulatory classes are due to
differences in operating mode distributions. Differences in the emission rates for age group 0-3 between
regulatory classes are also due to different application of the T&M adjustment factors (Section 2.1.3.3). For
example, as discussed in Section 2.1.3.1.3, the zero-mile MY 2003-2006 emission rates by operating mode
are equivalent for HHD, MHD, and Urban Bus, but the T&M adjustment factors are applied differently for
each regulatory class, and the operating mode distributions are difference, resulting in the differing gram
per mile emission rates observed for 2003-2006 in Figure 2-46 and Figure 2-48.

The MY 2007-2009 emission rates reflect the use of diesel oxidation catalysts and are derived by reducing
the CO and THC emissions in MY 2003-2006 by 80 percent and applying the model-year and regulatory
class specific T&M adjustment factors. For MY 2010-2018, the significant variation in the emission rates by
model year are due to the model year specific production volumes of the NOx FEL Group and the model
year split of the 2010-2013 and 2014 and later 0.2 NOx FELgroup (Section 2.1.1.5.2 and Section 2.1.3.2).
For example, the spike in THC emissions observed in the 2010 model year HHD vehicles (Figure 2-47) is
explained by the high THC emissions of the 0.5 NOx FEL group (Figure 2-38), and the high production
volumes of the 0.5 NOx FEL engines in MY 2010 (~50% of the total HHD, see Figure 2-12).

Finally, there are small reductions in emission rates attributable to the longer warranty period and useful
life requirements in the HD2027 rule.

0.4

E
o>

C£

0

1

0.2'

0.0





















[til





























n	















Reg Class

¦*- 41-LHD2b3
— 42-LHD45

46-MHD67

47-HHD8

48-Urban	Bus

49-Gliders

1980	2000	2020	2040

Model Year

Figure 2-46 Base running emission rates for THC from age 0-3 diesel heavy-duty vehicles averaged over a nationally

representative operating mode distribution.

90


-------
0.15

t£

0

1

0.10

0.05

0.00

;

























h















	









Wr



.......





1444-1-





















2010

2020

2030
Model Year

2040

2050

Reg Class

41-LHD2b3
— 42-LHD45

46-MHD67

47-HHD8

48-Urban	Bus

Figure 2-47 Base running emission rates for THCfrom age 0-3 diesel heavy-duty vehicles for MY 2007 and later
averaged over a nationally representative operating mode distribution.





_ A

























\~



















t













Reg

Class

41-LHD2b3

42-LHD45

46-MHD67

47-HHD8

48-Urban	Bus

49-Gliders

1980	2000	2020	2040

Model Year

Figure 2-48 Base running emission rates for CO from age 0-3 diesel heavy-duty vehicles averaged over a nationally

representative operating mode distribution.

2.1.4 Energy

2.1.4.1 1950-2009 Model Years

When we extended calculations to 40 ages, we duplicated the emission rates for MY 1960 for model years
back to 1950.

91


-------
2.1.4.1.1	LHD

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 (regClasslDs 20 and 30), varied by fueltype, model year group, engine technology, and
"size weight fraction" as discussed in the MOVES2010a energy updates report.57 The energy rates in
MOVES2010a were simplified to be a single set of energy rates for each regulatory class, fuel type and
model year combination by weighting across engine size, engine technology, and vehicle weight according
to the default population in the MOVES2010 sample vehicle population table. The resulting C02 (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 2-51 and Figure 2-52.
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 2.1.1.2), 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 C02.

In MOVES, C02 emissions were used as the basis for calculating energy rates. To calculate energy rates
(kJ/hour) from C02 emissions (Equation 2-32), we used a heating value (HV) of 138,451 kJ/gallon and C02
fuel-specific emission factor (fC02) of 10,180 g/gallon58 for conventional diesel fuel.

HV

fenergy = fCo2 T~	Equation 2-32

JC02

The energy rates for the MHD, Urban Bus, and HHD vehicle classes are shown in Figure 2-49. Compared to
other emissions, the uncertainties in the energy rates are smaller, in part because there is no classification
by age, model year, or regulatory class. Thus, the number of vehicles used to determine each rate is larger,
providing for a greater certainty of the average mean energy rate.

Operating mode-based energy consumption rates are the same across MHD, Urban Bus, and HHD
regulatory classes. However, the distribution of time spent in the operating mode varies between these
regulatory classes based on differences in their activity and tractive power demand. Thus, the C02 (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 2-51 and Figure
2-52.

92


-------
6

5 -

4 -


-------
20 -

¦ HHD FEL 0.20, MY 2010-2013
p HHD FEL 0.35, MY 2010-2015
p HHD FEL 0.50, MY 2010-2015



N=78
N=31
N=35

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 2-50 Average HHD C02 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 2-29, the majority of vehicles within the NOx FEL groups of 0.35 and 0.5 are MY 2010-
2013 vehicles, for both HHD and MHD vehicles. As discussed in the next subsection, we expect improved
energy efficiencies in MY 2014 and later vehicles due to the phase-in of the Phase 1 Heavy-duty
Greenhouse Gas Emission Standards. Due to the small sample of vehicles in the 0.35 and 0.5 NOx FEL
groups, we assume that the MY 2010-2016 energy rates are representative of both the MY 2010-2013 and
the 2014-2016 model year groups.

Table 2-29 HDIU Vehicles with Valid C02 Measurements By Regulatory Class, Model Year Group and NOx FEL Group



NOx FEL Group

Total
Valid

Tested
Vehicles

Reg
Class

Model Year
Group

0.2

0.35

0.5

LHD

2010-2013

52

0

10

62

64

2014-2016

27

0

5

32

32

MHD

2010-2013

21

23

9

53

55

2014-2015

19

0

0

19

19

HHD

2010-2013

78

26

35

139

139

2014-2015

44

5

0

49

55

Using this method, the energy rates for each model year are unique based on NOx FEL based production
volume weighting, as can be observed in the model year variability among the MY 2010-2013 C02 (g/mile)
emission rates and fuel economy values (miles per gallon) shown in Figure 2-51 and Figure 2-52.

2.1.4.3 2014-2060 Model Years

2.1.4.3.1 LHD45, MHD, Urban Bus, arid 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

94


-------
section (Section 2.1.4.2). The MY 2014-2018 running energy rates were calculated as a weighted average
using the MY 2014-2015/2016 rates from the 0.2 NOx FEL group, and the MY 2010-2016 rates from the
0.35 and 0.5 NOx FEL groups. Although the 0.35 and 0.5 NOx FEL groups contain measurements from MY
2010-2013 vehicles (Table 2-29), applying the MY-specific production volume weighting of the FEL groups
means that the MY 2014-2018 emission rates are primarily or entirely based on the data from the MY
2014-2015/2016 vehicles in the 0.2 NOx FEL group. The MY 2014-2018 running energy rates developed in
this step are considered the "baseline" - the reductions in energy rates expected from the Medium- and
Heavy-Duty Greenhouse Gas (GHG) Phase 1 Rule59 were applied to the "baseline" energy rates as
described below.

MOVES accounts for the improved fuel efficiency achieved by the HD GHG Rulemakings in two ways. First,
the running, start, and extended idle rates for total energy consumption are reduced to be consistent with
the HD GHG rules. Second, the truck weights and road-load coefficients are updated to reflect the lower
vehicle curb weights through lightweighting of materials, lower resistance tires, and improved
aerodynamics of the vehicle chassis. Vehicle weights and road-load coefficients are discussed in the
Population and Activity Report.

The HD GHG Phase 1 rule59 was implemented starting with 2014 model year and increased in stringency
through model year 2018. The reductions in start and running energy rates reflect the improvements
expected from improved energy efficiency in the powertrain. The estimated reductions for heavy-duty
diesel energy rates from the HD GHG Phase 1 rule are shown in Table 2-30.

The MY 2014-2016 running energy rates were not adjusted for the HD GHG Phase 1 rulemaking because
the impact of Phase 1 is assumed to be included in the measurements from the MY 2014-2015/2016
vehicles sampled from the HDIUT program. Instead, we renormalized the Phase 1 GHG reductions using
the MY 2014-2016 as the baseline using Equation 2-33.

Renormalized Phase 1 reductions in yeart

1 — (reductions inyeari)

1 — (Average reductions in 2014 thru 2016)	Equation 2 33

For example, the renormalized reductions for LHD and MHD in 2017-2020 are calculated as:

Renormalized Phase 1 reductions for LHD and MHD in 2017 thru 2020
1 - (9%) 91%

= 1 " 1 - (5%) = 1 " 95% = 1 " 96% = 4%	2'34

We applied the renormalized reductions to estimate the MY 2017 and later running energy rates, as shown
in Table 2-30. As discussed in Section 2.2.4, because the start energy rates were not updated with more
data from model year 2014-2016 vehicles, the reduction in energy consumption from starts due to HD
GHG Phase 1 rule was modeled by directly using the reductions estimated from the rule.

95


-------
Table 2-30 Estimated Reductions in Diesel Engine Energy Consumption Rates from the HD GHG Phase 1 Program60

Regulatory
Class

Fuel

Model
Years

Estimated Reduction from the
MY 2013 Baseline (applied to
starts)

Renormalized Reductions to
MY 2014-2016 Energy Rates
(applied to running)

HHD and
Urban Bus

Diesel

2014-2016

3%

-

2017-2020

6%

3%

LHD and
MHD

Diesel

2014-2016

5%

-

2017-2020

9%

4%

MOVES also incorporated the Medium- and Heavy-Duty GHG Phase 2 rule61 as implemented. The Phase 2
program begins in 2021 and phases in through model year 2027. The Phase 2 rule breaks 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)

The Phase 2 Rule set separate standards for engines and vehicles and ensured improvements in both. It
also set separate standards for fuel consumption, C02, N20, CH4 and HFCs.0

Because the Phase 2 rulemaking set different standards for vocational vehicles and tractor-trailers and
because single-unit vocational vehicles and tractor-trailers are mapped to the same regulatory classes
(MHD and HHD) under the default MOVES framework for emission rates, we modeled changes in running
energy rates due to Phase 2 using the EmissionRateAdjustment table. The EmissionRateAdjustment table
includes the following data fields, many of which are shared with the EmissionRate table:

polProcessID (primary key)
sourceTypelD (primary key)
regClassID (primary key)
fuelTypelD (primary key)
beginModelYearlD (primary key)
endModelYearlD (primary key)
emissionRateAdjustment
dataSourcelD

Table 2-31 summarizes the energy rate reductions stored in the EmissionRateAdjustment table which are
applied to the running rates in MOVES for MY 2021 through 2027 heavy-duty diesel vehicles.

0 HFCs are not modeled in MOVES, and the N20 and CH4 standards are not considered technology forcing on
emissions.

96


-------
Table 2-31 Estimated Reductions in Diesel and CNG Engine Energy Consumption Rates due to the HD GHG Phase 2

Program61

VEHICLE SOURCE TYPE
(SOURCE TYPE ID)

FUEL

MODEL
YEARS

REDUCTION FROM MY
2020 ENERGY RATES

Long-haul Combination Truck
(62)

Diesel & CNG

2021-2023

6.9%

2024-2026

11.4%

2027+

15.3%

Short-haul Combination Truck (61)

Diesel & CNG

2021-2023

6.8%

2024-2026

11.3%

2027+

14.4%

Other Bus, School Bus, Refuse Truck,
Single-Unit Short-Haul, Single-Unit
Long-Haul, Motorhomes
(41, 43, 51, 52, 53, 54)

Diesel & CNG

2021-2023

7.8%

2024-2026

12.3%

2027+

16.0%

Transit Bus (42)

Diesel & CNG

2021-2023

7.0%

2024-2026

11.8%

2027+

14.4%

Thus, for LHD45, MHD, HHD and Urban Bus, the running energy rates for MY 2021 through 2027 are
estimated with a chain of calculations starting with the HDIUT-based estimates by operating mode and
regulatory class, then reduced by applying the HD GHG Phase 1 reduction in Table 2-30 and further
reduced by applying the HDGHG Phase 2 reductions listed in Table 2-31. The reductions shown in Table
2-31 reflect a combination of improvements to the engine and other systems, but exclude improvements
to aerodynamics and tire rolling resistances. The projected improvements due to aerodynamics and tire
rolling resistance are reflected in new road load coefficients as described in the Population and Activity
Report.

In MOVES5, we updated the running energy consumption rates for all HD vehicles based on the HD GHG
Phase 3 rule. We maintained the same relative energy consumption rates across operating modes as
already existed, adjusting them down according to the changes in standards.

The change in emission standards for Phase 3, relative to Phase 2, can be found in the HD Phase 3 docket
in the HD TRUCS Output Calculator.62 The changes in standards by regulatory vehicle type are presented
below in Table 2-32. Bolded rows indicate those that we used to update MOVES energy consumption rates.
All other regulatory groupings are optional or, in the case of Heavy-Haul Tractors, represent a very small
portion of the HD fleet, so we ignore them for the purposes of updating energy consumption rates in
MOVES.

97


-------
Table 2-32 Phase 3 fleet-average C02 emission percent reductions

Regulatory Grouping

Model Year

2027

2028

2029

2030

2031

2032+

LHD Vocational

17%

22%

27%

32%

46%

60%

MHD Vocational

13%

16%

19%

22%

31%

40%

HHD Vocational

0%

0%

13%

15%

23%

30%

MHD All Cab and HHD Day Cab Tractor

0%

8%

12%

16%

28%

40%

Sleeper Cab Tractors

0%

0%

0%

6%

12%

25%

Heavy Haul Tractors

0%

0%

1%

1%

3%

5%

Opt

onal Custom Chassis: School Bus

13%

16%

19%

22%

31%

40%

Opt

onal Custom Chassis: Other Bus

0%

0%

13%

15%

23%

30%

Opt

onal Custom Chassis: Coach Bus

0%

0%

0%

0%

0%

0%

Opt

onal Custom Chassis: Refuse Hauler

0%

5%

10%

15%

16%

16%

Opt

onal Custom Chassis: Concrete Mixer

0%

0%

0%

0%

0%

0%

Opt

onal Custom Chassis: Motor Home

0%

0%

0%

0%

0%

0%

Opt

onal Custom Chassis: Mixed-Use Vehicle

0%

0%

0%

0%

0%

0%

Opt

onal Custom Chassis: Emergency Vehicle

0%

0%

0%

0%

0%

0%

We mapped every HD source type and regulatory class combination in MOVES to the Phase 3 regulatory
groupings, presented in Table 2-33. We applied the percent reductions from Table 2-32 in
EmissionRateAdjustment. Combination long-haul trucks (source type 62) are covered by the MHD All Cab
and HHD Day Cab Tractor category when they are Class 6 or Class 7 and the Sleeper Cab Tractors category
when they are Class 8.

Table 2-33 Phase 3 regulatory groupings mapped to MOVES vehicle types

Regulatory Grouping

Source Type ID(s)

Regulatory Class ID(s)

LHD Vocational

41, 42, 43, 51, 52, 53, 54

42

MHD Vocational

41, 42, 43, 51, 52, 53, 54

46

HHD Vocational

41, 42, 43, 51, 52, 53, 54

47, 48

MHD All Cab and HHD Day

61

46, 47

Cab Tractor

62

46

Sleeper Cab Tractors

62

47

We expect compliance with the Phase 3 standards to be achieved primarily through the adoption of
electric vehicles. Therefore, we added fleet averaging for heavy-duty energy consumption and C02
emission rates (discussed in the Emission Adjustments Report) and updated HD ZEV sales projections in
MOVES (discussed in the Vehicle Population and Activity Report). When the HD ZEV sales are combined
with the fleet averaging algorithm, the resulting change in MOVES output energy consumption and C02
emission rates for HD ICE vehicles in model year 2027 and beyond is small.

98


-------
2,1,4.3,2 LHD2b3

Energy reductions for LHD2b3 are modelled differently than for 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 C02 per mile or gallons of fuel per 100 miles. For simplicity, we apply the diesel chassis-certified
reductions to all LHD2b3 vehicles since most of the diesel LHD2b3 vehicles are chassis-certified/ The LHD
engine-certified vehicles are subject to the light-heavy duty reductions discussed in the previous section
that are applied to LHD45 vehicles. In addition, the fuel economy of medium-duty passenger vehicles
(MDPVs) are covered by the Light-duty GHG rule.63

Because MOVES includes energy rate measurements from LHD vehicles for model years 2014-2016, we
renormalized the Phase 1 reductions starting in MY 2017 so they could be applied to the MY 2014-2016
rates using Equation 2-33. Example calculations for LHD2b3 diesel in MY 2018-2020 are provided in
Equation 2-35.

Renormalized Phase 1 reductions for LHD2b3 in 2018 thru 2020 =

1 - 15%	1 - 15%	85%	,

= 1	=1	=1	= 1 - 88%) = 12%)	Equation

<2.3% + 3% + 6%j	1 - 3.8% 96.2%	2-35



Table 2-34 describes the expected changes in C02 emissions for diesel chassis-certified LHD2b3 vehicles
due to improved engine and vehicle technologies due to the HD GHG Phase 1 program. Note that the
impacts of the HD GHG Phase 1 program on gasoline LHD2b3 energy rates are discussed in Section
3.1.3.2.1. Since nearly all HD pickup trucks and vans will be certified on a chassis dynamometer, the C02
reductions for these vehicles are not treated as separate engine and road-load reduction components, but
represented as total vehicle C02 reductions and applied to all LHD2b3 vehicles in MOVES. MOVES models
the HD pickup truck/van standards by lowering the energy rates stored in the emissionRate table. No
change is made to the road-load coefficients or weights of passenger or light-duty truck source types.
Instead, the energy consumption rates for LHD2b3 were lowered by the percentages shown in Table 2-34
for the corresponding model years.

Table 2-34 Estimated Total Vehicle Reductions in Energy Consumption Rates for LHD2b3 Diesel Vehicles due to the HD

GHG Phase 1 Program

Regulatory
Class

Fuel

Model years

Reduction from
MY 2013 Energy
Rates

Renormalized Reductions Applied to
MY 2014-2016 Energy Rates
(running process)

LHD2b3

Diesel

2014

2.3%

-

2015

3%

-

2016

6%

-

2017

9%

5%

2018-2020

15%

12%

p As discussed in Section 1.4, engine-certified LHD2b3 vehicles are classified in MOVES as LHD45 vehicles for model
year 2017 and later.

99


-------
Table 2-35 shows the projected improvements in fleet average C02 emissions due to the HD GHG Phase 2
program for chassis-certified diesel and gasoline LHD2b3 vehicles. These reductions were applied in the
EmissionRate table for energy and the running process.

Note that we expect the Phase 2 requirements for LHD2b3 to be met via electrification starting in MY2025.
Projected electrification rates are described in the Population and Activity Report. The modeling of
electrification on diesel exhaust emissions due to fleet averaging is described in the Emission Adjustments
Report.

Table 2-35 Estimated Total Vehicle Reductions in Fleet Average Energy Consumption Rates for LHD2b3 due to the HD

GHG Phase 2 Program

REGULATORY CLASS

FUEL

MODEL YEARS

REDUCTION FROM MY
2020 EMISSION RATES

LHD2b3

Gasoline and

2021

2.50%



Diesel

2022

4.94%





2023

7.31%





2024

9.63%





2025

11.89%





2026

14.09%





2027+

16.24%

We do not model any change in medium-duty vehicle energy consumption rates as a result of the HD GHG
Phase 3 rule. Instead, they are covered by the LMDV20275 rule which is discussed in the GHG and Energy
Consumption Report.

2.1.4.4 Model Year Trends

Figure 2-51 and Figure 2-52 display the C02 (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 GHG and Energy Report. The C02 (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 C02
(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. Some of the small differences in C02 (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 as proportion of source types represented in each
regulatory class changes.

100


-------
1500

— 1000

ra
ce

CM

O
o

500

1980

2000	2020

Model Year

2040

Reg Class

41-LHD2b3
— 42-LHD45

46-MHD67

47-HHD8

48-Urban	Bus

49-Gliders

Figure 2-51. Base running emission rates for C03from age 0-3 diesel heavy-duty vehicles averaged over a nationally

representative operating mode distribution.

20

15

c

o

ra
O

at
CL

0)

I

10

0



J

Hu

p

jiv-





J



¥

y









	













Reg Class

-+¦ 41-LHD2b3
42-LHD45

46-MHD67

47-HHD8

48-Urban	Bus

49-Gliders

1980	2000	2020	2040

Model Year

Figure 2-52. Fuel economy for age 0-3 diesel heavy-duty vehicles averaged over a nationally representative operating

mode distribution

101


-------
2.1.5 Evaluation of Fleet-average Running Rates with Real-World Measurements
As one evaluation of the MOVES diesel exhaust running rates, Table 2-36 compares preliminary MOVES3
emission rates estimated in fuel-specific units (g/kg-fuel) to fuel-specific emission rates estimated from a
remote sensing and tunnel measurements in calendar years 2015-2017.

Haugen et al. (2018) conducted exhaust plume measurements from 1,844 in-use heavy-duty diesel trucks
at the Peralta weigh station near Anaheim, CA in 2017, of which over 63% of the fleet were model year
2011 or later. Wang et al. (2019), conduced sampling of the Ft. McHenry Tunnel in Baltimore, MD during
winter and summer of 2015. The model year distribution of the Ft. McHenry diesel fleet was not measured.
Wang et al. (2018) estimated the heavy-duty emission factors separately from the light-duty vehicles using
a linear regression which accounted for the fraction of the fleet is composed of heavy-duty vehicles. The
emission rates from both studies are compared to MOVES emission rates estimated from a national scale
run with a preliminary version of MOVES3 conducted for calendar year 2016 for all on-road heavy-duty
diesel vehicles. No effort was made to match the vehicle operation of the studies or to match the fleet and
fuel characteristics (model year distribution, regulatory class distribution). As such, the comparison is only
intended to be a rough comparison, to assure that MOVES provide estimates that are in the range of
feasible values measured from in-use fleets.

Table 2-36 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. PM2.5 is lower than the Ft. McHenry estimates, but
within the standard error of the winter measurements, and close to the 95% confidence range of the
summer measurements (approximately two times the standard error). Given the expected differences in
vehicle operation and fleet composition, the comparisons increased our confidence that MOVES is
estimating representative in-use running emission factors for heavy-duty diesel vehicles.

Table 2-36. Comparison of MOVES Emissions with Remote Sensing and Tunnel Measurements (g/kg-fuelj



Peralta CA 2017
HDV (Haugen et
al. 2018)

Winter Ft.
McHenry MD
2015 (Wang
et al. 2019)

Summer Ft.
McHenry MD
2015 (Wang
et al. 2019)

MOVES3
National
Heavy-duty
Fleet 2016

THC

2.2 ±0.4

NA

NA

0.68

CO

5.9 ±0.9

4.6 ± 2.0

7.5 ± 2.6

5.2

NOx

12.4 ±0.6

29.6 ±4.7

17.9 ± 1.4

12.2

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

102


-------
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 1.2, we define eight stages which are differentiated by soak time length (time
duration between engine key off and engine key on) between a cold start (> 720 minutes of soak time) and
a hot start FTP (< 6 minutes of soak time). More details on how start emission rates are calculated as a
function of soak time, can be found later in this section and in the MOVES light-duty exhaust emission rate
report.12 The impact of ambient temperature on cold starts is discussed in the Emission Adjustments
Report.

The next subsections discuss the derivation of heavy-duty diesel start emissions by pollutant and model
year group. Start emissions are currently a small contributor to total exhaust emissions from heavy-duty
diesel vehicles. No T&M or other age effects are currently applied to the diesel start emissions.

2.2.1 THC, CO, and NOx

The pre-2010 model year emissions are discussed in Section 2.2.1.1 and 2010+ model year emission rates
are discussed in Sections 2.2.1.2.

2,2,1,1 1950-2009 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 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 2-37.

Table 2-37 Average Start Emissions Increases (g/start) for pre-2010 Model Year Light Heavy-Duty Diesel Vehicles for

Regulatory Class LHD2b3 and LHD45 (regClassID 41 and 42)

THC

CO

NOx

0.13

1.38

1.68

For pre-2010 model year HHD and MHD trucks, analogous data were unavailable. To provide at least a
minimal amount of information, we measured emissions from a 2007 Cummins ISB which is used in both
LHD and MHD vehicles on an engine dynamometer at the EPA National Vehicle and Fuel Emissions
Laboratory in Ann Arbor, Michigan. Among other idle tests, we performed a cold start idle test at 1,100
RPM lasting four hours, long enough for the engine to warm up. Essentially, the "drive cycle" we used to
compare cold start and warm emissions was the idle cycle, analogous to the FTP and ST01 cycles used for
LHD vehicles. Emissions and temperature stabilized about 25 minutes into the test. The emission rates

Q 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.

103


-------
through time are shown in Figure 2-53. 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.

time [hrs]

Figure 2-53 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)

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 difference between
cold start and warm idle over a 25-minute period of the elevated cold starts as shown in Equation 2-36.

Grams per Start =

= stabilization time x (cold start average rate	— hot running average rate)

(v-1 emissions	v-1 emissions\	Equation 2-36

t=0	t=25	/

The results are shown in Table 2-38. The measured THC increment is zero. The NOx increment is negative
since cold start emissions were lower than warm idle emissions.

Table 2-38. Cold-start Emissions Increases (g/start) in Grams on the 2007 Cummins ISB

THC

CO

NOx

0.0

16.0

-2.3

We also considered NOx data from University of Tennessee,64 which tested 24 trucks with PEMS at
different load levels during idling. Each truck was tested with a cold start going into low-RPM idle with air-
conditioning on. We again used Equation 2-36 to integrated the emissions over the warm-up period to get
the total cold start idling emissions. We calculated the warm idling emissions by multiplying the reported
warm idling rate by the stabilization time. We used the stabilization period from our engine dynamometer
tests (25 minutes). Then, we subtracted the cold start-idle emissions from the warm idle emissions to

104


-------
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 2-39 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 2-39. MOVES Inputs for Pre-2010 HHD and MHD Diesel Start Emissions (grams/start) for Regulatory Class 46,

47, and 48. No Differentiation by Model Year or Age.

THC

CO

NOx

0.0

16.0

0.0

As discussed in the Emission Adjustments Report, 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 applied to CO, PM2.5, or NOx diesel
start emissions because no clear trend was found with the data.

2,2,1,2 2010-2026 Model Years

The cold start emissions for 2010 model year and later LHD, MHD, and HHD diesel engines were updated
for MOVES3 based on new data. However, because of the small sample size and lack of real-world data,
notable uncertainty about real world heavy-duty diesel start emissions remains.

Similar to the approach taken for light-duty vehicles, the cold start emissions are defined as the difference
in emissions between a test cycle with a cold start and the same test cycle with a hot start. Heavy-duty
diesel engines are certified using the Heavy-Duty Diesel Engine Federal Test Procedure (FTP) cycle65. The
test procedure for certification requires that manufacturers run the engine over the FTP cycle with a cold
start and then repeat the cycle with a warm start. Starting in model year 2016, EPA began collecting
certification data that contained separate cold and hot results for each engine certified. The data included
the following engine families from 2016 and 2017 model years shown in Table 2-40.

Table 2-40 Engine Data Analyzed to Estimate the Cold Start Emission Rates for HD Diesel Engines

Category

Number of Engines

Manufacturers

LHD

5

Ford, Isuzu, Hino, FPT

MHD

6

Ford, Hino, Cummins, Detroit Diesel

HHD

11

Cummins, PACCAR, Detroit Diesel,
Volvo, Hino

The certification data was used to determine the grams emitted per cold start using Equation 2-37.

105


-------
Grams per Start

= [Cold FTP Emission Results (g/(hp — hr))

— Hot FTP Emission Results (g/(hp — hr))]	Equation 2-37

* FTP Cycle Work (hp — hr)

The amount of work (hp-hr) performed over the FTP cycle is required to convert the FTP emission results in
grams per horsepower-hour into grams, but it is not provided as part of the certification data submitted by
the manufacturers to EPA. Furthermore, the FTP cycle work is unique to each engine and is generally
calculated based on the engine's maximum speed, curb idle speed, and the maximum torque curve.
Therefore, we needed to develop a surrogate from the information that is provided by manufacturers for
certification for each engine. We determined that the rated power of an engine correlates well to the FTP
cycle work. This analysis was based on FTP cycle work and rated power data from ten HD engines. As
shown in Figure 2-54, 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 2-54 - 0.0599 (hr) times the
rated power (hp) plus 4.4297 (hp-hr).

40

35

130

!-
20
15

250	300	350	400	450	500	550

Rated Power (hp)

Figure 2-54: Relationship between HD Diesel Engine Rated Power and FTP Cycle Work
2.2.1.2.1 Heavy-heavy duty

Analysis of cold and hot start FTP certification data from eleven HHD diesel engines determined the grams
per start for THC, CO, NOx, and PM2.5. The average and standard deviation of the THC, CO, and NOx
emission levels of the eleven engines are shown in Table 2-41. The PM2.5 emissions are summarized in
Table 2-47. The sample included both MY2016 and MY2017 engines, ranging in displacement between 7.7
and 14.9 liters, and in rated power between 260 and 605 HP. The default cold start emissions values in
MOVES are the mean values shown in the table.

FTP workvs Rated Power - Diesel

V = 0.0559x +4.4297









	RJ

= 0.8503
•









1

*





1*

	















*









106


-------
Table 2-41: Cold Start Emissions for MY 2010-2026 Heavy Heavy-Duty Diesel Engines

Grams per Start

THC

CO

NOx

Mean

0.08

6.6

8.4

Standard Deviation of Data

0.1

5.6

1.7

2,2,1,2,2 Medium-heavy duty

The certification data from six MHD diesel engines were used to develop the THC, CO, and NOx grams
emitted per start. The average and standard deviation of the emissions from the six engines are shown in
Table 2-42. The sample included MY2016 and MY2017 engines, ranging in displacement between 5.1 and
8.9 liters, and in rated power between 230 and 380 HP. The default values in MOVES are the mean values
shown in the table.

Table 2-42 Cold Start Emissions for MY 2010-2026 Medium Heavy-Duty Diesel Engines

Grams per Start

THC

CO

NOx

Mean

0.20

2.5

6.4

Standard Deviation of Data

0.2

2.7

1.8

2,2,1,2,3 Light-heavy duty

Analysis of five LHD diesel engines from the certification data determined the grams per start for THC, CO,
and NOx shown in Table 2-43. 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 in MOVES are
the mean values.

Table 2-43 Cold Start Emissions for MY 2010-2026 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 apply the cold start THC, CO, and NOx emission rates from the 2016 MY and 2017 MY engines to all
engines MY 2010 through 2026. They are subject to the same HD diesel emission standards, and the
aftertreatment systems on these engines are similar and generally include both a diesel particulate filter
and selective catalytic reduction system.

2,2,1,2,4 Incorporation of Tier 3 Standards for Light Heavy-Duty Diesel

The Tier 3 exhaust emission standards affect light heavy-duty diesel vehicles in the LHD2b3 regulatory class
(regClassID 41). Reductions are applied to start rates for NOx only, phasing in from MY2018 to MY2021 as
previously described for running emissions in Section 2.1.1.5.5. No reductions applies to THC and CO rates.

107


-------
2,2,1,3 2027-2060 Model Years

Start emissions for THC and CO are the same for model years 2010 to 2060, but as noted in Section 2.1.1.6,
the HD2027 NOx standards include duty-cycle standards, off-cycle standards and changes to warranty and
useful life requirements. To account for the HD2027 standards, we updated NOx start emission rates to
reflect the changes in the duty-cycle standards using the method described in this section.

We did not estimate the impact of the off-cycle standard on start emissions, in part because the baseline
MY 2010 and later start emission rates in MOVES are not based on in-use data but are based on emissions
data from the FTP duty-cycle. Additionally, because the heavy-duty diesel start emission rates in MOVES do
not vary with age due to insufficient data, we did not estimate changes due to the changes in warranty and
useful life.

Because engines meeting the HD2027 standard are not yet in production, to update the NOx start emission
rates for MY2027+, we estimated the NOx cold start emission rate (g/start) from a CARB Stage 1 HDD
engine66 tested on the FTP duty-cycle cycle after different periods of use (aging). Table 2-44 contains the
NOx Cold and Hot FTP measurements in Columns (B) and (C) for different aging periods. Column (E), "Cold -
Hot," is calculated as the difference between Columns (B) and (C). The cold start, Column (F), is then
calculated by multiplying the difference in Column (E) by the work performed on the FTP cycle, Column (D),
as shown in Equation 2-38.

N0X Cold Start f—) = [cold (¦—) — Hot.,

Vstart/ L Vhp ¦ hr/ Vhp ¦ hr

g

X FTP work (hp ¦ hr) Equation 2-38

Table 2-44 Calculation of NOx 12-hour Cold Starts from the CARB Stage 1 HHD Engine from the Cold and Hot FTP Cycle

Aged hours

(A)

(B)

(C)

(E)

(D)

(F)

FTP composite
(g/hp-hr)

Cold

(g/hp-hr)

Hot

(g/hp-hr)

Cold - Hot
(g/hp-hr)

FTP Work
(hp-hr)

Cold Start
(g/start)

0

0.008

0.025

0.005

0.02

31.4

0.63

333

0.012

0.042

0.006

0.036

31.4

1.13

656

0.018

0.061

0.009

0.052

31.4

1.64

1000

0.024

0.092

0.01

0.082

31.4

2.58

1000 hr Post Ash
Clean

0.026

0.109

0.009

0.1

31.4

3.14

The Stage 1 HHD engine was deemed representative of an engine-certified to a 0.02 g/hp-hr NOx standard
based on the FTP composite measurements in Column (A). Table 2-44 demonstrates that the larger cold
start measured with increased aged hours, and after the DPF ash clean out at 1000 hours. We used the
1000 hr, Post Ash Clean cold start emission rate (3.14 g/start shown in Table 2-44) to represent the 12-
hour cold-start (operating mode 108) emission rate.

To estimate the 12-hour cold-start NOx emission rate for HHD diesel vehicles subject to the HD2027
standards, we interpolated the HHD 12-hour cold-start between the Stage 1 cold start (3.14 g/start) and
the MOVES baseline (MY2010-2026) 12-hour cold-start (8.4 g/start), and their respective FTP duty-cycle
standards using Equation 2-39 as shown in Figure 2-55 and Table 2-45. For example, the interpolation
yielded an estimated 12-hour cold start of 4.02 g/start for the 0.05 g/hp-hr FTP standard.

108


-------
Start ERFTPx HHD12 hour

/MOVES start HHD12 hour - Stagel startx

V Baseline FTP — Stagel FTP /
x (FTPX — Baseline FTP) + MOVES start HHD12 hour

Equation 2-39

Where:

•	Start ERFTPx HHD12 hour = the estimated NOx start emissions for an FTP duty-cycle
standard, x, for heavy heavy-duty diesel emissions for a 12-hour cold-start (operating
mode 108).

•	Stagel start = 1000 Post Ash Clean start emission rate from the CARB Stage 1 HHD diesel
engine = 3.14 g/start (Table 2-44)

•	Stagel FTP = Composite FTP level of the CARB Stage 1 engine = 0.02 g/hp-hr

•	MOVES start HHD12 hour= MOVES3 baseline start emission rate (= 8.4 g/start) for MY
2027 heavy heavy-duty diesel engine for a 12-hour soak (operating mode 108)

•	Baseline FTP =	baseline FTP composite NOx standard = 0.2 g/hp-hr

•	FTPX = composite FTP standard in the HD2027 standards

9

_ 8

i—

OJ 7

4-> /

356

4—•
i_

(TJ r

do 5

"O

o 4

U

^ 3
o

^ 2

r^j

T—1

Q 1
X

X 0

c













































































































) 0.05 0.1 0.15 0.2 0.25
FTP Composite Standard (g/hp-hr)

Figure 2-55 Calculated relationship between the HHD NOx 12-hour cold-start and the composite FTP NOx standards
Table 2-45 HHD Cold Start Emissions for Baseline and HD2027 Standards

Scenario

Applicable Model
Years

Weighted Average
FTP standard
(g/hp-hr)

Cold Start
emissions (g/start)

Baseline

Model Year 2010-2026

0.2

8.40

HD2027 Standards

Model Year 2027+

0.05

4.02

109


-------
We assumed that the relative difference in cold start emission rates by regulatory class is the same in the
baseline and HD2027 standards. This calculation was combined with the estimate of emissions by start
operating mode and is described in Equation 2-40 in Section 2.2.3.

2.2.1.4 Model Year Summary

Figure 2-56 through Figure 2-58 display the cold start (operating mode 108) emission rates across model
years for heavy-duty diesel vehicles. The figures show the large difference in start emission rates before
and after model year 2010. Model year 2010 corresponds to the implementation of (SCR) aftertreatment,
as well as the different datasets and methodologies. The rates for LHD2b3 are lower starting in MY2018
due to the phase-in of Tier 3 standards. The HD NOx emission rates change in MY2027 due to the HD2027
rule.

0.20'

1980

2000	2020

Model Year

2040

Reg Class

-+¦ 41-LHD2b3
— 42-LHD45

46-MHD67

47-HHD8

48-Urban	Bus

49-Gliders

Figure 2-56 Heavy-duty Diesel THC Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class and

Model Year

110


-------
15

B 10'

1980

2000	2020

Model Year

2040

Reg

Class

41-LHD2b3

42-LHD45

46-MHD67

47-HHD8

48-Urban	Bus

49-Gliders

Figure 2-57 Heavy-duty Diesel CO Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class and Model

Year

\

\

Reg Class

—	41-LHD2b3

—	42-LHD45

46-MHD67

47-HHD8

48-Urban	Bus

49-Gliders

1980	2000	2020	2040

Model Year

Figure 2-58 Heavy-duty Diesel NOx Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class and

Model Year

111


-------
2.2.2 Particulate Matter (PM2.5)

2,2,2,1 1950-2010 Model 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 2-46.

Table 2-46 Average PM2.5 emissions (grams) from MY2004 HHD 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 1950 through 2006 model year vehicles. For 2007 through 2009 model
years, we applied a 90 percent reduction to account for the expected use of DPFs, leading to a
corresponding value of 0.011 g/start. The value is the same for all heavy-duty diesel regulatory classes.

As introduced in Section 2.1.2.1.8, in MOVES, the PM2.5 emission rates are estimated as the elemental
carbon (EC) and non-elemental carbon PM (nonEC). We estimated the EC and nonEC from the total PM2.5
starts rates by applying the EC/PM fraction of 46.4 percent from the PM2.5 speciation profile developed
from the idle mode of the UDDS tests from the E55/59 program for pre-2007 trucks.6 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.6

2,2,2,2 2010-2060 Model Years

The cold start emissions for 2010 model year and later LHD, MHD, and HHD diesel engines were updated in
MOVES3 based on new data. We updated the cold start particulate matter emission rates based on the
certification data and data analysis methods discussed in Section 2.2.1.2. The resulting cold start emission
rates for each HD diesel engine regulatory group are shown in Table 2-47. 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 2-47 is of similar magnitude to that of HHD and MHD).
Instead of using the certification test data, we used the data from MHD diesel to represent the LHD diesel
PM2.5 emission rate in MOVES, because of the overlap in engines and aftertreatment systems between the
two categories.

112


-------
Table 2-47: 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



a Instead of using the test data, we used the data from MHD diesel to represent the LHD
diesel PM2.5 emission rate in MOVES as noted above.

We are applying the new cold start PM2.5 emission rates from the model year 2016 and 2017 engines to MY
2010 and newer engines because the PM standards are the same and all the MY 2010 and later engines
generally include both a diesel particulate filter (DPF) and selective catalytic reduction (SCR) system.

2.2.2.3 Model Year Summary

Figure 2-58 and Figure 2-59 display the cold start (operating mode 108) emission rates across model years
for heavy-duty diesel vehicles. As expected, large reductions are shown in model year 2007 with the
implementation of diesel particulate filters. Further changes are due to the incorporation of the 2010 and
later certification data.

^ 0.0H
3

CO
D>

C3
Q1

1	0.06'
+¦»

if)

2

o

o

7Z 0 03'

re

0.00'

Reg Class

41-LHD2b3

42-LHD45
46-MHD67

-+¦ 47-HHD8

48-Urban	Bus

49-Gliders

1980	2000	2020	2040

Model Year

Figure 2-59 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 Adjusting Start Rates for Soak Time

The discussion to this point has concerned the development of rates for cold start emissions from heavy-
duty diesel vehicles. In addition, it was necessary to derive rates for additional operating modes that

113


-------
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 opModelD=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.

2.2.3.1 Adjusting Start Rates for Soak Time - MY 2009 and Earlier

The soak adjustment ratios we used for THC, CO, and NOx for MY 2009 and older HD diesel vehicles are
illustrated in Figure 2-60 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.12 The soak
adjustments are taken from the non-catalyst soak adjustments derived in a CARB report67 and reproduced
in a MOBILE6 report.68

Soak Time (minutes)

Figure 2-60. 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 Report12

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 2-48 for THC, CO, and NOx.

114


-------
Table 2-43. Heavy-Duty diesel THC, CO, and N0X Start Emissions (g/start) by Operating Mode for 2003 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

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 2-49 below. They are estimated by assuming a
linear decrease in emissions with time between a full cold start (>720 minutes) and zero emissions at a
short soak time (< 6 minutes).

Table 2-49. Particulate Matter Start Emission Rates (g/startj by Operating Mode (soak fraction) for all HD Diesel

vehicles through MY 2003

Operating
Mode

1950-2006
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

2.23.2 Adjusting Start Rates for Soak Time - MY 2010 and Later

As described in the preceding section, the start rates are based on data collected from light-duty vehicles
in the 1990's. The question arose as to whether they could be considered applicable to heavy-duty diesel
vehicles with aftertreatment systems designed to meet the 2007/2010 exhaust emissions standards. To

115


-------
address this question, we initiated a research program in 2016, with the goal of examining the
relationships between soak time and start emissions for a set of heavy-duty vehicles. Two test programs
were conducted to revise the 2010 MY and later soak curves for heavy-duty diesel vehicles in MOVES3 and
later versions. The testing consisted of both chassis and onroad testing of MY 2015 and MY 2016 vehicles.

The first test program included a MY 2015 day-cab tractor with a MY 2015 HHD diesel engine tested on a
heavy-duty chassis.69 The vehicle was relatively new and had 10,000 miles on the odometer. The testing
consisted of running two repeats of a transient drive cycle developed by the National Renewable Energy
Laboratory (NREL). The vehicle speed trace is shown below in Figure 2-61. 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 2-61. At least two repeats were conducted for each
soak period. The emission measurements included dilute gaseous measurements and triplicate particulate
matter filters.

Double NREL Cycle

60

Time (seconds)

Figure 2-61 National Renewable Energy Laboratory's Heavy-Duty Vocational Transient Cycle

The NOx, CO, THC, and PM2.s emission results in terms of grams or mg per mile from the tests over a range
of soak periods are shown in Figure 2-62 through Figure 2-65.

116


-------
8.0
7.0
T 6.0

4 50

£ 4.0

1 30

iS 2.0

1.0
0.0

c

NOx





























»



»









•

•



*















*







































\



















1







































ID 20 30 40 50 60 70 SO 90 100

SoakTime (hours)

Figure 2-62 MY 2015 Heavy-Duty Vehicle N0X Emissions by Soak Time

0.06
0.05
~ 0.04

S

£ 0.03
o

i/i

- 0.02

UJ

0.01
0.00

HC











»















t

$









•







*
*



•













•

•

• *



















•1

. *



















!•
t



















10 20 30 40 50 60 70 BD 90 100

Soak Time (hours)

Figure 2-63 MY 2015 Heavy-Duty Vehicle THC Emissions by Soak Time


-------
0.50
0.45
	0.40

OJ

~ 0.55

3 °-30

£ 0.25
"5 0-20
¦j= 0.15
m 0.10
0.05 J

0.00 '
t









CO

































•



*





•	





A t

1



















<



















•























•

















•



















t;





















J.





















¥





















) 10 20 30 40 50 60 70 SD 90 100

Soak Time (hours)

Figure 2-64 MY 2015 Heavy-Duty Vehicle CO Emissions by Soak Time

9.0 ,

s.o

"7.0

QJ

1 6'°

E 5.0

I"

! 5»

^ 2.0 \
i.O

o.o 1
c









PM























•





































































t.





















L «



•

1



















/

*













• •







































»

t 10 20 30 40 50 60 70 BO 90 1[

SoakTime (hours)

Figure 2-65 MY 2015 Heavy-Duty Vehicle PM2.s 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 2-66. The route

118


-------
consisted of approximately 700 seconds of driving in a neighborhood/urban environment over
approximately 2.7 miles.

Soak Route Speed Profile

400

Trip Counter -iMtofldi]

Figure 2-66 Onroad Soak Drive Route

The emission results, in terms of total emissions over the route, from the onroad tests are shown in Figure
2-67 through Figure 2-69.

r n

Total NOx Over Cycle





2.5

JS

£ 2.0
o

i/i

«" 1.5
E

UJ

ra 1.0
£

0.5
0.0

(









•

























•

•

m





•

r

•



•

w





























0

5 10 15 20 25 3
Soak Time (hours)

Figure 2-67 MY 2.016 Heavy-Duty Vehicle NOx Emissions by Soak Time

119


-------
0.8

0.7

"bD 0.6

l/l

J 0-5

1/1

Ln 0 4

E

^ 0.3

ra

£ 0.2

0.1

0.0 '
t



Total HC Over Cycle























































•





















•



••

• •





• * .

A













f











0

) 5 10 15 20 25 31

Soak Time (hours)

Figure 2-68 MY 2016 Heavy-Duty Vehicle THC Emissions by Soak Time

a n

Total CO Over Cycle





8.0

— 7.0

M

£ 6.0
o

S 5.0

1/1

E 4.0

LU

75 3.0

4—1

o

t- 2.0
1.0
0.0

t









•























a













w

•

~











a





m «

#



w







A













f













m











0

5 10 15 20 25 3
SoakTime (hours)

Figure 2-69 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 multiplied by the ratio of 3 minutes divided by 18

120


-------
minutes (the midpoint times of operating mode 101 and 102). The NOx, CO, and THC soak period ratio
results for each vehicle are shown below in Figure 2-70.

HD Diesel Truck Start Emission Ratio

1.20

1.00

.2 0.80

*5

cc

o 0.60

LO

cn

£ 0.40

0.20

0.00















a















?













?_



















/•'



mM	-V







4

/















100 200 300 400 500
Soak Time (min)

600

700

800

HD Van NOx
—•— Day Cab NOx

HD Van CO
— •— Day Cab CO

—•¦¦¦ HD Van HC
--•--Day Cab HC

Figure 2-70 Soak Emission Ratios from a MY 2015 HD Day-Cab and a MY 2016 HD Van

The 2010 MY and later heavy-duty diesel soak ratios for MOVES were determined by averaging the results
from the two trucks. The resulting soak adjustment ratios are shown in Table 2-50. 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.

Table 2-50 HD 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

121


-------
For MY2027+ vehicles subject to HD2027 standards, we used Table 2-42 to estimate the MOVES NOx
emission rates for each MOVES heavy-duty regulatory class (LHD45, MHD, and HHD), and for each MOVES
start operating mode classified by different soak times. We assumed that the relative difference in
emission rates by regulatory class and by operating mode is the same for MY2010-2026 and MY2027 and
later meeting the HD2027 standards.

Stait ERpTP=x,reg class=y,soak=z

= Start ERr

^¦Duty cycle standard x,HHD, 12

/MOVES staitregciass=y soak=z
\ MOVES start HHD 12_hour

Equation 2-40

Where:

•	Start ERftp= the start NOx emission rates for the HD2027 standards with FTP x (0.035 or
0.05) for regulatory class y (LHD45, MHD, and HHD), and soak length z

•	Start ERDuty cycie standard x,hhd, 12-hour = the estimated start emissions for an FTP duty-
cycle standard, x, for heavy heavy-duty diesel emissions for a 12-hour soak (operating
mode 108)

•	MOVES start regclass=y soak=z = MOVES3 baseline start emission rate for MY 2027 for
regulatory class y (LHD45, MHD, and HHD), and soak length z

•	MOVES start hhd,i2-hour= MOVES3 baseline start emission rate for MY 2027 HHD diesel
engine for a 12-hour soak (operating mode 108)

For example, Figure 2-71 compares the estimated MOVES NOx start emission rates for HHD diesel vehicles
for MY2010-2026 and MY2027 and later.

i Without HD2027

HD2027 Standards Rate for MY2027+

-e

03

-i—¦

V)
U)

O

8
7
6
5
4
3
2
1
0

l

i

1

<6min 6-30 min 30-60 min 60-90 min 90-120 min 120-360 360-720 720+min

min	min

MOVES Start Operating Mode
Figure 2-71 Duty-cycle-based NOx start emissions for HHD Diesel comparing MY2010-2026 and MY2027+ rates.

122


-------
The PM2.5 start rates by operating mode for MY 2010 and newer vehicles are presented in Table 2-51
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 2-51 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). The ambient temperature effects for starts with warm and
hot soaks (operating mode 101-107) are documented below and recorded in the MOVES
startTempAdjustment table.

Because the THC temperature effects in MOVES are modeled as additive adjustments, the adjustment
calculated for cold starts needs to be reduced for warm and hot starts. Due to lack of data, we multiply the
soak fractions described earlier in Figure 2-60 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.R The additive cold start
adjustment for THC emission factors are displayed in Table 2-52, 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.

R The temperature effects from pre-2010 technology engines are applied to all model years. We hope 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.

123


-------
Table 2-52 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

-0.0211x(Temp - 75)

0.52

106

-0.0254x(Temp - 75)

0.62

107

-0.0349x(Temp - 75)

0.86

108

-0.0406x(Temp - 75)

1

2.2.4 Start Energy Rates

The start energy rates (in units of kJ) were developed for MOVES200471, and updated in MOVES2010 as
documented in the MOVES2010a energy updates report.57 Figure 2-72 displays the cold starts in grams of
C02 emissions calculated from the energy rates using the carbon content for conventional diesel fuel as
documented in the MOVES3 Greenhouse Gas and Energy Report.

As shown, there is more detail in the pre-2000 energy rates. The spike in C02 g/startfor model years 1984-
1985 reflects variability in the data used to derive starts, which was consistent with the more detailed
approach used to derive the pre-2000 energy rates in MOVES2004. The only updates to the start energy
rates post-2000 is the impact of the Phase 1 Heavy-Duty GHG standards, which began phase-in in 2014 and
have the same reductions as the running energy rates as presented in Table 2-30 and Table 2-34. 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).

124


-------
1970	1990	2010	2030	2050

Model Year

Figure 2-72 Heavy-Duty Diesel Cold Start C02 Rates (Operating Mode 108) by Model Year and Regulatory Class

r

re
*-»

w

s


-------
The energy rates for heavy-duty starts have not been updated due to relatively small contribution the
starts have to the energy inventory. Table 2-54 displays the relative contribution of total energy
consumption estimated from a national run of MOVES for calendar year 2016, using a draft version of
MOVES3 developed for the proposed Clean Trucks Rule.72 As shown, the estimated energy consumed due
to starts is very small in comparison to the energy use of running activity.

Table 2-54. Relative contribution of total energy consumption from each pollutant process by regulatory class for

heavy-duty diesel vehicles in calendar year 2016

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
hours), typically overnight, including higher engine speed settings and extensive use of accessories by the
vehicle operator. Extended idling most often occurs during rest periods by long-haul trucking operators
where the truck is used as a residence (sometimes referred to as "hotelling"). Operators idle to power
accessories such as air conditioning systems or heating systems. Heavy-duty engine and truck
manufacturers recommend trucks not idle at low engine speeds for extended periods, because it can
"create engine wear and carbon soot buildup in the engine and components."73 Additionally, idling for
extended periods allows the vehicle's exhaust to cool below the effective temperature required for
emission aftertreatment systems in modern trucks such as selective reduction catalysts and diesel
oxidation catalysts. As a result, extended idle is treated as a separate emission process in MOVES which
uses a different emission rate than the idling that occurs during the running emission process.

Extended idling does not include vehicle idle operation that occurs during normal road operation, such as
idling at a traffic signal or the "off-network" idle that might occur during a delivery. Although frequent
stops and idling can contribute to overall emissions, these modes are included in the normal vehicle hours
of operation. Extended idling is characterized by idling periods that last hours rather than minutes.

In the MOVES model, long-haul combination trucks (sourceTypelD 62) is the only source type assumed to
have extended idling activity. These trucks are only associated with MHD, HHD and Glider1 regulatory
classes. As an alternative to extended idling, long-haul truck operators can also use auxiliary power units
(APUs) or plug into facility's power (shore power) to power their cabin and accessories during hotelling.

s The default hotelling activity in MOVES3 is estimated from telematics data in which all idle events with duration
greater than one hr from long-haul combination trucks are assigned to extended idling.

T Glider extended idle emission rates are documented in Section 2.5.

126


-------
The emission rates for auxiliary power units (APUs) are discussed in Section 2.4, and the energy
consumption rates for shore power are discussed in the Greenhouse Gas and Energy Consumption
technical report.

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 1950-2006 Model Years

The MOVES extended idling emission rates for pre-2007 model years were derived from data collected in
several distinct test programs under different types of idle conditions. For MOVES3 and later versions,
weightings were adjusted from those in previous versions of MOVES to better account for new information
on typical extended idling engine idling speeds and loads. These adjustments are described below in
Section 2.3.1.2. Appendix D summarizes the data and calculations for the pre-2007 model years.

2,3,1,1 Data Sources

The references included in this section provide more detailed descriptions of the data and how the data
were obtained:

•	Testing was conducted on 12 heavy-duty diesel trucks and 12 transit buses in Colorado by
McCormick et al.74 Ten of the trucks were Class 8 heavy-duty semi-tractors, one was a
Class 7 truck, and one of the vehicles was a school bus. The school bus data was not used
to calculate extended idle rates. The model years ranged from 1990 through 1998. Typical
Denver area wintertime diesel fuel was used in all tests. Idle measurements were collected
during a 20-minute time period. All testing was done at 1,609 meters above sea level (high
altitude).

•	Testing was conducted by EPA on five trucks in May 2002 (Lim et al.).75 The model years
ranged from 1985 through 2001. The vehicles were put through a battery of tests including
a variety of idling conditions.

•	A total of 63 trucks (nine in Tennessee, 12 in New York and 42 in California) were tested
over a battery of idle test conditions including with and without air conditioning (Irick et
al.).76 Not all trucks were tested under all conditions. Only results from the testing in
Tennessee and New York are described in the IdleAire report (Irick et al.).76

•	The California test data was collected on 42 diesel trucks in parallel with roadside smoke
opacity testing (Lambert)77. All tests conducted by the California Air Resources Board
(CARB) at a rest area near Tulare, California in April 2002 are described in the Lambert77
Clean Air Study. All analytical equipment for all testing at all locations was operated by
Clean Air Technologies.

•	Fourteen trucks were tested as part of the E-55/59 Coordinating Research Council (CRC)
study of heavy-duty diesel trucks with idling times either 900 or 1,800 seconds long.78

•	The National Cooperative Highway Research Program (NCHRP)79 obtained the idling
portion of continuous sampling during transient testing to determine idling emission rates
on two trucks.

127


-------
•	A total of 33 heavy-duty diesel trucks were tested in an internal study by the City of New
York (Tang et al.)80. The model years ranged from 1984 through 1999. One hundred
seconds of idling were added at the end of the WVU five-mile transient test driving cycle.

•	A Class 8 Freightliner Century with a 1999 engine was tested using EPA's onroad emissions
testing trailer based in Research Triangle Park, North Carolina (Brodrick).81 Both short (10
minute) and longer (five hour) measurements were made during idling. Some testing was
also done on three older trucks.

•	Five heavy-duty trucks were tested for particulate and NOx emissions under a variety of
conditions at Oak Ridge Laboratories (Storey et al.).82 These are the same trucks used in
the EPA study (Lim et al.).

•	The University of Tennessee (Calcagno et al.) tested 24 1992 through 2006 model year
heavy-duty diesel trucks using a variety of idling conditions including variations of engine
idle speed and load (air conditioning).64

2,3,1,2 Analysis

We used the data sources referenced above to estimate the emission rates for particulate matter (PM2.5),
oxides of nitrogen (NOx), hydrocarbons (THC), carbon monoxide (CO) and carbon dioxide (C02). The data
were grouped by truck and bus and by idle speed and accessory usage to develop emission rates
representative of extended idle emissions.

The important conclusion from the analysis was that truck operator behavior plays an important role when
assigning emission rates to periods of extended idling. Factors such as accessory use and engine idle speed,
which are controlled by operators, affect engine load and emission rates during extended idling. The
impacts of other factors, such as engine size, altitude, model year within MOVES groups, and test cycle are
negligible.

We first evaluated the studies on engine idle speed. NREL's review of owner's manuals found that several
heavy-duty engine manufacturers recommend use of fast idle (> 1000 rpm) if the engine needs to idle for
extended periods.83 In a 2004 UC-Davis survey (Lutsey et al. 2004), respondents' average engine idle speed
was 866 rpm, with small peaks around 650 and 1000 rpm.84 About one-third of the respondents indicated
they changed their idle speed from its usual setting, which is consistent with the distribution of the
responses where about one-third of the idle engine speeds reported were 1000 rpm or faster. A 2015
study by Hoekzema (2015)85 suggested that even fewer trucks operated in a high idle condition. Drivers
surveyed for this study reported high idle operation (> 1000 rpm) just 18 percent of the time during idling
periods of an hour or more. Additionally, Hoekzema (2015) cited similar studies representing 764 trucks
that averaged engine speeds of 886 rpm during extended idle. Therefore, in MOVES3 and later versions,
we reduced the amount of high idle from 100 percent assumed in MOVES2014 to 33 percent, to better
match the references noted above.

The use of accessories (e.g., air conditioners, heaters, televisions, etc.) provides recreation and comfort to
the operator and increases load on the engine. There is also a tendency to increase idle speed during long
idle periods for engine durability. The emission rates estimated for the extended idle in MOVES assume
both accessory use and engine idle speeds set higher than used for "curb" (non-discretionary) idling. We
classify the extended idling that does not employ high speed idle without additional auxiliary loads as "curb
idle."

Emissions data from the references in the data sources section (2.3.1.1) was classified into one of three
idle conditions. The first condition, which has a low engine speed (<1,000 rpm) and no air conditioning is

128


-------
representative of curb idle. The second condition is representative of extended idle with higher engine
speed (>1,000 rpm) and no air conditioning. The third represents an extended idle condition with higher
engine speed (>1,000 rpm) and air conditioning. For the purpose of this analysis, the load placed on the
engine due to air conditioning is assumed to represent all forms of accessory load that may be used during
hotelling.

Note that some of the idle tests are of short duration. We believe it is reasonable to classify the short-
duration tests as curb idle in our calculations of extended idle emissions. We are using the short-duration
idle tests from the pre-2007 MY vehicles because idle emissions stabilize more quickly than later model
years because the pre-2007 vehicles lack the emission aftertreatment technologies that can lose
effectiveness as exhaust cools during longer idle periods.

For 1990 and earlier, we developed curb idle emission rates based on the analysis of the 18 heavy-duty
diesel trucks from 1975-1990 model years used in the CRC E-55/59 study and one MY 1985 truck from the
Lim study. The curb idle rates were then adjusted using ratios from 1991-2006 trucks to estimate the
elevated NOx emission rates characteristic of higher engine speed and accessory loading of extended idle.

In particular, as summarized in the tables in Appendix D, data from 188 vehicles were used to estimate
curb idle NOx emission rates for 1991-2006 model year heavy-duty diesel trucks. The curb idle NOx
emission rate of 91 g/hr was calculated by weighting the average NOx emission rate from each test by the
number of vehicles tested. Four studies and results from 31 vehicles included higher idle engine speed and
air conditioner use, which resulted in a weighted idle NOx emission rate of 227 g/hr. The ratio of the 1991-
2006 MY NOx emission rate from curb idle to idle with high engine speed and A/C was applied to the 1990
and earlier model year curb idle rate to get the calculated 1990 and earlier NOx emission rate with high
engine speed and A/C. A similar strategy was applied to the THC, CO, and C02 emission rates for 1990 and
earlier model years.

For both the MY 1950-1990 and 1991-2006 vehicles, using the data summarized in Appendix E, adjusted
emission rates were calculated for each pollutant by weighting the overall "high speed idle, A/C on" results
by 0.33 and the "low speed idle, A/C off" (i.e., curb idle) results by 0.67 to account for the fraction of idling
at high and low engine speeds.

The NOx, THC, CO, and PM2.5 emission rates from this data analysis are primarily from diesel HHD trucks. In
MOVES2014, we calculated the MHD extended idle emission rates as half of the corresponding HHD
emission rates. However, a study by Khan et al. (2009)86 found that MHD and HHD trucks had similar
emission rates during extended idle. Consequently, MOVES applies the same extended idle emissions rates
to MHD and HHD, as shown in Table 2-55.

MOVES stores PM2.5 emission rates according to elemental carbon (EC) and non-elemental carbon
(NonECPM), but the data sources used to calculate the extended idle emission rates reported only total
PM2.5. As mentioned in Section 2.1.2.1.8, an EC/PM fraction of 46.4 percent is applied for the running
exhaust idle operating mode (opModelD 1), and we also apply it to extended idle. The resulting EC and
NonECPM rates are also shown in Table 2-55.

No adjustment to the rates are made to account for tampering and mal-maintenance (T&M) because the
pre-2007 trucks do not have the exhaust aftertreatment technologies that are anticipated to see large
emission increases when they are tampered or mal-maintained. While the 188 trucks used for these
estimates may not fully represent real-world emission deterioration, they do include real-world vehicles at

129


-------
a variety of ages and conditions and thus it would be "double-counting" to apply the exhaust running T&M
effects to these rates.

Table 2-55. Pre-2007 Extended idle emission rates (g/hour) in MOVES by pollutant for MHD and HHD

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 2007-2026 Model Years

The extended idle emission rates for model years 2007 to 2026 are based on the following data sources
and analysis.

2,3,2,1 Data Sources

The extended idle emission rates for model year 2007 and later heavy-duty diesel combination long-haul
trucks (sourceTypelD 62) diesel emission rates in MOVES are based on two test programs measuring
extended idle emissions from HHD diesel trucks. The Texas Transportation Institute (TTI) tested extended
idle emission from 15 heavy-duty diesel tractors ranging from model year 2005u to 2012.87 Another study
conducted by California Air Resources Board (ARB)88 tested five tractors (engine model years 2007 and
2010). As discussed in the analysis section (Section 2.3.2.2), the four MY 2005 and 2006 engines included in
the TTI study are included in the development of the 2007 and later model year emission rates for THC,
CO, NOx, and energy because there is no noticeable differences in the emission rate for these model years
from comparable MY 2007 and later engines. For PM2.5, these engines are only used for comparison and to
develop T&M adjustment factors.

The study (TTI or ARB), engine model year, engine manufacturer, odometer, the NOx certification level,
California Clean Idle certification, and engine aftertreatment are listed for each of the trucks in Table 2-56.
The last three columns in Table 2-56 are taken from the California Executive Order certification database.89
NOx certification level (g/bhp-hr) is the standard to which the engine was certified. Some 2010 and later
engines were certified above the 0.2 g/bhp-hr NOx 2010 federal standard due to the emissions averaging,
banking and trading (ABT) program, and EPA allowance of nonconformance penalty (NCP) engines in
2012.90 In these cases, the family emission limit for which the vehicle was certified is reported in Table
2-56. 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 2-56 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.

u 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.

130


-------
Table 2-56. HHD Diesel Tractors Used to Update the MY 2007-2026 Extended Idle Emission Rates

Study

Engine
MY

Engine

Odometer

NOx cert
(g/bhp-hr)

Clean
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 2-57) 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:

131


-------
•	Six of the fifteen TTI trucks were not able to be commanded into high idle.

•	The 'stabilized' idling emission tests did contain some high idle that appears representative
of automatic engine control strategies for 2007 and later trucks. Two of the trucks included
high idle during the winter stabilized tests due to automatic engine control strategies. We
assume that for 2007 and later technology trucks, operators and manufacturers rely on
automatic engine control strategies rather than the vehicle operators to employ high idle
conditionsv. 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.

•	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 2-73 through Figure 2-77.

Table 2-57. Ambient Test Conditions for the TTI Extended Idle Tests

Test ID

Temperature

Relative Humidity

Auxiliary Load

Hot (Summer)

100'F (37.8TC)

70%

Air conditioning

Cold (Winter)

3Cf F (-1.1° C)

N/A

Heating System

ARB tested five trucks on a chassis dynamometer on the ARB HHDDT 4-mode cycle, reporting the g/hr
results from the 10-minute 'Idle' mode. Before testing the 'Idle' mode, the vehicle was first warmed on a
pre-conditioning cycle, and then soaked for 10-20 minutes.91 Additional test conditions were not reported
by ARB, but we assumed that the ARB vehicles were tested at moderate temperatures, with no auxiliary
loading. Thus, we treated the ARB data as more representative of an extended idling truck that did not
require significant 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

In developing the extended idle emission rates, we averaged the emission rate from each of the tests,
within model year ranges that represent engine and aftertreatment technology groups that have similar
impacts on extended idle emissions. Where possible, we used all 35 tests (15 trucks x 2 conditions = 30 TTI
tests, and 5 ARB tests). Because there were more TTI tests, the average within each model year group is
weighted significantly towards the TTI tests. We chose to weight each test equally, because we believe the
TTI data are more representative of real-world extended idle conditions, because they were tested with
auxiliary loads at non-standard 'lab' temperatures.

The individual test results and the average emission rates by model year group are presented in the
following figures (Figure 2-73 to Figure 2-77). Within each figure, the tests are distinguished according to

vAs discussed earlier, our assumptions for pre-2007 trucks are different.

132


-------
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 C02, 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 2-73 displays the C02 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 C02 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 C02 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.

C02

15000

3

10000

C

o

C/)

'(/>

a>
a>

2 5000-

0-







A





•



•

•	• • A A A

•	• * * , *

• • 2 i
« S



i













condition



•

hot



•

cold



•

lab

SCR



•

0



A

1

2006

2008

2010

2012

Figure 2-73. C02 Emission Rates from the TTI and ARB Programs by Engine Model Year, and Average Emission Rate

(line) based on all the data.

133


-------
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 2-74 displays the CO individual test results. No trend is observed with respect to model year or use
of aftertreatment. The laboratory ARB tests are lower than the TTI tests, which could be due to the lower
fuel consumption of the tests. The CO emission rate is slightly lower than the emission rate for 1990-2006
MY of 55 g/hr. Similar to C02, a single average emission rate is calculated for all the tests results and is
applied to all 2007 and later model years.

150

15)100

GO
£=

O

CO

"co

'E
a)
a)

=5 50

o-

co

t

4

condition



•

hot



•

cold



•

lab

SCR



•

0



A

1

2006

2008

2010

2012

2014

Figure 2-74. CO Emission Rates from the TTI and ARB Programs by Engine Model Year, and Average Emission Rate

(line) Based on All the Data

Figure 2-75 displays the NOx individual test results. We initially expected the data to show a decrease in the
extended idle emission rates beginning in MY 2008 to account for the California Clean Idle Certification (all
MY 2008 and later trucks were clean-idle certified). However, no reduction was observed. We also
expected to observe a decrease in 2012, with the full implementation of SCR, but this was also not the
case. Therefore, we calculated average NOx emission rates for two model year groups (2005-2009) and
(2010-2026) as represented by a solid line in Figure 2-75. 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 2-78.

134


-------
250-

200-

(/3150 n
c
o
c/5
'in

§1001
a;

"O

50-

o-



NOx







~
~



s

A

condition

•	hot

•	cold

•	lab

SCR

•	0

*¦ 1

2006

2008

2010

2012

Figure 2-75. 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 2-76 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 N02, 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 trucksw and are used to represent the 2007-2009 emission rates in MOVES. The 2010-2012
represents DPF equipped trucks, with some penetration of SCR equipped trucks. The model year group
representing 2013 and later model years was developed because starting in 2013, Navistar began certifying
a heavy heavy-duty diesel (HHDD) engine equipped with SCR aftertreatment. In 2014 and 2015, Navistar
and all other engine manufacturers certified all their HHDD engines equipped with SCR aftertreatment.92
Therefore, emission rate for the 2013+ model year group was estimated by averaging the rates of all the
SCR equipped trucks in the data set, even though the dataset did not include any data on 2013 and later
model year engines.

w The 2005-2009 THC rates here are ~3 times smaller than the 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.

135


-------
THC

-10-

Ui

c

o


en

'E

a>

_5>

"O

5-

0-



A
A

condition



•

hot



•

cold



•

lab

SCR



•

0



A

1

2006

2008

2010

2012

2014

Figure 2-76. 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 2-77 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.x 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.

x The MY 2005-2006 PM2.5 emission rates measured from the TTI data are only ~3 times higher than the MY 2007-
2009 PM2.5 rates, and roughly ~10 times smaller than the PM2.5 rates for MY 1990-2006 (2.5 g/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
information from the TTI dataset due to limitations on time and resources.

136


-------
PM

0.5

0.4

o>

0.3

c

o

u>

c/>

a>0.2
¦a

0.1

0.0

<









•











i\







•

\ #

\

\

<

ft











k -	

condition

•	hot

•	cold

•	lab

SCR

•	0
A 1

2006

2008

2010

2012

2014

Figure 2-77. PM2.s 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 2-58 to estimate the EC and PM2.5 emission rates.

Table 2-58. Baseline elemental carbon to PM2.5 fraction assumed for extended idling

Model Year
Group

EC/PM

Source

Pre-2007

0.26

MOVES2014 Extended Idling34 Y

2007-2009

0.10

ACES Phase I91

2010+

0.16

ACES Phase II93

2.3.2.3 Tampering and Mal-maintenance

As discussed in Section 2.3.1.2, we did not incorporate tampering and mal-maintenance effects on the pre-
2007 extended idle rates. For the 2007 and later extended idle rates, we incorporated the effects of the
effect of tampering and mal-maintenance (T&M) for two reasons:

1. The twenty vehicles used to estimate the extended idle emission rates did not appear to
include any tampered or mal-maintained vehicles with elevated emission rates. In
addition, 14 of the 20 vehicles had odometer readings with less than 100,000 miles (Table
2-56).

Y 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.

137


-------
2. The 2007 and later technology includes aftertreatment technology, including diesel

oxidation catalysts (DOC) and diesel particulate filters (DPF). We anticipate that the failure
of these after-treatment systems would significantly increase extended idle emissions if
they were tampered or mal-maintained.

We incorporated the T&M effects for extended idle exhaust using different data and methodology than
was used to derive the tailpipe exhaust emission rates for two reasons:

1.	Extended idle emissions in MOVES are stored in the EmissionRate table, and are not
distinguished by vehicle age, as the running and start exhaust emission factors. To fit the
current MOVES structure, we incorporated the effects of T&M into a single emission rate
by model year that applies to all vehicle ages.

2.	We are less confident in the application of the emission effects of T&M failures estimated
for running emissions in Appendix B to extended idling emission. For example, we do not
think failure of selective catalytic reduction (SCR) aftertreatments systems should impact
extended idling NOx emission rates as much as running exhaust emissions because the SCR
systems is not fully operational during long idling periods. Instead, we estimated the
effects of T&M on 2007 and later extended idle emissions using pre-2007 extended idle
emissions as surrogate values for 2007 and later extended idle emission with failed
aftertreatment systems.2

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 2-59. As shown,
the MOVES EC/PM emission rates for MY 2007+ trucks are slightly higher than the 'Baseline' EC/PM
fractions in Table C-2, because the fleet emissions are assumed to include some emission contribution
from trucks with failed DPFs, which have a higher EC/PM fraction.

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 8 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.

z Starting in MY 2027, these vehicles have NOx emission standards for idle operation. Thus, we may reconsider this
approach in a future version of MOVES.

138


-------
Table 2-59. Extended Idle Emission Rates for 2007 and Later Model Year Heavy-Duty Vehicles

Model Year
Group

C02
(g/hr)

CO
(g/hr)

NOx
(g/hr)

THC
(g/hr)

PMz.5
(g/hr)

EC

(g/hr)

nonEC
(g/hr)

EC/PM

2007-2009

7151

39.3

100.5

8.5

0.087

0.012

0.076

0.13

2010-2012

7151

39.3

42.6

2.7

0.034

0.006

0.028

0.18

2013+

7151

39.3

42.6

1.6

0.021

0.004

0.017

0.20

2,3,2,4 MHD Regulatory Class

The extended idle emission rates for MHD are assumed to be the same as HHD for the following two
reasons. First, MHD trucks are estimated to account for only five percent of long-haul combination trucks
in the US and therefore, they are a minor contributor to the emissions from extended idling trucks. Second,
Khan et al. 200994 evaluated extended idle emission rates of pre-2007 MHD engines and did not observe a
pronounced difference in extended idle emission rates between MHD and HHD trucks. Taken together,
these imply that any difference in emissions modeled with unique MHD extended idling rates would be
minimal, so without any extended idling data on 2007 and later model year MHD trucks, we felt it was
most defensible to keep the MHD emission rates the same as the HHD emission rates.

2.3.3 2027-2060 Model Years

For MY2027 and later vehicles subject to HD2027 standards, we anticipate that reductions in the HHD and
MHD NOx extended idle emissions rates will be driven by the idle standard, rather than the duty-cycle
standards in the rule. The duty-cycle standards do not contain high duration extended idling (> 1 hour) that
is representative of truck hotelling activity. We did not estimate any change in extended idle emission rates
due to the lengthened warranty or useful life periods because MOVES extended idle rates do not vary by
age.

To calculate the effects of the new standard, we estimated extended idle emission rates that would comply
with the off-cycle NOx/C02 g/kg standard calculated in Table 2-13. We then used Equation 2-25 to calculate
the extended idle off-cycle NOx g/hr emission rate based on the MOVES extended idle C02 g/hr emission
rate, as shown in Table 2-60.

Table 2-60 Calculation of HHD and MHD Extended Idle NOx g/hr Emission Rates for MY2027+

Model Year
Group

MOVES
Extended
Idle Rates
C02 (kg/hr)

Idle Standard (g/hr)

Idle

Standard
NOx/COz
(g/kg)

Idle-standard compliant
NOx emission rate (g/hr)

2027-2028

7.191

9

1.17

8.42

2029+

7.191

8.7

1.13

8.14

2.3.4 Model Year Trends

Figure 2-78 through Figure 2-81 illustrate the extended idle emission rates for regClasslDs 46 and 47.

139


-------
As shown, the NOx and the CO extended idle emission rates have a relatively small decrease between the
pre-2007 and the 2007+ model years. For THC and PM2.5, we observe large decreases starting in MY 2007,
which is consistent with our understanding of the effect of diesel particulate filters. We observed a
decrease by ~29 times in extended idle PM2.5 rates between the pre-2007 and post-2007 extended idle
rates corresponding to the implementation of the DPFs, which is consistent with the ~27 times decrease in
PM2.5 running exhaust emission rates from PM2.5 certification data as discussed in Section 2.1.2.2.
Extended idle rates for NOx decrease in MY 2027 due to the HD2027 rule, but other pollutant rates remain
unchanged.

100'

a:
a;
2

4->

X

LU

X

O

50

0 ¦

Reg Class

46-MHD67

47-HHD8
49-Gliders

1980

2000

Model Year

2020

2040

Figure 2-78. Extended Idle NOx Emission Rates for HHD and MHD Diesel Vehicles by Model Year

140


-------
50-

40-

o
,c

2 30-

4)

ra
QL

ai

s 20-

UJ

0

1

10-

Reg Class

46-MHD67

47-HHD8
49-Gliders

1980	2000	2020	2040

Model Year

Figure 2-79. Extended Idle THC Emission Rates for HHD and MHD Diesel Vehicles by Model Year

~ 40

i2 20

o

o

f

Reg Class

46-MHD67

47-HHD8
49-Gliders

1980	2000	2020	2040

Model Year

Figure 2-80 Extended Idle CO Emission Rates for HHD and MHD Diesel Vehicles by Model Year

141


-------

-------
1.5e+05

o

-G

1.06+05


-------
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.101AA

The second study used to update APU emission rates was by Frey and Kuo (2009),96 who tested two APU
systems (APU ID 2 and 3), equipped with 2006 Kubota Z482 engines. The APU systems were tested at a
range of electric output loads to obtain the fuel consumption relationship with the electric power
demands, and the fuel-based emission rates. The study measured the in-use APU electric loads from a fleet
of 20 vehicles (10 trucks equipped with each APU system) for over a year. They then used the relationship
between electric power demand and the fuel-based emission factors with the average energy use of the
APU system to estimate average APU (g/hr) emission rates of C02, CO, NOx, THC, and PM for both a mild
temperature (50-68C F) scenario and a high temperature (lOOf F+) scenario. Frey and Kuo 2009 reported a
PM emission rate, but the emission rate is 'inferred from the literature' because their PM measurements
were semi-qualitative.

An additional two studies were used as a source of data to compare and evaluate the APU emission rates
obtained from the studies mentioned above. TTI 201297 conducted testing of two APU systems using their
environmental chamber at both lOOf F and Of F. The APU systems (APU 4 and 5) manufacturer, engine
make and model year were maintained confidential in the report. Storey et al. 200398 tested a Pony Pack
APU System (APU ID 6), equipped with a Kubota Z482 engine, in an environmental chamber at both 90f F
and Of F. This is one of the studies used by Frey and Kuo 200996 to determine the PM emission factor for
the APU's tested in their study. The engine year, engine displacement, and engine power were not
reported in the TTI 2012 and Storey et al. 2003 studies. For this reason, these studies were used only as
comparative data sets.

Table 2-62. 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 201495

2

Kubota Z482

2006

0.48

10.9/8.1

Tier 2

Frey and Kuo
200996

3

Kubota Z482

2006

0.48

10.9/8.1

Tier 2

Frey and Kuo 2009

4

Confidential Information

TTI 201297

5

Confidential Information

TTI2012

6

Kubota Z482









Storey et al. 200398

Table 2-63 contains the in-use emission rates measured from reviewed APU systems. As shown, the
emission and fuel rates for the APUs measured in the TTI 2014, and Frey and Kuo 2009 (APU ID 1, 2 and 3)
compare well with the APU emission rates reported from TTI 2009 Storey et al. 2003 (APU ID 4, 5, and 6).

AA 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.

144


-------
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 (C02, 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 C02,
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 C02
emissions and fuel use. For APU ID 5 and 6, the hotter temperatures had higher C02 emissions and fuel
use.

For CO, NOx, THC, and PM there are conflicting trends with respect to ambient temperature. For APU 2 and
3, NOx and PM emissions are higher at the hot conditions compared to mild conditions, consistent with the
higher fuel use. However, CO shows lower emissions at hot conditions, and THC shows a mixed trend. For
the other studies, there is no consistent trend between the hot and cold conditions.

Table 2-63. In-UseAPU Emission Rates

APU
ID

CO.
(g/hr)

CO

(g/hr)

NO,
(g/hr)

THC
(g/hr)

PM
(g/hr)

Fuel
(gal/hr)

Ambient
condition

Temperature

m

DPF

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

a Frey and Kuo 2009 report the mild condition for auxiliary

oads on the trucks is for am

Dient

temperatures ranging from 10-20°C (50-68TF)

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

145


-------
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 2-64.

Table 2-64. Average APU Emission Rates from non-DPF APU IDs 1, 2, and 3 according to Cold, Hot, and Mild Ambient

Conditions

C0z
(g/hr)

CO

(g/hr)

NOx
(g/hr)

THC
(g/hr)

PM
(g/hr)

Fuel
(gal/hr)

Ambient
condition

Temperature
££1

DPF

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 2-66.

We estimated elemental carbon (EC) fraction of PM from composition measurements made on APU ID 1 as
reported in Appendix J. For each test, we calculated the elemental carbon/total carbon ratio, and then
averaged the ratio across all cold and hot tests, separately for the DPF and the non-DPF tests as shown in
Table 2-65. 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 2-66.

Table 2-65. 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 2-66. Fleet-Average Non-DPF Equipped APU Emission Rates in MOVES

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)

3510

10.0

15.6

1.3

0.96

0.13

0.83

0.14

0.35

The HD GHG Phase 2 rule implements a phase-in standard that requires APUs installed in new tractors to
meet lower PM standards from MY 2021 through MY 2024 (beyond the Tier 4 nonroad standards).99 The
APU PM standards along with the current Tier 2 and Tier 4 nonroad standards for nonroad diesel engines
8
-------
Table 2-67: Nonroaa (8
-------
Table 2-68 APU Emission Rates in MOVES 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)

PMz.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

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.

2.5 Glider Vehicle Emissions

"Glider vehicles" or "Gliders" refer to vehicles with old powertrain (engine, transmission and/or rear axle)
combined with a new chassis and cab assembly. Most gliders are Class 8 heavy heavy-duty vehicles. They
typically use model year 2001 or older remanufactured engines that do not have to use emissions controls
such as DPF or SCR needed to meet the stringent PM and NOx standards starting MY 2007+. 102

Starting with MOVES3, we model the emission impacts of the glider vehicles as a separate regulatory class
(regClassID 49) because their population became significant starting with model year 2008 as described in
the Population and Activity Report.

For modeling purposes, all glider vehicles are presumed to be combination trucks (sourceTypelD 61 and
62) running on diesel fuel. EPA's in-house glider vehicle emission testing data103 suggest that glider
emissions have similar THC, NOx, PM2.5, and C02 running exhaust emission rates to the MOVES model year
2000 heavy heavy-duty vehicles (regClassID 47), while CO from glider vehicles is higher. Based on this
analysis, the MOVES running, start, and extended idling exhaust rates for gliders of all model years88 are
set equal to those of the model year 2000 heavy heavy-duty vehicles.

For example, Figure 2-83 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.

BB Glider emission rates are intended to be the same for all model years from 2000 through 2060. However, for EC
and non-EC PM, the pre-MY 2007 values were updated without changing the post-MY 2007 values. As discussed in
the activity report, glider activity in MOVES begins in model year 2008.

148


-------
regClassID H47B49

MY2000

IIII II IIII ..IIII

.. .. II II II II .. II II II

i .. .. II II II II .. II II II

.. II II II

.. II

~i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—r

..IIIIIII

MY2008

.1 .1 ll ll l! .. .1 ll ll ll ll I

MY2010+

I >1 il .1 ll.. .1.1 .1 J ¦

. _. I .1 .1 J al _¦ -I .1 J ll j

.1 .1 .1 .l ll ll

. . .1 I I.I .1 ill





















. .11 1 1 .

Mill



..ll

||

... 1 1 l l

. 1 1 1 l

|



. 1 1 l l

m
o

"0

. I I l

t—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—r

opModelD

Figure 2-83. Comparison of the running exhaust emission rates (0-3 age group) ofHHD (regClassID 47) vs. Gliders (regClassID 49) for selected pollutants (NOx,

ECPM, NonECPM) and model year groups

149


-------
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
3.1.1 THC, CO, and NOx

The heavy-duty gasoline running rates were analyzed in three stages. The MY 1950-2007 emission rates
were originally developed in MOVES2010. When we extended calculations to 40 ages, we duplicated the
emission rates for MY 1960 for model years back to 1950. In MOVES2014, we updated the MY 2008-2009
heavy-duty gasoline rates to account for the Tier 2 and 2007 heavy-duty rulemakings. In MOVES3, we
updated the MY 2010-2026 emission rates based on the more recent testing data. In MOVES4, we revised
the MY2027+ emission rates to account for the impact of HD2027 rule. The analysis of PM2.s 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 1950-2007 Model Years

The heavy-duty gasoline emission rates for model year 2007 and earlier were carried over from previous
versions of MOVES. They are based on analysis of four medium heavy-duty gasoline trucks from the CRC E-
55 program and historical data from EPA's Mobile Source Observation Database (MSOD)104, which has
results from chassis tests performed by EPA, contractors and outside parties. The heavy-duty gasoline data
in the MSOD is mostly from pickup trucks which fall mainly in the LHD2b3 regulatory class. Table 3-1 shows
the total number of vehicles in these data sets. In the real world, most heavy-duty gasoline vehicles fall in
either the LHD2b3 or LHD45 class, with a smaller percentage in the MHD class. There are very few HHD
gasoline trucks now in use.

Table 3-1 Distribution of Vehicles in the Data Sets by Model Year Group, Regulatory Class and Age Group

Model year
group

Regulatory class

Age group

0-5

6-9

1960-1989

MHD



2

LHD2b3



10

1990-1997

MHD



1

LHD2b3

33

19

1998-2002

MHD

1



LHD2b3

1



Similar to the HD diesel PM, THC, and CO analysis described above, the chassis vehicle speed and
acceleration, coupled with the average weight for each regulatory class, were used to calculate STP
(Equation 1-6). To supplement the available data, we examined engine certification data as a guide to

150


-------
developing model year groups for analysis. Figure 3-1 shows averages of certification results by model
year.

30 -r	

25 	

20 		I--

-Q

" 15

O 10

5 	

* t	i

z

I CO
Nox
> HC

...s5

if H

" -- -- -- ~

» * t- *

1.4

-	1.2

-	1
0.8
0.6

.a
"5>

h 0.4 t

-	0.2

-	0

-0.2

1990	1995	1990	1995	2000

Model year

2005

2010

Figure 3-1 Brake-Specific Certification Emission Rates by Model Year for Heavy-Duty Gasoline Engines

Based on these certification results, we decided to classify the data into coarse model year groups: 1960-
1989, 1990-1997, and 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.

151


-------
4000"

3000"

2000

1000 -

0-
. 150 H

3

O

? 100-

.2 50 ~l

V)
in

oi

500 "
400
300
200-
100 -
0-

A	¦

t t A * t * - * ft t i t i • ' # ~ ft # t

• A

• A *	J * *	A

* • I s I ¦ ¦ 111 i;1! ¦' ~ ~ ~ *

¦i'i'	" " ' 	1

A

~
~

~

~

i	~ * ¦ ~,; ¦

		

1—r

n—r

Model.Years

~ 1960-1989
A 1990-1997
¦ 1998-2007

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

152


-------
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

NO,

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.

153


-------
2500 J
2000
1500 J
1000
500
0

^ 100
o

D) 75

JV

E 50

c

o

(O OE

tn
£

m 0

200
150
100
50
0

A A

~ A ~

A A * *

A

~ i *

A
~

. 1 »
• ~

1 i

A
A •

4 A A t * *

A	* *

A * ~	A * ~

~ ~

* ~ • *

1	1	1	1	1	T

"T—i—i—r

age

A 6+
~ 0-5

0 1 11 12 13 14 15 1621 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

andLHD45

3.1.1.1.2 MHD and HHD

Like the LHD rates described above, the 2007 and earlier MHD and HDD gasoline rates are based on
emissions data from the mix of LHD2b3 and MHD vehicles outlined in Table 3-1. The same model year
groups were used to classify the emission rates: 1950-1989, 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.

154


-------
3

o

30000-
20000
10000-
0

600'

CO

3 400
ro

i_

c

.2 200

cn

0

2000'
1500
1000-
500
0

































4











































*

*









































*

*











A

A

A +

o
o





L f _S ¦ ¦ A J

*

1 ~ " 1





¦ J

*

* ~ '
~

1 ¦ ¦ 1

¦













































































*











































+ *
* 4









































~

~

A









•

~ A
* A

X

o

t i t 1 4 1

\ 1

J

¦ ' *	

































































































































~



























*

~











* *





1

A.

z
o











•

• *
A ¦







i ¦

t ¦



+ k ¦ 1







* *

t ¦ ¦

. 1









































Model.Years

~ 1960-1989
A 1990-1997
¦ 1998-2007

0 1 11 1213141516212223242527282930333537383940
op Mode ID

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.

155


-------
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 1950-2060

OpModelD

THC

CO

NOx

0

2.85

1.45

1.67

1

2.43

1.79

1.85

11

3.12

1.66

1.88

12

3.36

3.12

1.13

13

3.53

3.16

1.11

14

3.53

3.16

1.11

15

3.53

3.16

1.11

16

3.53

3.16

1.11

21

2.78

1.67

1.42

22

3.08

2.59

1.23

23

2.97

3.31

1.05

24

1.80

1.54

1.03

25

1.80

1.54

1.03

27

1.80

1.54

1.03

28

1.80

1.54

1.03

29

1.80

1.54

1.03

30

1.80

1.54

1.03

33

2.45

2.41

1.33

35

2.16

2.41

1.19

37

2.16

2.41

1.19

38

2.16

2.41

1.19

39

2.16

2.41

1.19

40

2.16

2.41

1.19

Figure 3-5 displays the resulting emission rates by operating mode bin and age group for the LHD45, MHD,
and HHD gasoline vehicles, which were calculated by applying the multiplicative age effects in Table 3-3.

156


-------
7500-
5000-
2500"
0

^ 500'
o

O)

400'

| 300
! 200

$ 100
'E

LU 0"

1000-

750-

500

250-

o-

a +
*

+

~ *

* ± ~

* *

A. . *

i * *

A A *
~ * ~ •

: * *

* * * *

* * * *

• *

A

A
~

A ±

~ * + +

age

*	6+

•	0-5

~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 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.

157


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



MDPV (Tier 2

8.5k- 10K (Class

10K-14K



Bin 5)

2B)

(Class 3)

Units

g/mile

g/mile

g/mile

Fully Phased in MY

2009

2009

2009

THC

0.09 NMOG

0.195 NMHC

0.230 NMHC

CO

4.2

7.3

8.1

NOx

0.07

0.2

0.4

This section documents the THC, CO and NOx emission rates for regulatory class LHD2b3 vehicles in model
years 2008 and 2009. In conducting this analysis, we lacked any modal data on LHD vehicles and therefore,
we ratioed the modal emission rates measured from light-duty vehicles by the difference in standards.12 By
MY 2008, the certification results demonstrated that LHD2b3 were nearing the emission levels of light-duty
vehicles certified to the Tier 2 Bin 8 standard.34 Consequently, we relied on the MOVES2014 analysis of in-
use Tier 2 Bin 8 vehicles conducted for the light-duty emission rates.12 We applied this analysis to derive
MY 2009 emission rates, then calculated MY 2008 rates by interpolating between MY 2007 and MY 2009.

Although the light-duty rates are based on VSP, rather than STP, adapting them for the LHD2b3 rates was
deemed an acceptable approximation because the gasoline LHD2b3 gasoline vehicles are chassis-certified
to distance-based standards (g/mi). Accordingly, the vehicle emissions rates are less dependent on the
individual power and weight of the vehicle, and should scale approximately to the g/mile emission
standards.29

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.34,30 Table 3-5 displays the aggregated LHD2b3 standards, Bin 8 FTP standard and the ratio between
the standards by pollutant.

Table 3-5 Aggregate LHD2b3 Standard Ratios against Bin 8 Modal Rates



Aggregate LHD2b3
FTP standard

Bin 8 FTP
standard

Aggregate
LHD2b3/Bin 8

NMOG

0.18

0.1

1.8

CO

7.49

3.4

2.2

NOx

0.22

0.14

1.6

29	This approximation may 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 tons, which differs from the/Sca/e value of 2.06 metric tons.

30	As documented in MOVES2014 documentation, this analysis assumed that 5% of the gasoline LHD2b3 engines
Were engine"certified, but, actually, all gasoline fueled LHD2b3 vehicles are chassis-certified. However, the engine-

certification standard has a small impact on the calculated aggregated standard conducted for MOVES2014.

158


-------
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.12 The "split" ratios for running and start,
which were applied to the light-duty Tier 2 Bin 8 vehicle emission rates are shown in Table 3-6.

Table 3-6 Ratio Applied to Light-Duty Tier 2 Bin 8 Emission Rates to Estimate Regulatory Class LHD2b3 Emission Rates

for 2008-2009 MY



THC

CO

NOx

Running

2.73

2.73

1.95

Start

1.37

1.37

1.00

We also adopted the light-duty deterioration effects and applied them to the MY 2009 regulatory class
LHD2b3 (regClassID 41) emission rates. The light-duty emission rates have age effects that change with
each of the 6 age groups in MOVES, as shown in Table 3-7.

Table 3-7 Multiplicative Age Effect used for Running Emissions for Regulatory Class LHD2b32009 Model Year

ageGroupID

THC

CO

NOx

3

1

1

1

405

1.95

2.31

1.73

607

2.80

3.08

2.21

809

3.71

3.62

2.76

1014

4.94

4.63

3.20

1519

5.97

5.62

3.63

2099

7.20

6.81

4.11

After applying the steps described above (scaling the emission factors by ratio of FTP standards, and
applying light-duty deterioration trends), we restricted the scaled data so that the individual emission rates
by operating mode were never higher than MY 1998-2007 regulatory class LHD2b3 rates. This step
essentially "capped" the emission rates, such that none of the modal rates for MY 2009 are higher than
their counterparts for MY 2007 and earlier. MY 2008 rates are interpolated between MY 2007 and MY
2009 emission rates as discussed later.

This final step "capped" the model year 2009 emission rates in the highest operating modes, as shown in
Figure 3-6. For THC, emission rates in operating modes 28-30 and 38-40 were capped for some or all age
groups by the pre-2007 emission rates. For CO, emission rates in 12 of the 23 running operating modes (1,
16, 23-24, 27-30, 35-40) were capped by the pre-2007 rates. None of the NOx emission rates were
impacted by this step. Figure 3-6. shows the regulatory class LHD2b3 model year 2008-2009 emission rates
for CO, THC, and NOx. In the figure, rates "capped" by the pre-2007 rates exhibit the uncharacteristic
"stairstep" deterioration trends. Even with the "capping" effects, the rates for regulatory class LHD2b3

159


-------
(regClassID 41) are higher than those for light-duty trucks (regClassID 30), with a few exceptions. The few
exceptions are some of the age-dependent THC and or CO emission rates in operating modes 1, 30, 38, 39,
and 40. However, the majority of emission rates are considerably higher for the heavy-duty (LHD2b3) than
for the light-duty trucks. Similarly, when the FTP is simulated from the resulting rates, estimated
composites are substantially higher for LHD2b3 than for light-duty trucks.

The Light-duty Tier 2 standards shown in Table 3-4 phase-in at a rate of 50 percent in MY 2008 and are
considered fully phased in by MY 2009.106 For estimating emission rates in MOVES, we 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.

0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24

Vehicle age, years

Figure 3-6. Age Effects for CO, THC, and NOx Emission Rates for Regulatory Class LHD2b3 (regClassID 41) Vehicles in

Running Operating Modes for MY 2007, 2008 and 2009

2007

2009

2500-
2000-
1500-
1000-
500-

o-

o
o .

opModelD

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.

3.1.1.2.2 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.31 The percentage of gasoline-

31 Negligible portions are run on other fuels. The figures are aggregated from data supplied by Polk.

160


-------
fueled vehicles decreases as the GVWR class increases. Since these vehicles are a small portion of the fleet,
there is relatively little data on these vehicles, and therefore, the current 2008 and 2009 model year
emission rates are from MOVES2010.107 The rates are modeled by applying a 70 percent reduction to the
MY 2007 running rates starting in MY 2008, which is consistent with the emission standard reduction with
the "Heavy-Duty 2007 Rule."108,32 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 Years33

In MOVES3, we updated the THC, CO, and NOx emissions rates for MY 2010 and later vehicles for all
gasoline heavy-duty regulatory classes. The initial update is described here. Additional analysis to account
for the Tier 3 and HD2027 rules are described in Sections 3.1.1.3.1 and 3.1.1.3.2 below. The updated rates
are based on analysis of real-world PEMS-based emissions measurement data from two engine-certified
and one chassis-certified heavy-duty gasoline vehicles (Table 3-8) with model years between 2015-2017. As
explained in the PM2.5 section (Section 3.1.2.2), we also conducted chassis-dynamometer laboratory
testing on these vehicles but used the PEMS gaseous emissions data because it better represents emissions
in the real-world.

The Ford and Isuzu vehicles used the most popular engine configurations for recent model year heavy-duty
gasoline Class 4 vehicles. Each of the HD gasoline vehicles had three-way catalyst (TWC) technology to
control THC, CO, and NOx emissions. However, one key difference compared to light-duty gasoline vehicle
TWC configuration is that the engine-certified HD gasoline vehicles do not use a close-coupled TWC. There
might also be differences in catalyst precious metal loading and in-cylinder combustion control for
maximum TWC efficacy. The reason for these differences is that engine-certified and chassis-certified
gasoline spark-ignited vehicles have to meet different standards.

Table 3-8 Summary of MY 2015-2017 Heavy-Duty Gasoline Vehicles with Real-World PEMS-based Emissions

Measurement Data

Vehicle

Engine

Test Weight
(lbs)

Cert-
ification

Make

Model

MY

Odometer
(miles)

GVWR
(lbs)

GCWR
(lbs)

Family

Displ
(L)

Low

High

Isuzu

NPR

2015

48,000

14,500

20,500

FGMXE06.0584

6.0

8,620

12,940

Engine

Ford

E450

2016

31,000

14,500

-

GFMXE06.8BWZ

6.8

9,320

13,080

Engine

RAM

3500

2017

32,000

13,300

19,900

HCRXD06.45W0

6.4

14,557

18,020

Chassis

The testing was conducted by US EPA over various test cycles in the Ann Arbor, Michigan area. The test
matrix covered a range of vehicle operation that included:

1.	Two idling tests of 15- or 30-minutes duration

2.	Seven on-road driving routes that cover the full range of power demand by including transient low-
and medium-speed urban driving to steady-state high-speed highway driving

3.	Soak times ranging from zero minutes (hot start) to 720 minutes (cold start)

4.	Vehicle weight at low or high (Table 3-8)

32	The engine-certified standards for heavy-duty gasoline were reduced by 93% (THC), 80% (NOx), and 61% (CO) with
the MY 2008 standard.

33	The additional rate adjustments for MY2018-2060 LHD2b3 and MY2027-2060 LHD45/MHD/HHD vehicles are
described in subsections 3.1.1.3.1 and 3.1.1.3.2 respectively.

161


-------
5.	Air conditioning on or off

6.	Cabin windows down or up

A total of 202 tests across vehicles and operation modes was available for data analysis. These tests
covered about 412,000 seconds of post-QA operation. We removed the effect of warm and cold starts
from the operation since the running emissions are intended to be just the hot running operation; details
are discussed in Appendix 1.1. After removal of vehicle operation related to start emissions, the final data
set used for just the hot running emissions rates update was about 390,000 seconds.

The following steps were used to calculate the operating mode-based emission rates for each age and
regulatory class of LHD, MHD, and HHD:

1.	Assign operating modes as per the method described above for diesel vehicles and calculate the
average rate per operating mode per test per vehicle

2.	Calculate the average operating mode-based rate per vehicle (using only vehicle specific tests)

3.	Estimate emission rates for operating modes with limited or missing data.

4.	Calculate the operating mode-based emission rate as the production weighted average of the
three test vehicles.

5.	Adjust emission rates by vehicle age.

In Step 1, the operating modes (Table 1-4) were assigned to the 1-hz data using the STP equation (Equation
1-6) with road-load coefficients for single-unit short-haul truck (sourceType 52) for the 2014-2020 model
year range as defined in the sourceusetypephysics table in MOVES3 database. The coefficients for single-
unit short-haul trucks are the same for all the regulatory classes within this source type (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/scofe values (Table 1-3) of 5 (LHD2b3 and LHD45), 7
(MHD), and 10 (HHD). The selection of these new/sco/e 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/sco/e 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.

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

162


-------
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.34

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/sco/e 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/scofe 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/sco;e 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 2018.35 To calculate
emission rates for MY 2018 and later, we applied reductions representing the Tier 3 phase-in for MY 2018-
2022 for LHD2b3 vehicles (as shown in Table 3-9) to the emission rates representing MY 2010-2017
estimated from the above. The reductions for each model year during the phase-in were estimated by
extracting the corresponding MOVES rates for MY 2007-2022, and calculating the fractions relative to MY
2017. The basis and rationale for the Tier 3 reductions for gasoline LHD2b3 vehicles developed for the Tier
3 rulemaking are documented in the MOVES2014 heavy-duty exhaust report.109

The LHD2b3 MY 2018+ rates contain the same heavy-duty gasoline age effects as were applied to the MY
2010-2017 rates (Table 3-3). The resulting emission rates for THC, CO and NOx are shown in Figure 3-7
through Figure 3-9.

34	Sales of Class 2b gasoline trucks are much larger than for Class 3, 4, 5, and 6.

35	All LHD2b3 chassis-certified complete vehicles are subject to Tier 3. All LHD2b3 gasoline fueled vehicles are chassis-
certified complete vehicles.

163


-------
Table 3-9 Tier 3 Reductions by Model Year for Gasoline LHD2b3

Model Year

THC

CO

NOx

2018

35%

38%

41%

2019

44%

48%

52%

2020

53%

59%

63%

2021

62%

68%

74%

2022-2060

71%

78%

85%

Beginning in MY 2027, criteria pollutant rates for medium-duty vehicles are regulated by the LMDV2027
standards. The details on our updates to MOVES to account for LMDV2027 can be found in the LD Exhaust
Report.

3,1,1,3,2 LHD45, MHD, HHD 2027-2060 Model Years

In order to account for the HD2027 standards in MOVES4, we revised the running exhaust emission rates
for NOx, THC, CO, (and PM25 described later in 3.1.2.3) for MY2027+ heavy-duty (LHD45, MHD, HHD)
gasoline vehicles using the methodology described in this section. Unlike the HD2027 standards for diesel
vehicles, the rule does not include off-cycle standards for gasoline vehicles. Due to the relatively simple
treatment of aging for MOVES HD gasoline vehicles, we did not estimate any impact from the lengthened
warranty and useful life periods provisions on the emission rates.

The FTP duty-cycle standards shown in Table 2-9 apply to both heavy-duty compression-ignition engines
and heavy-duty spark-ignition engines. We updated the NOx exhaust emission rates for gasoline, assuming
that emissions are reduced for all operating modes based on the reduction in the NOx FTP standards from
the current 0.2 g/hp-hr standard. Table 3-10 shows the estimated reduction in NOx emission rates, which is
consistent with the ratio of the MY2010 FTP emission standards and the HD2027 final FTP standards shown
in Table 3-10.

In addition, we also estimated emission rate reductions due to the HD2027 standards for HC and CO. We
estimated reduced THC and CO emission rates assuming that those emissions would be reduced due to
improvements in the three-way catalyst emission controls. We used available data from production HD
Spark-ignition (SI) engines and from the heavy-duty gasoline technology demonstration program to
estimate our modeled emissions levels.110 We assumed a 65 percent reduction in THC emissions would
occur at a NOx standard of 0.1 g/hp-hr. We assumed additional decreases in THC emissions to reflect
tighter final NOx standards in MY 2027. We derived Equation 3-1 assuming a linear decrease in THC
emissions between the estimated THC emissions emitted at the 0.1 g/hp-hr NOx FTP level, and zero THC
emissions at a hypothetical 0 g/hp-hr NOx FTP level. We then used Equation 3-1 to estimate the reductions
in THC emissions using the NOx levels for the control scenarios (Table 3-10).

164


-------
^gasoline,THC,NOx FTP

I N Ox FTP Standard \ /-in	^

= 1 —	n	X (A ~~ ^gasoline,THC,0.1 NOx FTP )

\ 0.1 , , y ,	/	Equation 3-1

\	bhp ¦ hr /

. (NOx FTP Standard\	^r-n/\

= 1"(—	) X (1 - 65o/o)

\	bhp-hr /

Where:

*	/?gasoline,THC, NOxftp = percent emission reductions in heavy-duty gasoline THC emissions for
NOx FTP standards more stringent than the 0.1 NOx FTP standard, calculated values shown
in Table 3-10

•	NOx FTP Standard = HD2027 NOx FTP standards

We assumed a 60 percent reduction in CO for MY2027+ engines (see Table 3-10) based on EPA testing.

Table 3-10 Running Emission Rate Reductions From Heavy-duty Gasoline Vehicles Due to HD2027Standards,
Rgasoline, Across All Heavy-duty Gasoline Regulatory Classes and Operating Modes

Regulatory
ClassA

FTP/SET NOx
standard (g/hp-hr)

NOx

THC

CO

LHD, MHD,
HHD

0.035

82.5%

87.8%

60%

1 We applied the same standards to represent the SI engines modeled by t

ie LHD, MHD, and HHD

regulatory classes, unlike the final standards for compression-ignition engines

Then Equation 3-2 was used to revise the MOVES emission rates to account for the HD2027 standards.
Since spark-ignition engines are not subject to the HDIUT program, we did not estimate operating mode-
specific effectiveness of reductions of the in-use emissions compared to duty-cycle standard emissions, as
was done for diesel running emissions. Instead, we assumed these reductions apply uniformly across all
running exhaust operating modes.

ERpinal Standards (l ^gasoline) ^ ^^MOVES_baseline

Equation 3-2

Where:

•	ERcontroi — MOVES running exhaust emission rates for HD2027 standards based on the
reduction in the FTP duty-cycle standard

•	Rgasoline~ percent emission reductions in heavy-duty gasoline emissions from Table 3-10

•	ERmovesbaseline = MOVES running exhaust emission rates before the rate revision

3,1,1,4 Inspection and Maintenance Program Effects for LHD2b3 Gasoline Vehicles
In MOVES3.1, we updated the meanBaseRatelM values for THC, CO, and NOx running exhaust emission
rates to better reflect our understanding how state and local inspection and maintenance (l/M) programs
reduce gaseous emissions from LHD2b3 gasoline vehicles.

As background, in addition to the meanBaseRates described in Section 1, the MOVES emissionRateByAge
table includes a field, meanBaseRatelM, that is used to estimate emissions under a relevant l/M program.
These calculations are explained in more detail in the MOVES Adjustments report.

165


-------
In MOVES3.0.4 and earlier versions of MOVES, we set the values of meanBaseRatelM for all HD exhaust
rates to equal the associated meanBaseRate—essentially assuming no benefit from HD l/M programs. For
MOVES3.1 and later, we reconsidered this choice for gasoline LHD2b3 trucks (regClass 41). Given the
similarity of the engine technology and aftertreatment systems between gasoline LHD2b3 and light-duty
trucks (regClass 30), and the similarity in the way these trucks are tested in contemporary l/M programs,
we updated the gasoline LHD2b3 HC, CO and NOx running exhaust values for meanBaseRatelM to reflect
the same proportional reduction (that is meanBaseRatelM/meanBaseRate) that we model for each
operating mode bin and age for the light-duty trucks. For more detail on the meanBaseRate and
meanBaseRatelM values for light-duty trucks, see the MOVES3 LD report.

3,1,1,5 Model Year Trends

Figure 3-7 through Figure 3-9 display the THC, CO, and NOx non-IM running exhaust emission rates by
model year and regulatory class (HHD and Urban Bus). The emission rates are estimated in grams per mile
(g/mile) using nationally representative operating mode distributions and average speeds. The model year
groups used to estimate the emission rates are evident: 1950-1989, 1990-1997, 1998-2007, 2008-2009,
and 2010-2060. Note that not all the changes in the gram per mile emission rates are due to changes in the
operating mode specific emission rates. For example, the MY 1995-1997 operating mode specific emission
rates are the same as the 1990-1995 emission rates for all regulatory classes. However, there is an
observed spike in the HHD gram per mile THC emission rate, which is attributed to a shift in the
distribution of HHD gasoline activity among different source types in MOVES.

Figure 3-7 shows that the THC emission rates follow decreasing trends with model year that correspond
with tighter emission factors. The drop-in emission rates in model year 2008-2009 is attributed to the
different methodology used to develop those rates discussed earlier. Even though the increasing trend
between 2008-2009 and 2010 and later model year groups may not be intuitive from a technical
perspective, we have increased confidence that the MY 2010 and later THC emission rates represent the
real-world emissions since they were developed based on in-use testing of MY 2010 and later vehicles.
Emission rates also change starting in MY2027 due to the implementation of the HD2027 standards.

166


-------
Model Year

Figure 3-7. Base running emission rates for THCfrom age 0-3 gasoline heavy-duty vehicles averaged over a nationally

representative operating mode distribution.

The CO emission rates are shown in Figure 3-7 The CO emission rates for LHD2b3 vehicles (the largest
regulatory class of heavy-duty gasoline) follow a generally decreasing trend with model year. The trends
for LHD45 and MHD show unexpected variation across model years, including an increase in CO emission
rates for LHD45 and MHD vehicles. We have the most confidence in the most recent model year data, and
the variability in the model year trends reflects uncertainty in the earlier heavy-duty gasoline emission
rates.

167


-------


IS****



n











T























V









Reg Class

-+¦ 41-LHD2b3
42-LHD45

46-MHD67

47-HHD8

1980	2000	2020	2040

Model Year

Figure 3-8. Base running emission rates for CO from age 0-3 gasoline heavy-duty vehicles averaged over a nationally

representative operating mode distribution.

Figure 3-9 shows that the NOx emission rates follow decreasing trends with model years that correspond
with tighter emission standards.

168


-------
Reg Class

-+¦ 41-LHD2b3
42-LHD45

46-MHD67

47-HHD8

1980	2000	2020	2040

Model Year

Figure 3-9. Base running emission rates for NOxfrom age 0-3 gasoline heavy-duty vehicles averaged over a nationally

representative operating mode distribution

3.1.2 Particulate Matter (PM2.5)

The available studies from which to develop PM25 emission for heavy-duty gasoline are particularly limited.
This includes limitations on second-by-second data from which to develop operating mode specific rates,
as well as studies representative of in-use and fleet average emissions. At the same time, heavy-duty
gasoline is a relatively small contributor to the total PM2.5 emissions inventory when compared to heavy-
duty diesel and light-duty gasoline. As a result, the limited analysis conducted for MOVES2010 has been
carried over for the 2009 and earlier vehicles as discussed in Section 3.1.2.1. For MOVES3, we 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 1950-2009 Model Years

For MOVES, the MY 1960-2009 heavy-duty gasoline PM2.5 emission rates were calculated by multiplying
the MOVES2010b light-duty gasoline truck PM2.5 emission rates by a factor of 1.40, as explained below.
When we extended calculations to 40 ages, we duplicated the emission rates for MY 1960 for model years
back to 1950.

Since the MOVES light-duty gasoline PM2.5 emission rates comprise a complete set of factors classified by
particulate sub-type (EC and nonECPM), operating mode, model year and regulatory class, the heavy-duty
PM2.5 emission factors are also a complete set. No change to the PM emission rates is made between MY
2003 and 2009, because the HD 2007 Rule PM standards are not expected to change in-use emissions for
heavy-duty gasoline vehicles. As presented in the next subsection, the simulated age 0-3 HD gasoline MY
1960-2009 emission rates on the UDDS is ~6.6 mg/mile, while the standard for 2008+ spark-ignition
vehicles is 20 mg/mile108

169


-------
3.1.2.1.1 Data Sources

The factor of 1.4 used to convert light-duty gasoline PM rates to heavy-duty rates was developed based on
PM2.5 emission test results from the four heavy-duty gasoline trucks tested in the CRC E55-E59 test
program. The specific data used were collected on the UDDS test cycle. Each of the four vehicles in the
sample received two UDDS tests, conducted at different test weights. Other emission tests using different
cycles were also available on the same vehicles but were not used in the calculation. The use of the UDDS
data enabled the analysis to have a consistent driving cycle. The trucks and tests are described in Table
3-11.

Table 3-11 Summary of Data Used in HD Gasoline PM Emission Rate Analysis

Vehicle

MY

Age

Test cycle

GVWR
[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

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:

170


-------
•	MY group 1983-1984, age 20+

•	MY group 1986-1987, age 15-19

•	MY group 1991-1993, age 10-14

•	MY group 2001, age 0-3

The simulated PM2.5 UDDS emission factors for the older light-duty gas truck group using MOVES2010b are
38.84 mg/mi (ignoring sulfate emissions which are on the order of lxlO"4 mg/mile for low sulfur fuels). This

65.22-Sf

value leads to the computation of the ratio:	^jr = 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 lxlO"5 mg/mile for low sulfur fuels, this

2.71^f

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:

Ratiofinal = RatioolderWtFrac + Rationewer(l - WtFrac)	Equation 3-3

= 1.679x0.75 + 0.578x0.25 = 1.40

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.36

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 /sco;e 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

36 This approximation may 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 tons, which differs from the/Sca/e value of 2.06 metric tons.37 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.

171


-------
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 3-12. No reductions are made between 2003 and
2009, because the 2007 HD rule is not anticipated to cause reductions in heavy-duty gasoline PM2.5
emissions.

172


-------
Table 3-12. Derivation of MHD and HHD Elemental Carbon Emission Rates from LHD2b3 Rates using fscaje 17.1/2.06

THC emission ratios. Using Model Year 2001 as an Example

opModelD

LHD2b3 EC emission
rates (mg/hr)

fscale 17.1/2.06 THC
emission ratios

MHD and HHD EC emission
rates (mg/hr)

0

0.59

1.000

0.59

1

0.54

1.000

0.54

11

0.60

1.000

0.60

12

0.79

2.263

1.78

13

1.38

3.677

5.08

14

2.62

5.095

13.37

15

5.55

5.443

30.22

16

64.52

5.427

350.13

21

8.38

1.000

8.38

22

2.92

1.154

3.37

23

2.08

2.173

4.52

24

2.92

2.825

8.24

25

10.94

4.842

52.95

27

20.50

7.906

162.10

28

126.42

8.796

1,112.05

29

523.16

6.471

3,385.32

30

2,366.75

7.102

16,809.50

33

26.59

2.121

56.40

35

10.76

4.780

51.42

37

13.29

4.010

53.28

38

43.61

8.979

391.56

39

75.73

9.522

721.06

40

74.96

5.300

397.26

The resulting PM2 5 emissions by regulatory class for LHD, MHD and HHD are shown in Figure 3-10. In
general, PM2.5 emission rates are of similar magnitude for each regulatory class between model year 1980
and 2009. There is significant variation in the model years, with some unexpected trends (e.g., LHD45 has
higher emission rates than HHD and MHD for most of these model years). These unexpected trends and

173


-------
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 2010-2026 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 MY 2010 and later THC, CO, and NOx emission
rates (Section 3.1.1.2.2) did not include PM2.5- Lacking appropriate PM data by operating mode, we
populated the MY 2010+ HD gasoline PM2.s rates by copying MY 2010+ HD diesel rates.

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 3.1.1.2.2) over various
chassis-dynamometer tests are shown in Table 3-13 The average PM2.5 rate over all vehicles and test cycles
is 1.35 mg/mi. The average PM2.5 emission rate for MY 2016 age 0-3, LHD diesel (comparable to the tested
gasoline vehicles) using nationally representative operating mode distributions and average speeds is 1.4
mg/mile (See Figure 2-34). Since those numbers were comparable given the uncertainty of the PM2.5
emission rates, and no modal HD gasoline PM2.5 data was available, we decided to use the HD diesel PM2.5
rates for HD gasoline. These rates also include the tampering and mal-maintenance age effects for model
year 2010-2026 (see Appendix B.8).

Table 3-13 PM2,s 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

1 The vehicles are described in section 3.1.1.2.2.

The draft diesel LHD2b3 and LHD45 PM2.5 rates were copied to the gasoline LHD2b3 and LHD45 rates,
respectively, from a MOVES version used for the preliminary HD2027 Rule analysis.108 Since the diesel
MHD rates were notably higher than the diesel LHD and HHD rates, the diesel HHD rates were used for
gasoline MHD and HHD. Note that after this analysis, the heavy-duty diesel PM2 5 emission rates in MOVES
were updated to account for the updated HDIUT sample and model year split described in Section 2.1.2.2.
For this reason, the zero-mile PM2.5 emission rates from heavy-duty gasoline are constant for 2010 and
later model years, whereas the heavy-duty diesel PM2.5 emission rates are reduced starting in model year
2013, and the heavy-duty gasoline rates for 2013 and later are generally higher than the comparable
heavy-duty diesel rates.37 Gasoline engine rates for LHD45 and LHD2b3 are higher than the HHD rates for
all 2010 and later years, whereas this trend is only seen in the initial model years (2010-2013) of the heavy-
duty diesel PM2.5 rates from the HDIUT program (Figure 2-34). We may revisit the HD gasoline rates in the

37 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.

174


-------
future to incorporate any new HD gasoline emissions data, or at least to make them consistent with the
updated MOVES HD diesel rates.

The gasoline rates were copied from the diesel rates as PM2.5, and then allocated to EC and nonECPM using
gasoline-specific fractions based on the Kansas City study of light-duty cars and trucks as described in the
Speciation Report.6 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.

The MOVES heavy-duty gasoline PM2.5 zero mile emission rates are constant for MY 2013 through 2026.
There are differences between the 2010-2012 and 2013-2026 model year groups due to different
tampering and mal-maintenance assumptions applied to the emission rates (see Section 2.1.2.3), which
primarily impact the ages 4-5 and older ages. The Tier 3 rulemaking sets PM FTP emission standards for
Class 2b and Class 3 of 8 mg/mile and 10 mg/mile, respectively, which began phase-in starting with model
year 2018 vehicles.51 We did not model reductions in the gasoline PM2.5 emission rates with the phase-in of
Tier 3, because the data on the tested heavy-duty gasoline vehicles (Table 3-13) suggests that the heavy-
duty gasoline vehicles are well in compliance with the Tier 3 standard. In addition, the diesel rates on
which the gasoline rates are based also are well in compliance with the Tier 3 standards as discussed in
Section 2.1.1.5.5.

3.1.2.3	2027-2060 Model Years

In order to account for the HD2027 standards in MOVES4, we revised the PM2 5 running exhaust emission
rates for MY2027+ heavy-duty (LHD45, MHD, HHD) gasoline vehicles using the methodology described in
3.1.1.3.2 for other criteria pollutants.

To meet the PM standards (5 mg/hp-hrfor MY2027+), manufacturers are expected to improve fuel control
and limit the need for catalyst protection. Therefore, we assumed a 50 percent reduction in PM2.5,
consistent with the 50 percent more stringent PM standard and revised the MOVES emission rates using
Equation 3-2.

Medium-duty (regulatory class 41, LHD2b3) vehicles are subject to the LMDV2027 standards for PM
beginning in MY 2027. We document our inclusion of the LMDV2027 standards in MOVES in the LD Exhaust
Report.

3.1.2.4	Model Year Trends

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

175


-------
0.20

0.15

(15

en

O)

5
0.

0.10

0.05

0.00





















_

hrt









\ 
-------
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.

0.03-

0.02

0.01

EC

NonEC

0.00

Figure 3-12 Heavy Duty Gasoline Running Exhaust PM2.5 Emission Rates by Elemental Carbon and Non-Elemental
Carbon (nonEC) Fractions for the 0-3 Age Group by Model Year and Regulatory Class using Nationally Representative

Operating Mode Distributions

3.1.3 Energy

3.1.3.1 1950-2009 Model Years

When we extended calculations to 40 ages, we duplicated the emission rates for MY 1960 for model years
back to 1950.

3.1.3.1.1	LHD

The energy rates for gasoline LHD (LHD2b3 and LHD45 regulatory classes) pre-2009 energy rates are
unchanged from MOVES2010a. In MOVES2010a, the energy rates for LHD2b3 and LHD45, along with the
light-duty regulatory classes, were consolidated across weight classes, engine size and engine technologies,
as discussed in the MOVES2010a energy updates report57.

3.1.3.1.2	MHD and HHD

The energy rates for gasoline MHD and HHD pre-2009 energy rates are unchanged from MOVES2014. The
rates were developed using the same data set we used to develop the THC, CO, and NOx exhaust emission
rates. Similar to the analysis for the diesel running exhaust energy rates, we made no distinction in rates by
model year, age, or regulatory class. To calculate energy rates (kJ/hour) from C02 emissions, we used a
heating value (HV) of 122,893 kJ/gallon and C02 fuel-specific emission factor (fC02) of 8,788 g/gallon for

177


-------
gasoline (see Equation 3-20). STP was calculated using Equation 1-6. Figure 3-13 presents the gasoline
running exhaust energy rates in MOVES for these regulatory classes.

6 -i

c
o

5 -

4 -

u
+•>
re

S?

v
c

u

1

3 -

1 -

0

—i	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

Operating mode

Figure 3-13. Gasoline Running Exhaust Energy Rates for MHD (1960-2009) and HHD (1960-2009)

A linear extrapolation to determine rates at the highest operating modes in each speed range was
performed analogously to diesel energy and NOx rates (see Section 2.1.1.4.2).

3.1.3.2 2010-2060 Model Years

The real-world PEMS-based emissions measurement data from two engine-certified and one chassis-
certified heavy-duty gasoline vehicles used to update the THC, CO, and NOx emission rates (Section
3.1.1.2.2) included C02 emissions data which was used to update the energy rates. The energy rates are
derived using the measured C02 values and the conventional gasoline specific values for carbon content
(0.0196 g/KJ) and oxidation fraction (1.0) and the molecular mass of C02 (44), and atomic mass of Carbon
(12). These values are described in the MOVES GHG and Energy Rates report.

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.

3.1.3.2.1 LHD2b3 2014-2060 Mode! 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

178


-------
rates using the gasoline reductions documented in Table 3-14 (Phase 1) and Table 2-35 (Phase 2) in
Section 2.1.4.3.

Table 3-14 Estimated Total Vehicle Reductions in Energy Consumption Rates for LHD2b3 Gasoline Vehicles due to the

HD GHG Phase 1 Program

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 HD GHG Phase 2 reductions for gasoline 2b3 vehicles are identical to diesel 2b3 vehicles, which are
described in Section 2.1.4.3.2 and listed in Table 2-35.

3,1,3,2,2 LHD45, MHD, and HHD 2016-2060 Model Years

Updates to the energy rates were made to the heavy-duty gasoline energy rates for model years 2016-
2020 based on the Phase 1 Medium and Heavy-Duty Greenhouse Gas Rule111 discussed in Section 2.1.4.3
and shown in Table 3-15.

Table 3-15 Heavy-Duty Gasoline Reductions due to the Heavy-Duty GHG Phase 1 Rule111

Regulatory Class

Model
Years

C02 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 C02 emission
reductions expected from the Heavy-Duty GHG Phase 2 rule, as shown in Table 3-16, which have separate
reductions for vocational and combination trucks.

As noted above, the HD GHG Phase 2 reductions to energy rates are not incorporated into the energy rates
in the emissionRate table in the MOVES database, but are applied at run-time using the values in the
emissionRateAdjustment table in the MOVES database.

179


-------
Table 3-16 Heavy-Duty Gasoline Reductions due to the Heavy-Duty GHG Phase 2 Rule112

SOURCE TYPE
(SOURCETYPEID)

REGULATORY CLASS

MODEL YEARS

C02 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%

We updated HD gasoline energy consumption rates in MOVES5 to account for the HD GHG Phase 3 rule. To
do so, we applied the same methodology as we did for diesel, which is discussed in Section 2.1.4.3.1.

3.1.3.3 Model Year Trends

Figure 3-14 and Figure 3-15 display the C02 (g/mile) emission rates and fuel economy values calculated
from the energy rates using the carbon content and energy density conversion factors for conventional
gasoline38 as documented in the MOVES3 Greenhouse Gas and Energy Report.The C02 (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; this is an artifact of using updated data and analysis.

The LHD emission rates show substantial variability in the early model years (pre-1985) - we do not expect
LHD vehicles to have lower fuel economy than MHD and HHD for these years but have not revisited these
emission rates due to the small number of pre-1985 gasoline vehicles remaining in the onroad fleet. As
discussed in Section 2.1.4.1, the detailed methodology used in MOVES2004 (which modeled different
emission rates according to vehicle weights, engine technologies, and engine sizes) introduced variability
into the energy rates within the current MOVES regulatory class emission rates for pre-2010 LHD.

The figures show that, since model year 2010, there are decreasing trends in C02 (g/mile) with
corresponding increases in fuel economy, due to the lower energy rates as well as lower source 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 have little change for model years 2027-
2060. However, some small differences in C02 (g/mile) or fuel economy values observed within model year

38 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 C02 g/mile.

180


-------
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.

Reg Class

41-LHD2b3

42-LHD45

46-MHD67

47-HHD8

1980	2000	2020	2040

Model Year

Figure 3-14. Base running emission rates for C02from age 0-3 gasoline heavy-duty vehicles averaged over a nationally

representative operating mode distribution

181


-------
20

Reg Class

-+¦ 41-LHD2b3

42-LHD45

47-HHD8

46-MHD67

0

1980

2000

2020

2040

Model Year

Figure 3-15. Fuel economy for age 0-3 gasoline heavy-duty vehicles averaged over a nationally representative

operating mode distribution

3.2 Start Emissions

Representative in-use data on vehicle start emissions for heavy-duty gasoline vehicles is even less common
than running data. While some data was available (Table 3-18, Table 3-23), the MOVES analysis also relies
on deterioration patterns from light-duty vehicles, as well as ratios to the relevant engine emission
standards. For LHD2b3 gasoline vehicles, manufacturers comply with chassis (g/mile) emission standards.
For the larger regulatory classes, engine emission standards apply. We used the engine emission standards
to estimate differences in emissions between the LHD2b3 regulatory class and the heavier regulatory
classes. Most of this analysis has been carried over from MOVES2010b and MOVES2014, but the cold start
emissions for LHD45, MHD, and HHD gasoline engines of 2008 model year and later were updated for
MOVES3 based on certification data.

The heavy-duty spark ignition engine emissions standards113 for the Federal Test Procedure (FTP) are
shown in Table 3-17. Note that the standards for model years 1990 through 2004 for CO and THC vary by
weight class, but not by model year, whereas those for NOx vary by model year, but not by weight class.
Also, for model years 2005-2007, a single standard is applied for NMHC+NOx, but by 2008, separate but
lower standards are again in effect. Also note that by model year 2008, the standards for the three gaseous
pollutants are the same across regulatory class.

Table 3-17 FTP Standards (g/hp-hr)for Heavy-Duty Gasoline Engines for Model Years 1990-2008+113

182


-------
Table 3-17 FTP Standards (g/hp-hr) for Heavy-Duty Gasoline Engines for Model Years 1990-2008+113

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

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 1950-2004 Model Years

To develop start emission rates for MY 1960-2004 heavy-duty gasoline-fueled vehicles, we extracted data
available in EPA's Mobile-Source Observation Database (MSOD).104 These data represent aggregate test
results for heavy-duty spark-ignition (gasoline powered) engines measured on the Federal Test Procedure
(FTP) cycle. The GVWR for all trucks was between 8,500 and 14,000 lbs, placing all trucks in the LHD2b3
regulatory class. The 1960-2004 LHD2b3 start rates are unchanged from LHD2b3 start emission rates in
MOVES2010b.

Table 3-18 shows the model-year by age classification for the data. The model year groups in the table
were designed based on the progression in NOx standards between MY 1990 and 2004. Standards for CO
and THC are stable over this period, until MY 2004, when a combined NMHC+ NOx standard was
introduced. However, no measurements for gasoline HD trucks were available for MY2004 and later.

Start emissions are not dependent on power, and therefore, the emission rates do not need to be
calculated differently to distinguish different/sco;e values as was done for running exhaust rates. As
discussed later, start emission rates are separated by regulatory classes to account for differences in the
emission standards and/or available test data.

183


-------
Table 3-18 Availability of Emissions Start Data by Model-Year Group and Age Group for LHD2b3 Vehicles

Model-year Group

Age Group (Years)

Total

0-3

4-5

6-7

8-9

10-14

1960-1989







19

22

41

1990





1

29



30

1991-1997

73

59

32

4



168

1998-2004

8









8

Total

81

59

33

52

22

247

3.2.1.1.2 Estimation of Mean Rates

As with light-duty vehicles, we estimated the "cold-start" as the mass from the cold-start phase of the FTP
(bag 1) less the "hot-start" phase (Bag 3). As a preliminary exploration of the data, we averaged by model
year group and age group and produced the graphs shown in Appendix F. Sample sizes were small overall
and very small in some cases (e.g., 1990, age 6-7) and the behavior of the averages was somewhat erratic.
In contrast to light-duty vehicle emissions, strong model-year effects were not apparent. This may not be
surprising for CO or THC, given the uniformity of standards throughout. This result was more surprising for
NOx, but model year trends are no more evident for NOx than for the other two. Broadly speaking, it
appeared that an age trend may be evident.

If we assume that the underlying population distributions are approximately log-normal, we can visualize
the data in ways that illustrate underlying relationships. As a first step, we calculated geometric mean
emissions, for purposes of comparison to the arithmetic means calculated by simply averaging the data.
Based on the assumption of log-normality, the geometric mean (xg) was calculated in terms of the
logarithmic mean (x/) as shown in Equation 3-4.

Xg = elnxi	Equation 3-4

This measure was not appropriate for use as an emission rate but was useful in that it represents the
"center" of the skewed parent distribution. As such, it was less strongly influenced by unusually high or
outlying measurements than the arithmetic means. In general, the small differences between geometric
means and arithmetic means suggest that the distributions represented by the data do not show strong
skew in most cases. Because evidence from light-duty vehicles suggested that emissions distributions
should be strongly skewed, this result implied that these data are not representative of "real-world"
emissions for these vehicles. This conclusion appeared to be reinforced by the values in Figure F-3which
represent the "logarithmic standard deviation" calculated by model-year and age groups. This measure (s/),
is the standard deviation of natural logarithm of emissions (x/). The values of s/ 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 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 s/. For

184


-------
this calculation, we assumed values of 0.9 for CO and THC and 1.2 for NOx. These values approximate the
maxima seen in these data.

The re-estimated arithmetic means were calculated from the geometric means, by adding a term that
represents the influence of the "dirtier" or "higher-emitting" vehicles, or the "upper tail of the
distribution," as shown in Equation 3-5.

_ 4r	Equation 3-5

xa = xge 2

For purposes of rate development using these data, we concluded that a model-year group effect was not
evident and re-averaged all data by age group alone. Results of the coarser averaging are presented in
Figure 3-16 with the arithmetic mean (directly calculated and re-estimated) and geometric means shown
separately.

185


-------
o i i i I i i i I i i i I i i i I i i i I i i i I i i i

0	2	4	6	8 10 12 14

Age (years)

16

0	2	4	6	8	10 12 14

Age (years)

8i

Age (yeans)

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 sj

We then addressed the question of the projection of age trends. As a general principle, we did not allow
emissions to decline with age. For THC and NOx, we assumed the emission rates stabilized at the maximum
level reached at the 6-7 and 8-9 age groups, respectively as shown in Table 3-19. For CO emissions, we

186


-------
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 s/, it was
necessary to re-estimate corresponding standard deviations for the parent distribution s, as shown in
Equation 3-6.

s = lx*es2(es2 - 1)	Equation 3-6

After recalculating the standard deviations, the calculation of corresponding standard errors was simple.
Because each vehicle is represented by only one data point, there was no within-vehicle variability to
consider, and the standard error could be calculated as s/^Jn. We divided the standard errors by their
respective means to obtain CV-of-the-mean or "relative standard error." Means, standard deviations and
uncertainties are presented in Table 3-19 and in Figure 3-17. Note that these results represent only "cold-
start" rates (opModelD 108). Soak time adjustments other start opModes are detailed in Section 3.2.3.

Table 3-19. 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

187


-------
10	15

Age (years)

20

25



18



16



14

g 12

£

10

¦o

8

o



(J

6

Q.



t

4



2



0

(b



; ino























	^

































10	15

Age (years)

20

25

12

10 -

(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.

When we extended calculations to 40 ages, we duplicated the emission rates for MY 1960 for model years
back to 1950.

188


-------
3,2,1,1,4 2005-2007 Model Years

For the 2005-2017 model year emission rates, we applied reductions to the 1960-2004 emission rates, by
comparing the standards between the two model year ranges. For CO, the approach was simple. We
applied the age zero values in Table 3-19 to the 2005-2007 model year group. The rationale for this
approach is that the CO standards do not change over the full range of model years considered.

For THC and NOx, we imputed values for the 2005-2007 and 2008-2017 model-year groups by multiplying
the age zero values for the 1960-2004 emission rates in Table 3-19 by ratios expressed in terms of the
applicable standards. Starting in 2005, a combined THC+NOx standard was introduced. It was necessary for
modeling purposes to partition the standard into THC and NOx components. We assumed that the
proportions of NMHC and NOx would be similar to those in the 2008 standards, which separate NMHC and
NOx while reducing both. We calculated the THC value by multiplying the 1960-2004 value by the fraction
/HC as shown in Equation 3-7.

fHC

0.14 g/hp -hr \ ,. _	.

(0.14 + 0. 20) g/hp - hr) (1' 0 9^hp ftr)	Equation 3-7

1.1 g/hp — hr

= 0.37

This ratio represents the component of the 2005 combined standard attributed to NMHC. We calculated
the corresponding value for NOx as shown in Equation 3-8.

fNOx

'	0-20 g/hp-hr	\	_ hr

^(0.14 + 0. 20) g/hp — hr) '	^	Equation 3-8

4.0 g/hp — hr

= 0.147

For these heavy-duty rates, we neglected the THC/NMHC conversions, to which we gave attention for
light-duty.

3,2,1,1,5 2008-2017 Model Years

For the 2008-2017 model years, the approach to projecting rates was modified to adopt two refinements
developed for light-duty rates. First, start emission rates for the LHD2b3 gasoline vehicles were estimated
from composite rates by applying the "start split-ratio" shown in Table 3-6 to a set of rates representing
light-duty trucks in Tier-2/Bin 8. Second, we updated the deterioration effects for start NOx exhaust from
MOVES2014114, by applying the ratios shown in Table 3-20.

189


-------
Table 3-20. Deterioration Reduction Ratio for 2008-2017 gasoline LHD2b3 NO, Starts

ageGroupID

Deterioration Reduction
Ratio

3

1.00

405

0.85

607

0.79

809

0.73

1014

0.62

1519

0.62

2099

0.62

These ratios were initially developed for LDT in a draft version of MOVES3115 but the LD age effects were
further updated prior to MOVES3 release.39

For THC and CO, the multiplicative age effects are unchanged from the effects from MOVES2014.116 The
resulting multiplicative age effects for start emission rates for LHD2b3 vehicles used in MOVES3 for model
years 2009-2017 are shown in Table 3-21. The start emission rates for model year 2008 are estimated by
averaging the MY 2007 and 2009 emission rates across all age groups and operating modes assuming a
phase-in of 50% of the Tier 2 standards and the HD 2007 Rule in MY 2008 as we assumed for LHD2b3
gasoline running emissions as discussed in Section 3.1.1.2.1. The relative age effects for LHD2b3 MY 2008
and MY 2009-2017 are shown Figure 3-21.

Table 3-21 Multiplicative Age Effect Used for Start Emissions for Gasoline LHD2b3 Vehicles for 2003-2017 Model Years

Adopted from the Deterioration Effects for Light-Duty Trucks

ageGroupID

THC

CO

NOx

3

1

1

1

405

1.65

1.93

1.47

607

2.20

2.36

1.74

809

2.68

2.54

2.01

1014

3.30

3.00

2.00

1519

3.66

3.35

2.26

2099

4.42

4.06

2.56

Using these deterioration rates for starts results in start emission rates for MY 2010+ gasoline LHD2b3
vehicles having a higher relative deterioration than running emission rates (compared to Table 3-3).40 We

39	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.12

40	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.

190


-------
recognize this is inconsistent with our knowledge of light-duty start deterioration.12 We plan to address
this data gap with data collected on LHD2b3 in future versions of MOVES.

3,2,1,1,6 Incorporating Tier 3 Standards: 2013 and Later Model Years

Emission rates representing the phase-in of Tier-3 standards for the start-exhaust process were developed
for MOVES2014 as described in gasoline running emissions section of the MOVES2014 heavy-duty exhaust
report.109 Like the MY 2008-2017 rates, the LHD2b3 Tier 3 start rates are based on light-duty truck emission
rates scaled to higher emission standards for the LHD2b3 regulatory class. The reduction in start emissions
due to Tier 3 is relatively lower than the reductions in running emissions presented in Section 3.1.1.3.1.

The LHD2b3 start rates during and following the Tier 3 phase-in have relatively lower deterioration than
the start rates for the model years preceding the onset of the phase-in (MY 2008-2017) as documented in
the MOVES2014 light-duty exhaust report.11641 For MOVES3, we adjusted the NOx start emission rates by
applying the deterioration ratios in Table 3-20 to the MOVES2014 NOx start rates. The multiplicative age
effects for LHD2b3 cold start rates for THC, CO and NOx after the complete phase-in of Tier 3 phase-in
model year 2022 are shown below in Table 3-22 The age effects of the phase-in years of Tier 3 (MY 2018-
2021) are a weighted average of the MY 2010-2017 and the MY 2022 start emission rates using the phase-
in assumptions documented in the MOVES2014 heavy-duty exhaust report.109

Table 3-22 Multiplicative Age Effect Used for Start Emissions for Gasoline LHD2b3 Vehicles for 2022-2060 Model Years

ageGroupID

THC

CO

NOx

3

1

1

1

405

1.54

1.73

1.38

607

1.94

1.97

1.57

809

2.26

1.96

1.74

1014

2.78

2.33

1.72

1519

3.09

2.59

1.95

2099

3.73

3.15

2.21

We do not model any impact of the HD2027 rule on the Gasoline LHD2b3 start emissions because HD2027
standards affect only engine-certified gasoline light-heavy-duty vehicles (modelled as regulatory class
LHD45).

3,2,1,1,7 Inspection and Maintenance Program Effects for LHD Gasoline Vehicles

In MOVES3.1, we updated the meanBaseRatelM values for THC, CO, and NOx start exhaust emission rates

to better reflect the l/M programs for LHD2b3 gasoline vehicles.

Due to a lack of data and analysis, in MOVES3.0.4 and earlier versions of MOVES, we assumed no benefit
from l/M programs for starts. However, as explained in for running emissions in Section 3.1.1.3.2, we

41 In MOVES3, the deterioration effects for all model year light-duty vehicles were updated using updated data and
analysis. The light-duty Tier 3 emission rates no longer have different deterioration values. We may update the
LHD2b3 start deterioration effects to be consistent with the light-duty vehicles in an upcoming version of MOVES.

191


-------
updated the gasoline LHD2b3 HC, CO and NOxstart values for MeanBaseRatelM in MOVES3.1 and later to
reflect the same proportional reduction (that is meanBaseRatelM/meanBaseRate) that we model for each
operating mode bin and age for the light-duty trucks. For more detail on the meanBaseRate and
meanBaseRatelM values for light-duty trucks, see the MOVES Light Duty Report.12

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 1950-2007 Model Years

Since bag data were lacking for MY 1950-2007 vehicles in classes LHD45 and MHD, we estimated cold start
values relative to the LHD2b3 start emission rates.

For CO and THC, we estimated rates for the heavier vehicles by multiplying them by ratios of standards for
the heavier class to those for the lighter class. The value of the ratio for CO based on 1990-2004 model
year standards is shown in Equation 3-9.

37.1 g/hp — hr

fCO

11 4 g/hp - hr	Equation 3-9

= 2.58

The corresponding ratio for THC for 1990-2004 model year vehicles is 1.73, as shown in Equation 3-10.

1.9 at hp — hr	^

f =	a/ 	= l. 73	Equation 3-10

,HC 1.1 g/hp -hr

The ratios derived in the previous two equations (2.58 and 1.73) were applied to estimate the start
emission rates for 1960-2004 and 2005-2007 model year groups for the LHD45, MHD, and HHD gasoline
vehicles (Table 3-25 ). Note that the ratios for CO and THC do not vary by model year group because the
standards do not; See Table 3-16.

For MY 1950-2007, NOx start emission rates for medium and heavy-duty vehicles are equal to the LHD2b3
start emission rates, because the same standards apply to all the HD regulatory classes. The approaches for
all three regulatory classes in all model years are summarized in Table 3-25 .

3,2,1,2,2 2008-2060 Model Years

The cold start emissions for 2008 model year and later LHD45, MHD, and HHD gasoline engines have been
updated for MOVES3 based on new data. Similar to the approach taken for light-duty vehicles and for
diesel vehicles (see Section 2.2.1.2), the cold start emissions are calculated as the difference in emissions
between a test cycle with a cold start and the same test cycle with a hot start. Heavy-duty gasoline engines
are certified using the Heavy-Duty Gasoline Engine Federal Test Procedure (FTP) cycle.117 The test
procedure for certification requires that manufacturers run the engine over the FTP cycle with a cold start
and then repeat the cycle with a warm start. Starting in model year 2016, EPA began collecting certification
data that contained separate cold and hot results for each engine certified. The data that was analyzed for
this MOVES3 update includes the following engine families from the 2016 and 2017 model years shown in
Table 3-23.

192


-------
Table 3-23 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 3-11.

Grams per Start = [Cold FTP Emission Results (g/(hp — hr))

— Hot FTP Emission Results(g/(hp — hr))]	equation 3-

* FTP Cycle Work (hp — hr)

The amount of work (hp-hr) performed over the FTP cycle is not provided as part of the certification data
submitted by the manufacturers to EPA. We only had cycle work data from one 19.3 hp-hr HD gasoline
engine. While we acknowledge that FTP cycle work is unique to each engine because it is created based on
the engine's maximum speed, curb idle speed, and the maximum torque curve, we estimated cycle work
for all HD gasoline engines using our one engine data source.

The analysis of cold and hot start FTP emissions data from three 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 3-24. The MY 2016 and 2017 engines ranged
in displacement between 5.4 and 7.2 liters, and ranged in rated power between 297 and 332 HP. The new
default cold start emissions values for MOVES3 are the mean values shown in Table 3-24. 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 3-24 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 PM25 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 3-21 and Table 3-22.42 The start rates for THC, CO, and NOx for this model
year group for each age are graphed in Figure 3-21.

We do not model any impact of the HD2027 rule on the Gasoline LHD45, MHD or HHD start emissions due
to the lack of sufficient data to model the impact.

42 In a future update to MOVES, we may 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.

193


-------
3,2,1,3 Summary

Table 3-25 summarizes the data and methods used to estimate THC, CO, and NOx start emission rates from
heavy-duty gasoline vehicles as discussed in Sections 3.2.1.1 and 3.2.1.2. Figure 3-18 through Figure 3-20
displays the cold start (operating mode 108) emission rates across model years for heavy-duty gasoline
vehicles.

Table 3-25 Summary of Cold Start Emission Rates for Heavy-Duty Gasoline Vehicles

Regulatory
Class

Model Year
Group

CO

THC

NOx

LHD2b3

1950-2004

Data analysis, values from Table 3-19

2005-2007

Data analysis, values
from Table 3-19

Reduce in proportion
to standards from 1960-2004

2008 - 2017

Section 3.2.1.1.5 Based on Tier 2 Bin 8 LDT rates and deterioration

2018 +

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

LHD45,
MHD, HHD

1950-2004

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. No change for HD2027 rule.

Updated based on FTP
certification data,
deterioration based
on the 2008 + LHD2b3
vehicles. No change
for HD2027 rule.

Note: soak time adjustments are detailed in Section 3.2.3.

The outcomes of the methods described in the table above are summarized graphically in Figure 3-18
through Figure 3-20 for cold-start emissions. The decline in start emissions with the adoption of more
stringent standards begins with the reduction in model year 2005 and ends at the completion of the phase-
in of Tier 3 standards for LHD2b3 vehicles in model year 2022. Note that there is a slight increase in THC
start emissions for LHD45 vehicles in model year 2008, which is the first model year using the new start
certification data discussed above in Section 3.2.1.1.5.

194


-------
Model Year

Figure 3-18 Heavy-duty Gasoline THC Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class and

Model Year. LHD45, MHD and HHD are equivalent.

200-

2 100-

Reg Class

-- 41-LHD2b3
42-LHD45

46-MHD67

47-HHD8

1980	2000	2020	2040

Model Year

Figure 3-19 Heavy-duty Gasoline CO Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class and

Model Year. LHD45, MHD and HHD are equivalent.

195


-------
Reg Class

41-LHD2b3

42-LHD45

46-MHD67

47-HHD8

1980	2000	2020	2040

Model Year

Figure 3-20 Heavy-duty Gasoline NOx Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class and

Model Year. LHD45, MHD and HHD are equivalent.

2004

2007



2008

2017

2022

0 4 8 12 16 20 240 4 8 12 16 20 240 4 8 12 16 20 240 4 8 12 16 20 240 4 8 12 16 20 24

Vehicle age, years

Reg Class - LHD2b3 LHD45

Figure 3-21 Heavy-duty Gasoline Cold-Start Rates (opModelD 108) vs. Vehicle Age for Select Model Years LHD45, MHD

and HHD are equivalent.

196


-------
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 3-3 in Section 3.1.2.1.2) to scale up start emission rates from light-duty trucks (LDT)
for model years 1960-2003.

For 2004+ model years, the LHD2b3 start emission rates are 1.4 times the model year 2003 LDT emission
rates. We project constant start emissions using the 2003 model year emission rates rather than scaling to
the LDT PM2.5 rates with the 2004 and later model years because the LD rates increase due to the updated
data on emission rates and sales penetration of gasoline direct injection technology, and subsequently,
decrease beginning in model year 2018 with the implementation of the Tier 3 Vehicle Emissions and Fuel
Standards Program. We are not confident that such patterns will apply to HD gasoline due to limited data
regarding heavy-duty PM2.5 rates and uncertainty regarding (a) the expected penetration of gasoline direct
injection technology in heavy-duty gasoline vehicles and (b) the impact of Tier 3 on HD gasoline PM2.5
emissions (see Section 3.1.2.2). We do not model any impact of the HD2027 rule on the Gasoline LHD2b3
start emissions because HD2027 standards affect only engine-certified gasoline light-heavy-duty vehicles
(modelled as the LHD45 regulatory class).

The start PM2.5 emission rates for heavy-duty gasoline vehicles exhibit the same relative effects of soak
time, and deterioration as the LDT PM2.5 start emission rates.

Medium-duty vehicles are subject to the LMDV2027 PM standards beginning in MY 2027. We document
our incorporation of the LMDV2027 standards into MOVES in the LD Exhaust Report.

3.2.2.2	LHD45, MHD, and HHD

Due to a lack of PM2.5 start data, we use the same PM2.5 emission rates for LHD2b3 for all heavy-duty
gasoline for MY 1950-2007. For MY 2008 and later, for LHD45, MHD, and HHD, we updated the PM2.5 start
emissions data using certification data presented in Table 3-24. This causes the start emissions to increase
significantly for LHD45, MHD, and HHD between MY 2007 and MY 2008 as shown in Figure 3-22. We do not
model any impact of the HD2027 rule on the gasoline LHD45, MHD or HHD start emissions due to the lack
of sufficient data to model the impact.

We continue to apply the same age adjustments to the start PM2.5 as the LDT emissions due to lack of data.
We caution there is considerable uncertainty in the start heavy-duty gasoline PM2.5 emission rates,
especially for pre-2007 model years.

3.2.2.3	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.

197


-------
0.00-

1980	2000	2020	2040

Model Year

Figure 3-22. Heavy-duty Gasoline PM2.s Cold-Start Emission Rates (g/start) for Age Group 0-3 By Regulatory Class and

Model Year. LHD45, MHD and HHD are equivalent

E

TO

to

D)

0.09

ra 0.06

-------
HD Gasoline Truck Start Emission Ratio

100 200 300 400 500 600 700 800

Soak Time (min)

New Data NOx —New Data CO —New Data HC
MOVES NOx —•—MOVES CO —•— MOVES HC

u

1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0

Figure 3-23 HD Gasoline Start Emission Ratio Compared to Recent Data

3.2.4 Start Energy Rates

The heavy-duty gasoline start energy rates were originally derived in MOVES2004, and updated in
MOVES2010a as described in the corresponding reports.57 Figure 2-32 displays the C02 (g/mile) emission
rates for cold start (operating mode 108) calculated from the energy rates using the carbon content of
conventional gasoline as documented in the MOVES3 Greenhouse Gas and Energy Report. As shown, there
is substantial variability in the start rates between 1974 and 2000. As discussed in Section 2.1.4.1, the
detailed methodology used in MOVES2004 (which modeled different emission rates according to vehicle
weights, engine technologies, and engine sizes) introduced variability into the energy rate within the
current MOVES regulatory class emission rates.

Table 3-26 displays the relative contribution of running and start operation to total energy consumption
from the heavy-duty gasoline regulatory classes from a national run for calendar year 2016. Like diesel
vehicles, starts from gasoline vehicles are estimated to be a relatively small contributor to the total energy
demand of vehicle operation. Due to the small contribution to the total energy inventory, we have not
prioritized updating the heavy-duty gasoline start emissions rates.

199


-------
Table 3-26 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 2-53. The energy rates also increase with cold temperatures using the
temperature effects documented in the 2004 Energy Report.71

The start energy rates include the projected impact of the Phase 1 Heavy-Duty GHG standards, which
began phasing-in in 2014 and have the same reductions as the running energy rates, as presented in Table
2-30 and Table 2-34. 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'

w

2

o
o

CM

O
O

400'

t

to

5 300'

ra
0£

200'

100'



iJ





X

w





fx













\









Reg Class

41-LHD2b3

42-LHD45

46-MHD67

47-HHD8

1970

1990

2010
Model Year

2030

2050

Figure 3-24 Heavy-Duty Gasoline Cold Start C02 Emission Rates (g/start) by Model Year and Regulatory Class

(OpmodelD 108)

4 Heavy-Duty Compressed Natural Gas Exhaust Emissions

While natural gas lacks the ubiquitous fueling infrastructure of gasoline, the use of compressed natural gas
(CNG) as a transportation fuel has grown for public transit, government, and corporate fleets. Such fleets
typically utilize centralized, privately-owned refueling stations. Fleet vehicles are typically operated as
back-to-base, which means the vehicles return to the same base location each day for refueling. Within

200


-------
this segment, some of the most prevalent use of in CNG vehicles has occurred among city transit bus fleets
and in solid waste collection or refuse truck fleets.118

MOVES models CNG fuel for all heavy-duty source types. CNG transit buses are mapped to the urban bus
regulatory class (regClassID 48). CNG vehicles in other heavy-duty source types are mapped to the HHD
regulatory class (regClassID 47). 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.

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 CH4 rate. The methane fraction from CNG vehicles is 89% and 96%
for model year groups 1950-2001 and 2002-2060, respectively, as documented in the Speciation report.6

Total energy consumption is age independent and therefore stored in the EmissionRate table. Some of the
published studies did not report total energy consumption directly, so it was necessary to compute energy
from a stoichiometric equation based on the carbon content in the emitted pollutants or from reported
values of miles per gallon equivalent of diesel fuel. In the former case, we used 0.8037 as the carbon
fraction coefficient for non-methane hydrocarbons (NMHC) when the bus was equipped with an oxidation
catalyst and 0.835 without due to high ethene levels, using speciation profiles from Ayala et al. (2003)119
discussed later in this section. All other conversion factors to energy were taken from Melendez et al.
(2005).123

On a similar note, MOVES does not report particulate matter (PM2.5) as a single rate; it reports one rate for
PM from elemental carbon (EC) of 2.5 microns or less, and another rate for non-elemental carbon of 2.5
microns or less. These separate rates for PM (EC) and PM (NonEC) from the emissionRateByAge table are
added together for a total PM2 5 rate used for comparison to the measurements.

4.1 Running Exhaust Emission Rates

The pre-2010 running emission rates are relatively unchanged from MOVES201443, and are based on cycle
average rates as discussed in Section 4.1.1. The running exhaust emission rates for model year 2010 and
later CNG vehicles using second-by-second in-use emission measurements from heavy-duty vehicles are
discussed in Section 4.1.2.

4.1.1 1950-2009 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.

43 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.

201


-------
4.1.1.1 CNG Chassis Dynamometer Measurements

Chassis data was collected from programs that were conducted at several research locations around the
country on heavy-duty chassis dynamometer equipment. In our analysis for MOVES2014, we compiled 34
unique dynamometer measurements. Data from newer studies such as Clark et al. (2007)130 would provide
further validation and refinement to the rates discussed in this report, however they have not been
incorporated here.

The data considered consisted of distance-specific running emissions rates for each of the following
pollutants and total energy:

1.	oxides of nitrogen (NOx)

2.	carbon monoxide (CO)

3.	particulate matter (EC + non-EC)

4.	total hydrocarbons (THC)

5.	methane (CH4)

6.	total energy consumption

This data was collected on two driving cycles, the Central Business District (CBD) and Washington
Metropolitan Area Transit Authority (WMATA).

The CBD cycle is defined as a driving pattern with constant acceleration from rest to 20 mph, a short cruise
period at 20 mph, and constant deceleration back to rest, repeated for 600 seconds (see Figure 4-1).120 The
WMATA cycle was developed using GPS data from city buses in Washington, DC, and has higher speeds
and greater periods of acceleration than the CBD cycle (see Figure 4-2).

CBD





















I



































































































0	100	200	300	400	500	600

Time, s

Figure 4-1 Driving Schedule Trace of the Central Business District (CBD) Cycle121

202


-------
0	500	1000	1500	2000

Time (sec)

Figure 4-2 Driving Schedule Trace of the Washington Metropolitan Area Transit Authority (WMATA) Cycle123

Table 4-1 shows a summary of the number of unique CNG bus measurements by driving cycle for each
study. Navistar published a similar study of CNG and diesel buses in 2008, and this analysis shares many of
the same sources.122 All of the vehicles were in service with a transit agency at the time of testing. The
number of unique measurements are typically equal to the number of vehicles tested and the
measurements were typically reported as averages based on multiple runs with the same vehicle and
configuration over a specific driving cycle with the exception of measurements reported by Ayala et al.
(2002)125 and Ayala et al. (2003).119 In the Ayala et al. (2002) study the 2000 model year CNG bus was
tested and then retested after approximately two months of service, which we treated as independent
measurements. Ayala et al. (2003) retested the same 2000 CNG bus as in their previous study; however,
the bus had accumulated an additional 35,000 miles and was serviced by the OEM to be equipped with an
oxidation catalyst that was later removed for baseline testing. Ayala et al. (2003) conducted duplicate tests
under each vehicle/aftertreatment configuration, which we considered four independent measurements.

203


-------
Table 4-1. Summary of External Emissions Testing Programs by Driving Cycle and Number of Unique Measurements

and their Corresponding Model Years

Paper/Article

Lead Research Unit

Driving
Cycle(s)

Model Year
(Number of
Measurements)

Melendez
2005123

National Renewable Energy Laboratory
(NREL)

WMATA

2001 (4),
2004 (3)

Ayala 2003119

California Air Resources Board (CARB)

CBD

2000	(4),

2001	(2)

LeTavec 2002124

Atlantic Richfield Company (ARCO)

CBD

2001 (1)

Ayala 2002125

CARB

CBD

2000 (2)

Lanni 2003126

New York Department of Environmental
Conservation

CBD

1999 (3)

McKain 2000127

West Virginia University (WVU)

CBD

1999 (3)

Clark 1997128

WVU

CBD

1996 (10)

McCormick
1999129

Colorado School of Mines

CBD

1994 (2)

TOTAL



(34)

4.1.1.2 Determining Model Year Groups

Model year groups are intended capture differences in vehicles over time while still being manageable
from a computational viewpoint. Model year groups are defined based on availability of measurement
data (see Table 4-1), emissions standards, and/or new vehicle technologies that affect real-world
emissions.

4.1.1.2.1 1994-2001 Model Years

We evaluated the measured NOx, CO, PM2.5, and THC emission rates to establish model year groups and
chose to group all the CBD measurements from the literature into one model year group, spanning from
MY 1994 to MY 2001. Note that we decided to exclude one of the studies that had four MY 2001 buses
tested on the WMATA cycle from this part of the analysis. This was done because inclusion increased the
complexity of analysis by having to deal with two driving cycles within a model year group while providing
only an incremental increase in sample size.

204


-------
4.1.1.2.2	2002-2006 Model Years

Of the surveyed data, only one study had vehicles newer than MY 2001.44 130This paper, a joint study
between NREL and WMATA, had three MY 2004 vehicles. The MY 2004 vehicles have a visibly different
emissions profile than the other vehicles. While these buses were only tested on the WMATA cycle, they
were all equipped with oxidation catalysts and had substantially lower emissions, particularly for PM2.5,
compared to the 1994-2001 buses tested on the CBD cycle. As a result, we created a model year group
from MY 2002 to MY 2006 based on the MY 2004 buses tested on the WMATA cycle. This MY group ends
before MY 2007 when a new series of stringent emission standards went into effect, as described below.131

4.1.1.2.3	2007-2009 Model Years

MOVES2014 had a single set of emissions for 2007-and later buses. In MOVES3, we created two groups,
MY 2007-2009 and MY 2010+ (noted as MY 2010-2017 when comparing certification data). We decided to
split the groups in this way because: (a) changes to 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.132 Analysis of these data showed that from MY
2002 to MY 2017, there have been changes in average certification levels for all the pollutants considered
in this report. In particular, NOx and PM2.5 levels have dropped dramatically. This effect is largely
attributable to increasingly stringent emission standards, which have affected both diesel and CNG
engines.

Emission rates from analysis of certification data and number of CNG engine families in the certification
data are shown in Table 4-2 below. The current, and historically most stringent, heavy-duty compression-
ignition NOx standard of 0.20 g/bhp-hr was fully phased in by 2010 and MY 2010+ heavy-duty CNG engines
are required to meet this standard (even if they are not compression-ignition). Thus, the average NOx
certification value for the MY 2010-2017 group is considerably lower compared to the MY 2007-2009
group. At the same time, and mostly to meet the new NOx standard, heavy-duty CNG engines transitioned
from lean-burn to stoichiometric-combustion with TWC. This technology transition is the likely reason for
the increase in THC and CO certification emissions rates from MY 2007-2009 to MY 2010-2017.

The differences in MOVES emission rates across all model years are discussed below in Section 4.1.3.

44 Several papers have discussed more recent vehicles. Examples include Clark et al. (2007).129 Data from these newer
studies would provide further validation and refinement to the rates discussed in this report.

205


-------
Table 4-2 Model Year Group Based Certification Emission Rate for Heavy-Duty CNG Engine Families

Model Year
Group

Number of Engine
Families3,15

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

2010-2017®

159 for NOx and CO, 153
for THC, and 120 for

PM2.5

0.1051

4.413

0.0028

0.044

a For MY 2002-2006, the number of engine families is based on HD CNG urban bus
regulatory class. For MY 2007-2009 and MY 2010-2017, the number of engine families is
based on all HD CNG engine families.

b Some engine families did not report emission data for THC and/or PM2.5.
c MY 2002-2006 group emission rates are projected sales weighted average of HD CNG
urban bus certification emission rates. MY 2007-2009 and 2010-2017 group emission
rates are simple average of all HD CNG certification emission rates (no weighting for
projected sales).

d Certification data has measurements of organic material non-methane hydrocarbon
equivalent (OMNMHCE). For this analysis they were treated as NMHC values.133
e Only shown for comparison. Certification data for MY 2010-2017 is not used in
developing MY 2010+ rates, which are based on MY 2010+ CNG vehicles in the HDIUT
data set.

4.1.1.3 Creating Comparable MOVES Gasoline Emissions

Section 1.6 explains how MOVES operating modes relate to scaled tractive power (STP). Because we
lacked data on age and STP trends for pre-2010 CNG vehicles, we applied adjustments based on the rates
for gasoline MHD vehicles. To do this, we compared CNG emission data collected on the CBD and WMATA
cycles to what MOVES estimated for MHD gasoline vehicles on those same cycles.

Because the pre-2009 CNG vehicles form a small and diminishing portion of the MOVES 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 4-3 and
Figure 4-4. 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.

206


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

350
300

—. 250
jj. 200

<3J

| 150
P 100
50
0

I

I I

I I I I I I I I I I I I I I I I I I I I I I I

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 4-3 Operating Mode Distribution for the CBD Cycle

900
800
700
~ 600
,8, 500
gj 400
1= 300
200
100

t

ffi

I I I I I I I I I I I I I I I I I I I I I I I

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

Operating Mode

Figure 4-4 Operating Mode Distribution for the WMATA Cycle

4.1.1.3.2 Simulating Cycle Average 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:

• fuel type,

207


-------
driving cycle,
age group,
regulatory class,
model year, and
pollutant and process.

E ¦ , = ?>0"RP0"*T0M,cycle	Equation 4-1

a-,p,simcycle	n	^

U cycle

Where:

•	Dcyde = 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 1950-2006 Model Years

The operating mode based emissions rates for MOVES2010b45 MHD gasoline vehicles were adjusted by the
ratio of cycle-average emissions rates from chassis dynamometer measurements to simulated cycle
modeling (see Section 4.1.3). For MY 1994-2001 and MY 2002-2006, the adjustment ratios were based on
the CBD cycle and WMATA cycle, respectively.

For each model year group, a central model year was selected as the source for the MHD gasoline
operating mode based rates. For MY group 1994-2001, we used MHD gasoline rates from MY 1997
because it is one of the median years in the group. Alternatively, we could have used the other median
year, MY 1998. Even though the average rate for MY 1998 was significantly lower (44 percent of that of
MY 1997), based on Equation 4-2, we expect minimal differences in the final estimated CNG rates (Rcng
term) whether we use MY 1997 or MY 1998 as the median year since the lower operating mode rates (RMdg
term) will lead to lower simulated cycle-average rate (Emdg term), which in turn will lead to larger
adjustment ratio (ECng/Emdg). For MY group 2002-2006, 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

45 The PM exhaust emission rates were subsequently updated from MOVES2010b for MHD gasoline vehicles for pre-
2010 model years.

208


-------
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	, Ecng,cbd,i994-2ooi	Equation 4-2

KCNG,OM,1994-2001 — KMDG,OM,1997 * p

^MDG,simCBD,1997

n	n	. ECNGiWMAXAi2oo4	Equation 4-3

CNG,OM,2002—2006 — KMDG,OM,2004 * p

tMDG,simWMATA,2004

Where:

•	Rcng,om,myg = operating mode based emissions rate for CNG vehicles for model year group
(MYG) 1994-2001 or MY 2002-2006, in g/hr

•	Rmdg,om,my = operating mode based emissions rate for MHD gasoline vehicles for model
year 1997 or 2004 (corresponding to MYG), in g/hr

•	Ecng,cycle,myg = Chassis dynamometer cycle-average emissions rate for MY 1994-2001 or
2004 CNG buses tested on a CBD or WMATA cycle, respectively, in g/mile. See Table 4-3.

•	EMDG,simcycie,MY = Simulated cycle-average emissions rate for MY 1997 or 2004 MHD gasoline
vehicles for CBD or WMATA cycle, respectively, in g/mile. This cycle-average rate is
calculated using the Rmdg,om,my operating mode rates. See Table 4-3.

4.1,1,4.2 2007-2009 Model Years

Due to lack of published data on MY 2007-2009 in-use vehicles, we used certification emissions rates,
shown in Table 4-2, to scale the operating mode based emissions rates. Certification emissions rates are
reported in grams per brake horsepower-hour (g/bhp-hr) and are not directly used in formulating MOVES
emission rates because they do not include real-world effects such as deterioration134 which were present
in the chassis dynamometer measurements used to estimate emissions rates for MY 1994-2001 and MY
2002-2006. So, we created scaling factors that we could apply to the MY 2002-2006 emissions rates to
estimate rates for MY 2007-2009. This scaling factor is the right-most term in Equation 4-4 shown below.

n	n	, ECNGiWMAXAj20o4 , Ccng,2007—2009	Equation 4-4

KCNG,OM,2007-2009 — KMDG,OM,2004 * p	* r

tMDG,simWMATA,2004 ^CNG,2002-2006

Where:

•	Ccng,2007-2009 = Average certification emission rate of all heavy-duty CNG engine families of
model year MY 2007-2009 in g/bhp-hr

•	Ccng,2002-2006 = Projected sales weighted average certification emission rate for CNG urban
bus engine families in MY 2002-2006, in g/bhp-hr

The adjustment ratio for energy consumption for MY 2002-2006 (Equation 4-3) is applied to all model
years in 2007-2009. For MY 2007+, we did not scale the energy consumption rates like we did for other
pollutants (Equation 4-4) because even though we have certification data on C02 emission rates for 2007-
2009 model years, we do not have certification data on C02 emission rates for MY 2002-2006. As a result,
MY 2007-2009 energy consumption rates are identical to the MY 2002-2006 rates.

209


-------
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

u

LU

1

CL

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

u

LU

1

CL

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

u

LU

1

CL

THC

TOTAL ENERGY

1994-2001c

all

1997

2.16

0.160

15.5

21.6

7.17

1.37

2002-2006c

all

2004

1.67

0.115

1.09

1.87

7.79

1.15

2007-2009d

all

2004

0.842

0.157

0.587

1.01

3.34

1.15

2010+®

Age 0- 3 (and 4-5) rates are based on analysis of 1 hz data from MY 2010+ CNG vehicles in
the HDIUT data set. Ages 6+ apply deterioration factors to age 0-3 rates as described in
main text.

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

c The ratios are calculated using Equation 4.2 or Equation 4.3 and the ECNG and EMDG values in this
table

d The 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.

fThe 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.

210


-------
4.1.2 2010-2060 Model Years

Running emission rates for MY 2010 and later CNG vehicles were based on information from in-use trucks,
and thus, unlike the calculations for earlier model years, the rates by operating mode could be calculated
directly. We then applied factors to account for deterioration with vehicle age, and adjustments to energy
rates to account for the phase-in of heavy-duty greenhouse gas standards.

We did not update CNG emission rates for the HD2027 rule since the average NOx FTP emission level for
MY 2010-2017 CNG engine families is already close to the HD2027 0.1 g/hp-hr standard. We expect any
further reductions due to the rule to be small.

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 fscaie 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 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 MOVES, ages 0-3 and 4-5 have no deterioration and the MOVES2014 light-duty PM2.5
deterioration factor for age 6-7 is applied to all CNG PM2.5 emission rates for ages 6+, thus making the PM2.5
and gaseous pollutant methods more (but not fully) aligned. Note that, unlike the factor for gaseous
pollutants, the MOVES PM2.5 deterioration factor does not vary between operating modes for a given age

group. See Section 3.1.2.1.3 for more details and Table 4-4 for a comparison between MOVES3 and later and MOVES2014.

211


-------
Table 4-4 Age-based Deterioration Factor for PM2.s Emission Rates for HHD and Urban Bus CNG Vehicles in Model Year

2010+

Age

MOVES2014

MOVES3 and later1

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

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, Phase 2, and Phase 3 Rules
To model energy consumption in MOVES, we split the CNG MY 2010+ group into MY 2010-2013, MY 2014-
2017, and MY 2018+ groups. The MY 2010-2013 energy consumption rates are identical across these
model years and based solely on the HDIUT data set analysis. For MY 2014-2017, the CNG energy
consumption rates of MY 2013 are reduced by the percentage reduction assigned to HHD vehicles in the
Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium and Heavy-Duty Engines
and Vehicles Phase 1 rule59 (see Table 2-30). Similarly, for MY 2018 and later, using MY 2017 rates as base
year, the energy consumption rates of CNG vehicles are further reduced as per the Greenhouse Gas
Emissions and Fuel Efficiency Standards for Medium and Heavy-Duty Engines and Vehicles — Phase 2 rule61
(see Table 2-31). Note that the Phase 1 reduction for CNG vehicles is identical across all allowed source
type and regulatory class combinations. However, for the Phase 2, different reductions for CNG vehicles
are applied by source type and regulatory class (see Table 2-31). The anticipated improvements in fuel
efficiency from the Phase 2 rules are stored in the EmissionRateAdjustment table.

The GHG Phase 1, Phase 2, and Phase 3 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
C02 emissions, and the reductions in fuel consumption (and C02 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 4-5 through Figure 4-9 display the THC, CO, NOx, PM2.5, and C02 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 analysis methodology and a shift in the CNG
vehicle fleet from lean-burn combustion to stoichiometric combustion with a three-way catalyst.

212


-------
Figure 4-5 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 catalyst (TWC) is not supported by the certification data presented in
Table 4-2 nor in recent studies comparing stochiometric and lean-burn combustion CNG engines.135,136 The
differences in the methodologies and limitations in the pre-2010 data likely contributed to apparent
increase in THC emissions starting in 2010 model year in MOVES.

a;

1

2	6'
at

re
ct

0

1





V







b











u





Reg Class

47-HHD8

48-Urban	Bus

1980	2000	2020	2040

Model Year

Figure 4-5. Base running emission rates for THC from age 0-3 CNG heavy-duty vehicles averaged over a nationally

representative operating mode distribution

Like THC, CO also shows a significant increase in emission rates with the updated analysis of the MY 2010+
vehicles (See Figure 4-6). However, the increase in CO is supported by certification data (Table 4-2) and in
more recent testing comparing stoichiometric-combustion with TWC based CNG buses.135,136

213


-------














V		,,,,,











































Reg Class

47-HHD8

48-Urban	Bus

1980	2000	2020	2040

Model Year

Figure 4-6. Base running emission rates for CO from age 0-3 CNG heavy-duty vehicles averaged over a nationally

representative operating mode distribution

Figure 4-7 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.135,136

214


-------
15-

10-

Q)
£
3

<1>
+-•
ra

en

X

O

Reg Class

47-HHD8

48-Urban	Bus

1980	2000	2020	2040

Model Year

Figure 4-7. Base running emission rates for NOxfrom age 0-3 CNG heavy-duty vehicles averaged over a nationally

representative operating mode distribution

Figure 4-8 shows significant decreases in the PM2.s emission rates between the 2001 and 2002 model year
emission rates based on the chassis-cycle average PM2.5 emission rates. An increase in PM2.5 emission rates
is shown from model year 2009 and 2010, which is inconsistent with the certification data (Table 4-2). The
two studies evaluating stoichiometric and lean-burn showed mixed results, with one study showing
stoichiometric engines with TWC emitting lower PM2.5 emissions136 and the other study showing
stoichiometric engines emitting higher PM2.5 rates.135

215


-------




































































Reg Class

47-HHD8

48-Urban	Bus

1980	2000	2020	2040

Model Year

Figure 4-8. Base running emission rates for PM2.5from age 0-3 CNG heavy-duty vehicles averaged over a nationally

representative operating mode distribution

Figure 4-9 shows general decreases in the C02 emission rates across model years, including the impact of
the HD GHG Phase 1 and Phase 2 rules discussed in 4.1.2.3. We show some rebound in HD CNG C02 energy
consumption rates due to the adoption of HD ZEV vehicles during the HD GHG Phase 3 rule and the fleet
averaging adjustment in MOVES.

216


-------
1500'

a;

E

TO 1000'

-------
NH3 emissions from starts. Extended idle emission rates are only estimated for gasoline and CNG as
discussed in the following sections.

5.1 Heavy-Duty Diesel

Ammonia is not formed during typical combustion processes in diesel engines but is emitted as a
undesirable byproduct of selective catalytic reduction (SCR) aftertreatment systems present on model year
2010 and later heavy-duty diesel vehicles to meet the 2010 NOx emission standards.

The SCR system functions by injecting aqueous urea into the exhaust flow, which then thermally
decomposes to NH3 and C02 (Equation 5-1). NOx is then reduced to N2 by reactions shown in Equation 5-2
through Equation 5-4.137

Excessive urea injected into the exhaust stream can lead to ammonia passing through the aftertreatment
system into the atmosphere, referred to as ammonia slip. NH3 oxidation (AMOX) catalysts can be used to
reduce ammonia slip but may not eliminate the emissions entirely.138,93

We estimate zero heavy-duty diesel NH3 start emissions. We also estimate zero NH3 extended idle
emissions. NH3 emissions during extended idling and startsshould be minimal, due to little or no dosing of
urea in the SCR system because current technology diesels cannot maintain the required operational
exhaust temperature at extended low loads.50 We do not estimate NH3 emissions from auxiliary power unit
exhaust, which are not expected to have SCR systems. The remainder of this section discusses ammonia
running emissions.

Ammonia measurements were not part of the HDIUT program we used to estimate the THC, CO, NOx, and
PM2.5 emission rates from heavy-duty diesel and CNG vehicles, as described in Section 2. As such, we relied
on other data sources for developing heavy-duty NH3 emission rates for diesel vehicles. We compared
fleet-average heavy-duty vehicle NH3 emission rates reported in the literature to in-use studies as
summarized in Table 6-1.

C0(NH2)2 + H20 -> 2NH3 + C02
4NH3 + 4N0 + 02 -> 4N2 + 6H20

Equation 5-1
Equation 5-2
Equation 5-3
Equation 5-4

4NH3 + 6N0 -> 5N2 + 6H20
2NH3 + NO + N02 -> 2N2 + 3H20

218


-------
Table 5-1. Fieet- average fuel- based NH3 emission rates (± 95% Confidence Intervals) from heavy-duty vehicles reported

from recent studies

Study

Study

Location

Number of

Heavy-duty vehicle fleet average



Year



vehicles

NHs emission rate (g/kg-fuel)

Preble et al.

2018

Caldecott Tunnel near

1,186

0.10 + 0.03

(2019)139



Oakland, CA





Haugen et al.

2017

Peralta Weigh Station near

1,844 (HDV)

0.09 + 0.02 (HDV)

(2018)142



Anaheim, CA

471 (MDV)
1,408 (high)
907 (low)

0.06 + 0.05 (MDV
0.08 + 0.02 (high)
0.06 + 0.05 (low)

Bishop et al.

2020

Perry Weight Station Salt

1,591 (HDV)

0.08 + 0.06 (HDV)

(2022)143



Lake City, UT

103 (MDV)
1,053 (high HDV)
538 (low HDV)

0.22 + 0.23 (MDV)
0.009 + 0.009 (high HDV)
0.23 + 0.02 (low HDV)

Wang et al.

2015

Fort McHenry Tunnel,

NA

0.10 + 0.07 (winter)

(2019)144



Baltimore, Maryland



0.03 + 0.08 (summer)

Preble et al. (2019)139 measured NH3 emissions rates from heavy-duty vehicles at the Caldecott Tunnel
near Oakland, California in 2018. They sampled the concentrations of NH3 and C02from the exhaust
plumes of individual heavy-duty vehicles as they entered the tunnel at a 4% grade traveling between 30
and 75 mph. From the NH3 and C02 concentrations, they estimated NH3fuel-based emission rates using the
carbon content of diesel fuel. By matching license plate images to state truck registration databases, they
were able to obtain vehicle information, including engine model year and aftertreatment system. The
average emission rates by different mode year and aftertreatment groups are shown in Table 5-2.

Preble et al. (2019) measured NH3 emissions from 2010 and later trucks equipped with SCR systems. They
were able to collect over 900 diesel truck NH3 emissions measurements identified by engine model year
and aftertreatment system. Collecting a large sample is important for capturing the fleet-average emision
rates, because 10% of trucks contributed 95% of the total fleet NH3 emissions. The Preble et al. (2019)
study measured a large number of model year 1994-2006 retrofit DPF trucks due to the large number of
drayage trucks servicing the nearby Port of Oakland. Between 2010 and 2012, all California drayage trucks
were required to be equipped with diesel particulate filters.140 The average ammonia emission rates for
the pre-2010 model year groups are low and uncertain, which is consistent with measurements of heavy-
duty ammonia emissions made from the Caldecott Tunnel in 2006.141 The fleet-average heavy-duty diesel
emission rates for Preble et al. (2019) are shown in Table 5-1. The sample size is larger than that listed in
Table 5-2 because the fleet-average includes all heavy-duty diesel vehicles measured, including trucks that
were not matched to the vehicle registration database.

219


-------
Table 5-2. Fuel-based NH3 emission rates (± 95% Confidence Intervals) from heavy-duty vehicles by aftertreatment and
engine model year measured at the Caldecott Tunnel by Preble et at (2019)139

Aftertreatment

Engine Model Year

NH3 (g/kg) fuel-based
emission rate

Number of
vehicles

Model year
ranges used
in MOVES

DPF + SCR

2010-2018

0.18 + 0.07

547

2010-2060

DPF

2007-2009

0.00 + 0.01

181

2007-2009

Retrofit DPF

1994-2006

0.01 + 0.01

114

Not used

No DPF

2004-2006

0.00 + 0.01

24

2004-2006

No DPF

1965-2003

0.02 + 0.02

62

1960-2003

The University of Denver's research group has conducted two studies of heavy-duty diesel emissions
measured using their remote sensing device called the Fuel Efficiency Automobile Test (FEAT) system that
measures pollution concentrations across the roadway. Haugen et al. (2018)142 measured NH3 emissions
from heavy-duty vehicles at the exit ramp of the Peralta Weigh Station near Anaheim, California. They
separately sampled emissions from heavy-duty vehicles with the elevated exhaust pipes ("high" in Table
6-1), and ground-level exhaust pipes ("low" in Table 6-1). Additionally, they classified samples into
medium-duty (defined as vehicles with GVWR < 26,000 lbs, or class 2 through 6 vehicles), and heavy-duty
vehicles (GVWR > 26,000 lbs, or Class 7 and 8 vehicles). The "low" sample has a high percentage of
medium-duty vehicles which had lower NH3 emission rates, and newer heavy-duty vehicles which have
higher than average NH3 emission rates. The heavy-duty and medium-duty vehicles measured in Haugen et
al. (2018), are over 99% and 92% diesel vehicles, respectively, with the remainder being compressed
natural gas vehicles. The research measured a large increase in fleet-average NH3 emissions in the 2017
campaign compared to previous measurements made at the Peralta Weight Station in 2008, 2009, 2010
and 2012, due the penetration of SCR-equipped vehicles into the in-use fleet. The average for the 2010 and
later chassis model years from this study was 0.14 g/kg/fuel, while the older model year vehicles had NH3
rates near zero.

Bishop et al. (2022)143 measured emissions from heavy-duty vehicles at the Perry, Utah Port of Entry (~50
miles north of Salt Lake City) in December 2020. Bishop et al. (2022) separately measured heavy-duty
vehicles with "high" and "low" exhaust pipe positions. In this study, the "low" exhaust tailpipe trucks were
almost exclusively 2011 and later heavy-duty vehicles trucks, which had significantly higher NH3 than the
older trucks included in "high" exhaust tailpipe group. The "low HDV" group had a mean emission rate
similar to that estimated from Preble et al. 2019 for DPF+SCR equipped trucks. Bishop et al. (2020)
measured a smaller number of Class 4, 5, and 6 vehicles, for which the mean estimate was highly
uncertain.

Wang et al. (2019)144 measured fleet average NH3 emissions for a week in both February and July/August,
2015 from the Fort McHenry Tunnel, which is along the 1-95 corridor with a traffic volume of ~55,000
vehicles per day. They measured emission concentrations from two of the four bores that contain heavy-
duty vehicle traffic. The tunnel includes a -1.8% down grade, followed by a 3.3% positive grade to the exit
of the tunnel. Using the measured concentration and the fraction of heavy-duty vehicles, they estimated

220


-------
fuel-based NH3 emission rates for both light-duty and heavy-duty vehicles. Because this was the earliest
study conducted, we would expect the NH3 emission rates to be lowest from this study due to a smaller
fraction of SCR-diesel vehicles present in the fleet. The summer measurements are lower than the other
studies, however, the winter measurements are similar. In addition, the difference is not statistically
significant due to the large confidence intervals of the mean.

Despite the different measuring systems, locations, and sampling years, the fleet-average emission rates
are statistically similar among the different studies. This provides confidence that the fuel-based emission
rates reported from the studies are not strongly impacted by measurement methods or the sampling
conditions of the location.

We developed heavy-duty NH3 emission rates in MOVES using the reported fuel-based emission factors by
model year and aftertreatment class from Preble et al. (2019) reported in Table 5-2. We chose to use
Preble et al. (2019) because they reported the emission rates exclusively for heavy-duty vehicles by model
year ranges. In addition, we also used the Preble et al. (2019) study to update the MOVES N0/N02
fractions (Section 7.1) and N20 emission rates.

To develop MOVES heavy-duty diesel emission rates by regulatory class, model year, and operating mode,
we multiplied the MOVES3 heavy-duty diesel vehicle fuel-consumption rates by regulatory class, model
year, operating mode (Fuel Rates RegMY,op) by the Preble et al. (2019) fuel-based NH3 emission rates
(FERModeiYear Group) from Table 5-1 shown below in Equation 5-5.

CD	— Una] Dntoc	V /7/7R	rAjllUllOn J-J

L,riReg,MY,age,op r UCL nuLC^Reg,MY,op * r L,[XModelYear Group

Figure 5-1 shows example NH3 emission rates for the LHD2b3 and HHD regulatory class for model year
2017. Even though the fuel-based emission rate is the same, the gram per hour rate is larger for the HHD
regulatory class due to higher fuel consumption rates.

We replicated the NH3 emission rates for each heavy-duty regulatory class and model year across all
vehicle ages. This differs from the ammonia rates for light-duty gasoline NH3 where we had a much larger
data sample and were able to estimate age effects. Preble et al. (2019) collected measurements of NH3 in
only their most recent campaign. Haugen et al. (2018) collected several measurement campaigns at the
Peralta Weight station, but had a limited number of vehicles from which to estimate both model year and
age specific emission rates. We recommend future studies to evaluate the impact of aging, deterioration
and mal-maintenance on NH3from heavy-duty diesel vehicles.

221


-------
12-

I 9"

JZ

d)

d)

ro

i_

c 6

0

If)


-------
5.2.2 1981-2060 Model Years

The heavy-duty gasoline vehicle emission rates are based on the new fuel-based emission rates for light-
duty trucks. We assume that the fuel-based ammonia emission rates are similar between light-duty and
heavy-duty gasoline vehicles because the same ammonia formation pathway is present for both vehicles
types. In modern gasoline vehicles, ammonia is formed from the catalytic reduction of NO in the three-way
catalytic converter during fuel-rich conditions.

Limited data is available to evaluate this assumption. Livingston et al. (2009) 147 measured ammonia
emissions from vehicles recruited in southern California. On average, they measured higher gram per mile
ammonia rates from the six medium-duty gasoline vehicles than the 35 light-duty vehicles. They attributed
the higher ammonia emission rates from the medium-duty vehicles due to:

1.	Larger exhaust volumes produced by the medium-duty vehicles

2.	Less stringent emission standards, which would lead to higher precursor exhaust emissions
of NOx and CO

3.	Potentially different catalyst activity

By estimating heavy-duty gasoline emission rates from light-duty gasoline vehicle fuel-based emission
rates, and heavy-duty gasoline fuel rates, we account for the larger exhaust volumes produced by heavy-
duty gasoline vehicles. While we do not account for items 2) and 3), which could lead to higher fuel-based
emission rates from heavy-duty gasoline vehicles, we believe our approach is reasonable given that heavy-
duty gasoline vehicles are anticipated to be a minor contributor of ammonia emissions in comparison to
light-duty gasoline vehicles.

The light-duty truck fuel-based ammonia emission rates are documented in the light-duty exhaust emission
rate report.12 The fuel-based emission rates are estimated for light-duty vehicles and light-duty trucks by
model year and vehicle age.

We estimated the ammonia emission rate for heavy-duty gasoline vehicles by multiplying the light-duty
truck fuel-specific emission rates by regulatory class, model year group, and age (LDT FERMYiage) by the
MOVES3 heavy-duty gasoline vehicle fuel-consumption rates by regulatory class, model year, operating
mode (HDG Fuel RatesRegMY,op) as shown in Equation 5-6.

HDG ERReg MY,age,op = HDG Fuel RatesReg MY,0p LDT FERMY,age	^

Example NH3 emission rates in MOVES for LHD2b3 and LHD45 gasoline vehicles for model year 2017 and
ages 0-3 are shown in Figure 5-2

223


-------
12.5-

^ 10.0-

u

o

JZ

d)

d) 7.5-
•%—*

ro

i_

c

0

Irt f- n


-------
blends of CNG fuel. In these cases, the gram per mile rates cited in Table 5-3 are those for L-CNG fuel (a
CNG blend produced from a liquified natural gas fuel tank), which most closely matches MOVES default
CNG fuel properties.

Karavalakis et al. (2013)152 tested a MY 2002 refuse hauler equipped with a lean-burn engine with
oxidation catalyst. This testing included many blends of CNG fuel over a single cycle (RTC cycle) that was
developed to be representative of typical refuse hauler operation, including engine load related to
compacting customer waste. Karavalakis et al. (2016a)154 performed similar testing with a MY 2011
stoichiometric waste hauler.

Karavalakis et al. (2016b)155 continued this work by testing a CNG school bus (MY 2005 lean-burn engine
with oxidation catalyst) and two Class 8 truck (MY 2012 and 2013 stoichiometric engine with EGR and TWC
systems). These vehicles were tested on a chassis dynamometer using cycles representative of typical
school bus operation or near-dock and local operation. Once again, several blends of CNG fuels were
tested.

Thiruvengadam et al. (2016)148 performed chassis dynamometer testing on five CNG vehicles: three Class 8
tractors, a refuse hauler and a transit bus. Table 5-3 presents an average of the results obtained for the
Class 8 tractors and results for the transit bus and refuse-hauler. The tests were done using some
regulatory cycles and other drive cycles representative of typical HD vehicle operation.

Zhu et al. (2024)150 carried out a study measuring tailpipe ammonia emissions from in-use goods
movement vehicles, including two Class 8 CNG trucks equipped with three-way catalysts. The study
measured vehicles following different routes representing goods movements in the South Coast Air Basin
using a mobile laboratory. Table 5-3 presents the range of g/mile emissions measured from both CNG
trucks.

Finally, the South Coast Air Quality Management District (SCAQMD) requested a study led by University of
California Riverside and West Virginia University156 focused on measuring in-use emissions for a cohort of
200 heavy-duty vehicles with GVWR > 14,000 lbs. It included vehicles of different vocations (transit buses,
school buses, refuse trucks, delivery vehicles and goods movement vehicles) and measured their emissions
with laboratory tests and real-world onroad tests. EPA obtained data from this work through a
collaborative agreement; however, at the time of developing this update, the second-by-second NH3 data
was not available. At the recommendation of peer reviewers, we included average values across tests and
vehicles of different vocations measured under the UDDS cycle from this study; these values were derived
from vehicle-specific g/mile values provided in Appendix K of that report.

225


-------
Table 5-3 Literature used to inform heavy-duty CNG emission rates in MOVES

Source

Fuel

Vehicle

Engine
MY

Engine and
Aftertreatment

Measurement
type

Drive/Duty
Cycle

nh3

Rate

(mg/mil

e)

Karavalakis et al
(2013)

CNGA

Waste hauler

2002

Lean-burn with
oxidation catalyst

Dynamometer
(RTC cycle)

Transport

65

Curbside

117

Compaction

6

Hajbabei et al
(2013)

CNG

Transit bus

2003,
2004

Lean-burn with
oxidation catalyst

Dynamometer

Central
Business
District Cycle

100-250

2009

Stoichiometric with
TWC

1300-
1450

Karavalakis et
al. (2016b)

CNGA

School bus

2005

Lean-burn with
oxidation catalyst

Dynamometer

Central
Business
District Cycle

CO

o

m

Class 8
tractor

2012,
2013

Stoichiometric with
TWC + EGR

Dynamometer

Local Haul
Cycle

950

Karavalakis et
al. (2016a)

CNGA

Waste hauler

2011

Stoichiometric with
TWC + EGR

Dynamometer
(RTC cycle)

Transport

780

Curbside

500

Compaction

320

Thiruvengadam
et al. (2016)

CNG

Transit bus

2008

Stoichiometric with
TWC + EGR

Dynamometer

UDDS Cycle

890

Refuse hauler

2008

Stoichiometric with
TWC

Dynamometer

UDDS Cycle

1,710

Class 8
tractor

2008,

2009,
2011

Stoichiometric with
TWC + EGR

Dynamometer

UDDS Cycle

690

Regional Haul
Cycle

820

Near Dock
Cycle

1,347

Local Haul
Cycle

1,557

Zhu et al.
(2024)

CNG

Class 8
tractor

2019c

Stoichiometric with
TWC + EGR

On-road
(mobile lab)

Goods
Distribution

450-
690B

Port Drayage

610-640

Goods

Distribution +
Elevation

420-530

SCAQMD study

CNG

School Bus

2013

Stoichiometric with
TWC + EGR

Dynamometer

UDDS

4,185

Transit Bus

2009

5,329

Refuse hauler

2011,
2014

3,546

Delivery

2018

2,552

Goods
Movement

2011,
2017,
2018

6,372

A CNG blend produced from LNG as described in the literature referenced

B Bolded values are used as the MOVES emission rate for each engine technology. For lean-burn engines we use 30
mg/mile and for stoichiometric engines we use 690 mg/mile.
c Source only provided model year of test vehicles

226


-------
We selected the g/mile emission rates from Table 5-3 we think are most representative of their respective
engine technology, accounting for a variety of factors, as the fleet-representative rates. We chose vehicles
which tend to be representative of the driving patterns of many types of HD vehicles while also covering
the full engine power and speed range of HD engines. Therefore we did not use any rates from refuse
haulers, which have unique operational patterns. We also focused on vehicles well within the model year
range of significant adoption of their engine technology. For example, we chose not to use the rates from
Thiruvengadam et al. because we believe they are from earlier models of stoichiometric engines and seem
to have higher emission rates than more modern stoichiometric engines. For stoichiometric engines, we
selected the goods distribution route from Zhu et al. as the most representative route, consistent with
their own determination where they write that "it is the most representative route of goods movement
vehicles in SCAB (South Coast Air Basin) whose activities include urban, highway, and loading/unloading
events." While this emission rate might be more representative of congested areas than operations
occuring in interstate highways, we prioritized the activity that is representative of urban cores and covers
the full range of engine speed and power. Future work will consider refining emission factors (for example,
adding vocational rates or developing operating-mode specific rates) as more detailed data becomes
available.

In updating MOVES with the chosen g/mile rates, we set the emission rates for HD CNG vehicles in the age
group in which they were tested (age group 10-14 for lean-burn vehicles and age group 4-5 for
stoichiometric engines). There is not enough data available across a broad range of vehicle ages to update
the deterioration trends for CNG ammonia emission rates. The peer reviewers in our MOVES5 peer review
noted that some testing showed ammonia emissions may decrease as vehicle ages because ammonia
emissions are driven by catalyst reactions. Given the uncertainty, we elected to model no deterioration in
CNG ammonia rates, such that CNG vehicles emit ammonia at the same rate regardless of vehicle age.

Because we did not have either the 1 Hz PEMS or dynamometer testing data needed to estimate emission
rates by operating mode, we applied the already existing operating mode trends, originally developed from
heavy-duty gasoline engines. Once again, as more detailed data becomes available, we will consider
performing further analysis to calculate CNG ammonia rates by operating mode from real-world CNG
vehicles.

Finally, since we have no data on extended idle emission rates for ammonia from CNG vehicles, we applied
the running idle emission rates (operating mode 1) for extended idle emissions. For lean-burn engines, we
use an extended idle rate of 139 mg/hour, compared to 290 mg/hourfor stoichiometric engines.

5.4 Summary

The following figures display the age 0-3 running ammonia emission rates for heavy-duty vehicles by
regulatory class and fuel type.

227


-------
—I—

1980

2000

2020

2040

Model Year

Figure 5-3: Base running emission rates for NH3from age 0-3 gasoline heavy-duty vehicles averaged over a nationally

representative operating mode distribution

Reg Class

—	41-LHD2b3

—	42-LHD45
46-MHD67

—	47-HHD8

0.09

0.06

0.03

0.00



























































-v...



Reg Class

—	41-LHD2b3

—	42-LHD45

46-MHD67

47-HHD8

48-Urban	Bus

49-Gliders

1980	2000	2020	2040

Model Year

Figure 5-4: Base running emission rates for NH3from age 0-3 diesel heavy-duty vehicles averaged over a nationally

representative operating mode distribution

228


-------
0.6

a;

|

3 0.41

aj
+-¦

CT3

0.2'

0.0























































Reg Class

— 47-HHD8
48-Urban Bus

1980	2000	2020	2040

Model Year

Figure 5-5: Base running emission rates for NH3from age 0-3 CNG heavy-duty vehicles averaged over a nationally

representative operating mode distribution

6 Heavy-Duty Crankcase Exhaust Emissions

Crankcase exhaust emissions, also referred to as crankcase blowby, are combustion gases that pass the
piston rings into the crankcase and are subsequently vented to the atmosphere. Crankcase blowby
includes oil-enriched air from the turbocharger shaft, air compressors, and valve stems that enters the
crankcase. The crankcase blowby contains combustion generated pollutants, as well as oil droplets from
the engine components and engine crankcase.157

6.1 Modeling Crankcase Emissions in MOVES

MOVES calculates crankcase emissions using two code modules: a gaseous and a particulate matter
crankcase emission calculator. Within these calculators, crankcase emissions are calculated in relationship
to tailpipe exhaust emissions. In MOVES, the tailpipe exhaust processes are running exhaust, start exhaust,
and extended idle exhaust (processID 1, 2, 90). The corresponding crankcase emission processes are
crankcase running exhaust, crankcase start exhaust, and crankcase extended idle exhaust (processID 15,
16, 17).

The gaseous crankcase calculator chains calculation of the crankcase emission rates to the tailpipe exhaust
emission rates for gaseous pollutants, but it does not change the tailpipe exhaust emission rates. On the
other hand, the particulate matter calculator has the ability to divide the particulate matter exhaust
emission rates stored in the emissionratebyage and emissionrate table into components representing the
contributions from tailpipe exhaust and crankcase emissions. Thus, the particulate ratios may be used to
adjust the particulate matter tailpipe exhaust emission rates to account for the crankcase contribution, as
was done in early versions of MOVES. In MOVES4 and later, however, the particulate matter tailpipe
exhaust emission rates only include tailpipe emissions and the exhaust ratios are set equal to one as shown

229


-------
in the subsequent sections for all model years, fuel types, regulatory classes and source types. More details
on the particulate matter crankcase calculator are provided in the MOVES Speciation Report.6

The crankcase ratios for non-methane hydrocarbons (NMHC), NOx, and PM2.5 are used to estimate the
crankcase emissions for each subspecies that is chained to their respective primary pollutant as shown in
Table 6-1. The crankcase emission ratios for both gaseous and particulate matter pollutants are stored in
the crankcaseEmissionRatio table. The table stores the crankcase emission rates by pollutant, process,
model year, source type, regulatory class, and fuel type. Regulatory class was added as a primary field in
MOVES4, and the crankcase emission ratios were updated according to regulatory class. The table
structure and code retain the ability to model separate crankcase emission ratios by source type, but we
use the same crankcase emission ratios across different source types within the same fuel type, regulatory
class and model year.

The PM 10 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.6

Table 6-1 Speciation of Chained Crankcase Pollutants from the Primary Pollutants

Primary Pollutant

Chained Crankcase Pollutant

Non-methane
hyd rocarbons
(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), lndeno(l,2,3,c,d)pyrene gas, volatile organic carbon (VOC),
Phenanthrene gas, Naphthalene gas, Pyrene gas

NOx

NO, NOz, HONO

PM2.5

EC, S04, H20(aerosol), NonECNonS04PM

By using crankcase to tailpipe emission ratios to estimate crankcase emission rates, MOVES implicitly
assumes that any increase in emissions due to aging and deterioration also occurs for crankcase emissions.
The data sets used to derive the crankcase emission rates for 2007 and later heavy-duty diesel engines are
based on engines and vehicles with relatively low miles and no expected deterioration in the emission
control system. If crankcase emissions do not exhibit the same increase in deterioration as tailpipe
emissions, this method would lead to an overestimation of crankcase emissions in older vehicles.

6.1.1 Heavy-Duty Diesel Crankcase Emissions

Crankcase emissions from pre-2007 diesel engines were typically vented to the atmosphere using an open
unfiltered crankcase system, referred to as a 'road draft tube'.157 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.158 159 160

Starting in model year 2001, federal regulations require closed crankcase systems for chassis-certified
diesel vehicles.161 Federal regulations permit 2006-and-earlier engine-certified diesel vehicles equipped
with "turbochargers, pumps, blowers, or superchargers" to vent crankcase emissions to the atmosphere.162
Starting in model year 2007, federal regulations no longer permit crankcase emissions to be vented directly
to the atmosphere, unless they are included in the certification exhaust measurements.162 Many heavy-
duty diesel manufacturers have adopted open crankcase filtration systems in model year 2007 and later

230


-------
engines.157 These systems vent the exhaust gases to the atmosphere after the gases have passed a
coalescing filter which removes oil and a substantial fraction of the particles in the crankcase blowby.157 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.163

A summary of published estimates of diesel crankcase emissions as percentages of the total emissions
(tailpipe + crankcase) are provided in Table 6-2. For the pre-2007 diesel technologies, hydrocarbon and
particulate matter emissions have the largest contributions from crankcase emissions. There is a
substantial decrease in PM2.s 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.

Table 6-2 Literature Review on the Contribution of Crankcase Emissions to Diesel Exhaust (Tailpipe + Crankcase)

Study

Model
Year

Tailpipe Exhaust
Aftertreatment

# Engines
or Vehicles

THC

CO

NOx

PM

Hare and Baines,
1977166

1966,
1973

None

2

0.2%-
3.9%

0.01%-
0.4%

0.01%
-0.1%

0.9%-
2.8%

Zielinska et al.
2008158

Ireson et al. 2011159

2000,
2003

None

2







13.5% -
41.4%

Clark etal. 2006165
Clark etal. 2006164

2006

None

1

3.6%

1.3%

0.1%

5.9%

Khalek et al. 200946
("ACES Phase 1")

2007

DPF-equipped

4

84.6%

33.5%

0.007

%

44.4%

NVFEL Testing

2015,
2018

SCR-DPF

2

19.0%-
57.8%

14.2%-
76.7%

2.3%-
7.5%

Not

measure
d

Note: The crankcase ratios shown here are a fraction of the total tailpipe and crankcase exhaust. The
crankcase ratios used in MOVES are a ratio of crankcase to tailpipe exhaust.

As discussed in the following subsections, we developed crankcase emission ratios by regulatory class and
model years groups using the available studies in Table 6-2, and additional information on requirements
for closed crankcase systems. The data on crankcase emissions are limited. The gaseous crankcase
emission rates for heavy-duty diesel vehicles are based on three studies in Table 6-2, totaling only seven
vehicles. As such, the rates have considerable, but unquantified, uncertainties.

6.1.2 LHD2b3 Crankcase Emissions

After 2001, all chassis-certified vehicles, including diesel vehicles, are required to avoid venting crankcase
emissions into the atmosphere.161 All LHD2b3 vehicles in MOVES are chassis-certified vehicles because the
small number of engine-certified LHD2b3 vehicles are classified as LHD45 vehicles as discussed Section 1.4.

MOVES uses two model year groups for crankcase emissions from LHD2b3 diesel vehicles. Model year
1960 to 2000 vehicles use the open crankcase ratios estimated for engine-certified vehicles as detailed in

231


-------
Section 6.1.3.1. For model year 2001 to 1960 we estimate zero crankcase ratios because all chassis-
certified diesel vehicles are required to have closed crankcase systems.

6.1.3 LHD45, MHD and HHD Crankcase Emissions

Diesel vehicles within the LHD45, MHD, and HHD regulatory classes are composed of 100% engine-certified
vehicles, which are permitted to emit crankcase emissions after model year 2007 if they are accounted for
in the engine-certification results. The crankcase emission rates for the engine-certified vehicles are
estimated by regulatory class and model year groups, to capture differences in crankcase emission ratios
reported from different studies, account for differences in the crankcase control in 2007 and later model
years, and to account for changes in tailpipe exhaust emissions which impact the crankcase to tailpipe
emission ratios.

6.1,3,1 1950-2006 Model Years

Table 6-3 displays the crankcase/tailpipe emission ratios used for pre-2007 diesel exhaust. For THC, CO,
and NOx, we selected the values measured on the MY 2006 diesel engine reported by Clark et al. 2006.165
These values compare well with the previous HC, CO, NOx values reported much earlier by Hare and Baines
(1977),166 which represent much older diesel technology. The similarity of the crankcase emission ratios
across several decades of diesel engines suggests that for pre-2007 diesel engines, crankcase emissions can
be reasonably well represented as a fraction of the exhaust emissions. The THC crankcase ratios presented
in Table 6-3 are also used for methane, total organic gases (TOG), non-methane hydrocarbons (NMHC) and
all the pollutants chained to NMHC listed in Table 6-1.

For PM2.5 emissions, we use a crankcase/tailpipe ratio of 20 percent. The 20 percent ratio falls within the
range of observations from the literature on diesel PM emissions. Zielinska et al. 2008158 and Ireson et al.
2011159 reported crankcase contributions to total PM2.5 emissions as high as 40 percent. Jaaskelainen
(2012)157 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.167 The crankcase
emission ratios shown Table 6-3 are applied to running, start and extended idle exhaust to estimate the
corresponding crankcase exhaust emissions.

Table 6-3 LHD45, MHD, and HHD 1960-2006 Diesel Crankcase Ratios for HC, CO, NOx, and PM2,5

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

232


-------
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. 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.158,159 Zielinska et al. 2008158 reported that the
EC/PM fraction of crankcase emissions from two pre-2007 diesel buses is 1.57 percent.

To account for the different speciation of exhaust and crankcase emissions, the crankcase emission factors
for PM species shown in Table 6-4 have been back-calculated such that the total crankcase PM2.5 emissions
are 20 percent of the PM2 5 exhaust measurements (consistent with Table 6-3) and have an EC/PM split of
1.57 percent. The start and extended idle crankcase ratios are the same, because the pre-2007 start and
extended idle exhaust EC/PM are the same ( both 46.4% as documented in Section 2.2.2 and Section
2.3.1). The running exhaust EC/PM ratio is 79% (Section 2.1.2.1.8).

The tailpipe exhaust fractions are set equal to 1 because the tailpipe emission rates are not assumed to
include any crankcase emissions. In other words, the crankcase emissions are estimated in addition to the
tailpipe emissions.

Table 6-4. LHD45, MHD, and HHD Exhaust and Crankcase Ratios for 1960-2006 Diesel by Pollutant, Process, and

Model Year Group for PM2.5 Species

Pollutant

Process

Start

Running

Extended Idle

EC



1

1

1

nonECnonS04PM

Tailpipe Exhaust

1

1

1

S04

1

1

1

H20



1

1

1

EC



0.007

0.004

0.007

nonECnon SO4PM

Crankcase

0.367

0.937

0.367

S04

0.367

0.937

0.367

H20



0.367

0.937

0.367

6.1,3,2 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 this section.

233


-------
The crankcase ratios for 2007-2009 HDD emissions are based on the ACES Phase 1 study163, which tested
four MY 2007 engines from different manufactures (Caterpillar, Cummins, Detroit Diesel and Volvo). The
ACES Phase 1 engines and exhaust control systems were new, and underwent 125 hours of "degreening"
before the test program. Thus, they represent low-mileage, properly functioning heavy-duty diesel engine
emissions. In reporting the emission rates, the engines were anonymized as A, B, C, D and a backup engine
B' which was tested at a secondary site. The B' backup engine is the same make, model as engine B. The
ACES study conducted hot FTP cycles that sampled tailpipe exhaust emissions and repeat tests that
sampled combined tailpipe and crankcase exhaust. The crankcase emission rates for each engine were
estimated by calculating the difference between the average emissions measured with and without
crankcase emissions routed into the sampling system as shown in Equation 6-1, where average emissions
are the total mass divided by the testing time.

CCemissions,i ExhaUStvvithCC.i ExhaUStvvyucc,	Ex'jUClt'iOfl 6-1

Where:

•	CCemissions ;= crankcase emissions (grams per hour) for engine i

•	Exhaustwithcc phot-FTP cycle average emission emissions (grams per hour) for each
engine i with the crankcase routed into the sampling system

•	Exhaustw/oCC ;= hot-FTP cycle average emission emissions (grams per hour) for each
engine i without the crankcase

Then the crankcase emission rates were averaged together in Equation 6-2 with the backup engine being
treated as additional tests of engine B.

, CCengineB CCengineB'.sitel C-'engineB',site2 ,	^ Equation

l^^engineA '	3	' ^^engineC ' ^^engineD,)

CpACESavg —	^

Since ACES Phase 1 hot-FTP contained a hot-start we assume starts are accounted for in the running
values, and thus for these model years, the crankcase ratio for starts is zero. Note that for similar 2010+
vehicles, where cold start crankcase emissions were measured, positive crankcase starts were measured
for only for CO as discussed in Section 6.1.3.3. And, for extended idle, based upon data from 2010+ HDD
vehicles using the same crankcase technology that show the crankcase emission rates for extended idling
are similar to the running rates, we use the hot-FTP rates from ACES Phase 1 for the crankcase extending
idling rates. The crankcase rates for running, starts and extended idle are listed in Table 6-5.

Equation 6-3 is used to calculate TOG crankcase emissions from the methane and NMHC crankcase values,
using the MOVES NMOG/NMHC value for 2007-2009 MY diesel exhaust.6

£	£	R NMOG	Fauation

tog(A) = c„4(A) + nm„c(A),_

Where each pollutant rate is in g/hr and the ratio of NMOG/NMHC is 1.343 from the hcspeciation table in
MOVES.6

234


-------
Table 6-5 The MY 2007 crankcase exhaust rates (CCACESavg) from ACES Phase 1 FTP cycle used for running and

extended idling (g/hr)

CO

NOx

THC

ch4

NMHC

TOG

Total PM2.5

8.15

0.79

1.76

0.38

1.37

2.23

0.04

The crankcase ratios for the ACES program in Table 6-5 are calculated using Equation 6-4 where the
ExhaustACES is calculated using Equation 6-2, but using Exhaustw/oCC; in place of the crankcase values.

nn	_	CCACESavg	Equation 6-4

^¦^"ACES ratio nn	i

CCACESavg + ExhaustACES

We then calculated the crankcase ratios for use in MOVES using the MOVES MY 2007-2009 exhaust base
rates for all diesel HHD vehicles (regClass 47) weighted by the activity of short and long-haul single-unit
and combination trucks (source types 52,53,61,62) in each operating mode estimated from a preliminary
MOVES3 national scale run. The values for ExhaustMOvES are listed in Table 6-6. The extended idle rates
are based on a single operating mode in MOVES. We used the HHD diesel emission rates to match the HHD
engines tested in the ACES Phase 1 program. We then used Equation 6-5 to estimate the base crankcase to
tailpipe exhaust ratio for 2007-2009 HD vehicles.

Table 6-6: The MY2007-2009 MOVES exhaust base rates (g/hr) for running (weighted by operating mode activity) and

extended idling



CO

NOx

THC

ch4

NMHC

TOG

Total PM2.5

Running Exhaust

11.46

288.51

2.93

1.73

1.20

3.34

0.88

Extended Idle Exhaust

39.26

100.45

8.49

5.00

3.49

9.69

0.087

_ CCAGESavg	Equation 6-5

ratio.base — r h t

E,xndUblMC)VES

We assume that crankcase emissions are proportional to the tailpipe exhaust emissions across regulatory
classes and source types. As such, we use the CCratio base to derive the crankcase ratio for all heavy-duty
diesel regulatory classes and source types. The heavy-duty base crankcase ratios are shown in Table 6-8.

As mentioned in the background section, many manufactures employ an open crankcase ventilation
system with a coalescing filter, but a substantial fraction opt for a closed crankcase system where the
crankcase vapors are either routed into the engine with the fuel injection or into the exhaust stream
upstream of the aftertreatment. For developing crankcase emission ratio estimates, we model heavy-duty
diesel closed crankcase systems as having zero crankcase emissions, and reduce the base open crankcase
emission ratios to account for the fraction of open crankcase systems in the vehicle fleet.

Available certification data on the prevalence of open and closed crankcase systems was incomplete at the
time of analysis, but the available data suggested that most manufacturers use either open or closed

235


-------
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.46 We then used manufacturer production volume data from 2016-2018 (complete
data for earlier years was not available) to weight the number of open and closed systems within each
regulatory class as shown in Table 6-7. In these model years, there were between 9 and 10 engine
manufacturers which produced heavy-duty engines, with between 34 and 41 certified engine families.
Using this method, we estimated that 0% of LHD engines, 90.5% of medium heavy-duty and 67% of HHD
engines have open crankcase systems. The average value across the 2016-2017 model years and by
regulatory classes was used for all 2007 and later engines.47

Table 6-7 Fraction of Engines with Open Crankcase Systems by Vehicle Regulatory Class and Model Year

Model Year

LHD

MHD

HHD

2016

0

0.912

0.725

2017

0

0.919

0.635

2018

0

0.884

0.640

Average

0

0.905

0.666

Finally, to estimate the MY 2007-2009 crankcase emission ratios for MOVES, we assume that crankcase
emissions are proportional to the exhaust emissions across each of the engine-certified regulatory classes
and source types. To estimate the crankcase emission ratio for each regulatory class, we multiplied the
open crankcase ratio, CCratio base, by the open crankcase fraction of each heavy-duty regulatory class,
OpenCCfrac,regciass as shown in Equation 6-6.

CCratio,regClass,modelyear — CCratio.base * OpenCCfrac	Equation 6-6

Where:

•	CCratio,regCiass,model year= the crankcase ratio used in MOVES by regulatory class and
model year

•	CCratio,base= the heavy-duty open crankcase ratio calculated from Equation 6-5

•	OpenCCfrac regCiass= the fraction of open crankcase systems by regulatory class
determined from certification data and manufacture production volume as shown in Table
6-7

46	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.

47	We recognize the uncertainties in applying open crankcase values from model year 2016-2018 engines to model
year 2007-2009 engines. However, the larger uncertainty in the crankcase emission rates is driven by the measured
crankcase emission rates. Assuming an HHD open crankcase percentage of 67% decreased the crankcase tailpipe
emission rates by roughly one third, which is well within the range of variability of emissions observed in the ACES
Phase 1 crankcase emission results.

236


-------
The MOVES 2007-2009 crankcase ratios for THC, CH4, NMHC, TOG, NOx, CO, and PM2.5 for each crankcase
process for LHD45, MHD, HHD and Urban Bus regulatory classes are shown in Table 6-8.

237


-------
Table 6-3 Crankcase/Tail pipe Ratios for Model Year 2007-2009 Engine-certified Vehicles by Heavy-duty Diesel

Regulatory Class

Process

Pollutant

HD baseline
(Equation 6-5)

Crankcase/Tailpipe ratio by regulatory class
(Equation 6-6)

LHD45

MHD

HHD

Urban Bus

Running

THC

0.600

0

0.543

0.400

0.400

ch4

0.223

0

0.201

0.148

0.148

NMHC

1.141

0

1.032

0.760

0.760

TOG

0.667

0

0.603

0.444

0.444

<

o
u

0.005

0

0.004

0.003

0.003

CO

0.711

0

0.644

0.474

0.474

NOx

0.003

0

0.002

0.002

0.002

PM

0.043

0

0.039

0.029

0.029

Starts

THC

0

0

0

0

0

ch4

0

0

0

0

0

NMHC

0

0

0

0

0

TOG

0

0

0

0

0

<

o
u

0

0

0

0

0

CO

0

0

0

0

0

NOx

0

0

0

0

0

PM

0

0

0

0

0

Extended
Idle

THC

0.207

0

0.187

0.138

0.138

ch4

0.077

0

0.070

0.051

0.051

NMHC

0.394

0

0.356

0.262

0.262

TOG

0.230

0

0.208

0.153

0.153

<

o
u

0.054

0

0.049

0.036

0.036

CO

0.208

0

0.188

0.138

0.138

NOx

0.008

0

0.007

0.005

0.005

PM

0.436

0

0.394

0.290

0.290

A C02 is not included in the MOVES crankcaseEmissionRatio table, in part due to its small fraction
compared to running exhaust C02. It is included here for comparison with other pollutants.

238


-------
For PM2.5 emissions, MOVES applies crankcase ratios to each of the intermediate PM2.5 species (EC,
nonECnonS04PM, S04, and H20). The MOVES PM2.5 speciation profile developed from the ACES Phase 1
study combined the crankcase and tailpipe emissions. As such, we model crankcase emissions as having
the same speciation as tailpipe emissions, and the crankcase fractions for the intermediate PM2.5 species in
Table 6-9 are the same as derived for total PM2.5 in Table 6-8. In MOVES4, we set the tailpipe ratios for
model year 2007-2009 equal to one because we now assume that the tailpipe emission factors only include
tailpipe exhaust. This approach was taken because the 2007-2009 model year running PM2.5 emission rates
are based on assumed reductions from MY 1998-2006 vehicles based on certification data (See Section
2.1.2.1.7). Assuming the resulting running PM emission rates in 2007-2008 only include tailpipe exhaust
was deemed equally valid as our previous assumption, and simplifies the calculation of crankcase
emissions by keeping all tailpipe ratios in the crankcase calculator equal to one. The extended idle
emissions for model year 2007-2009 are based on tailpipe measurements only, and using a extended idle
ratio equal to one is consistent with the measured data. The peer-reviewers comments from the MOVES3
crankcase review agreed with this update.168

The crankcase emission ratios for MHD regulatory class for the intermediate PM species are the same as
total PM2.5 (0.039 for running and 0.394 for extended idle) as shown in Table 6-8.

Table 6-9. MOVES Exhaust and Crankcase Ratios for Model Year 2007-2009 HHD Diesel by Pollutant and Process for

PM2.5 Species

Process

Model Year Group

Pollutant

Start

Running

Extended Idle

Tailpipe
Exhaust

2007-2009

EC

1

1

1

nonECnonS04PM

1

1

1

S04

1

1

1

H20

1

1

1

Crankcase

2007-2009

EC

0

0.0290

0.2929

nonECnon SO4PM

0

0.0290

0.2929

S04

0

0.0290

0.2929

H20

0

0.0290

0.2929

6.1,3,3 2010-2026 Model Years

The HDIUT program (see Section 2.1.1.2) is used as the source of baseline exhaust emission rates for model
year 2010-2026 diesel vehicles. As the HDIUT program measures tailpipe exhaust emissions from trucks in-
use, the crankcase is not routed to the tailpipe and therefore not accounted for in the data. To account for
crankcase emissions for 2010 and later diesel vehicles, we used direct crankcase measurements of NOx,
THC, CH4, and CO emissions from the US EPA's National Vehicle and Fuel Emissions Laboratory (NVFEL)
testing of two heavy duty trucks (MY 2015 and MY 2019). Each of the trucks had less than 10,000 miles and
represent properly functioning low-mileage heavy-duty vehicles. Testing was conducted on a chassis
dynamometer over a drive cycle that consists of a hot or cold start followed by an ARB transient cycle
(Phase 1), followed by four repetitions of the same ARB transient (Phase 2), 10 minutes of idling (Phase 3),

239


-------
and steady-state highway activity at 55 mph and 60 mph (Phase 4). The speed trace and a graphical
indication of the testing phases are shown in Figure 6-1.

0	1000	2000	3000	4000	5000	6000

Time (s)

Figure 6-1 Speed trace of the NVFEL testing program along with each phase

There were a total of thirteen repetitions between the two trucks, with soak times between repetitions of
either between one and three hours (hot start) or greater than twelve hours (cold start) as shown in Table
6-10.

Table 6-10: Testing information for the NVFEL test program

Truck ID

Number of Tests

Dates of Testing

Soak Times

1

6 (3 Hot Start, 3 Cold Start)

May 30, 2017-
June 1, 2017

Hot Starts 1-3 hrs
Cold Starts 12+ hours

2

7 (3 Hot Start, 4 Cold Start)

August 20, 2019 -
August 13, 2019

Hot Starts 1-3 hrs
Cold Starts 12+ hours

Table 6-llTable 6-11 reports the average emission rates across the repetitions by truck and phase, and by
phase only (weighting the two trucks equally).

240


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



Crankcase CO
(g/hr)

Crankcase NOx
(g/hr)

Crankcase THC
(g/hr)

Crankcase CH4
(g/hr)

Phase 1 Average

1.0900

0.4427

0.3463

0.0324

Truck 1 average

0.5057

0.5527

0.2373

0.0053

Truck 2 average

1.6742

0.3327

0.4552

0.0595

Phase 2 Average

0.6224

0.9086

0.4537

0.0155

Truck 1 average

0.1916

1.0316

0.2835

0.0001

Truck 2 average

1.0532

0.7855

0.6238

0.0309

Phase 3 Average

0.2461

0.7097

0.4529

0.0058

Truck 1 average

0.1670

0.7633

0.2634

0.0000

Truck 2 average

0.3252

0.6560

0.6424

0.0115

Phase 4 Average

0.2679

1.8447

0.5915

0.0053

Truck 1 average

0.1521

1.7766

0.3711

0.0000

Truck 2 average

0.3836

1.9127

0.8118

0.0106

Average of all phase averages

0.5677

0.9764

0.4611

0.0148

The average running rate was calculated by calculating a weighted average between Phase 2 (36.7%) and
Phase 4 (63.3%). The phase-weighting was calculated to match the national running operating mode
distribution for MY2015, HHD diesel vehicles in short- and long-haul single-unit and combination trucks
(source type 52,53,61,62) as estimated in a MOVES national run using a draft version of MOVES3 for
calendar year 2015.72 Phase 2 had transient operation below 50 mph and was mapped to operating modes
0-30 (36.7% of total activity) and Phase 4 had constant high speed data and was mapped to operating-
modes 33-40 (63.3% of total activity) as shown in Table 6-12. HHD diesel vehicles in short- and long-haul
single-unit and combination trucks vehicles were chosen as to match the NVFEL testing vehicles.

241


-------
Table 6-12. Assignment of MOVES Operating Modes to Test Phase from the NVFEL data, and Phase weighting
determined from the Operating Mode activity from a National draft MOVES run for MY 2015 HHD Vehicles in the
Short- and Long-haul Single-unit and Combination Trucks Source Types

OpModelD

Operating Mode
Description

Scaled Tractive
Power (STPj,
skW)

Vehicle
Speed (i/t,
mph)

% of Total
Activity from
MOVES run

NVEL Test
Phase

0

Deceleration/Braking





2.69



1

Idle



vt < 1.0

5.40



11

Coast

STPt< 0

1 < vt < 25

4.55



12

Cruise/Acceleration

0 < STPt< 3

1 < vt < 25

6.62



13

Cruise/Acceleration

3 < STPt< 6

1 < vt < 25

1.64



14

Cruise/Acceleration

6 < STPt< 9

1 < vt < 25

1.00



15

Cruise/Acceleration

9 < STPt< 12

1 < vt < 25

0.77



16

Cruise/Acceleration

12 < STPt

1 < vt < 25

1.36



21

Coast

STPt< 0

25 < vt < 50

3.37

Phase 2

22

Cruise/Acceleration

0 < STPt< 3

25 < vt < 50

1.34



23

Cruise/Acceleration

3 < STPt< 6

25 < vt < 50

1.54



24

Cruise/Acceleration

6 < STPt< 9

25 < vt < 50

1.52



25

Cruise/Acceleration

9 < STPt< 12

25 < vt < 50

1.04



27

Cruise/Acceleration

12 < STPt< 18

25 < vt < 50

1.86



28

Cruise/Acceleration

18 < STPt< 24

25 < vt < 50

0.98



29

Cruise/Acceleration

24 < STPt< 30

25 < vt < 50

0.53



30

Cruise/Acceleration

30 < STPt

25 < vt < 50

0.49



33

Cruise/Acceleration

STPt< 6

50 < vt

16.26



35

Cruise/Acceleration

6 < STPt< 12

50 < vt

14.90



37

Cruise/Acceleration

12 < STPt<18

50 < vt

13.61

Phase 4

38

Cruise/Acceleration

18 < STPt< 24

50 < vt

8.60

39

Cruise/Acceleration

24 < STPt< 30

50 < vt

5.05



40

Cruise/Acceleration

30 < STPt

50 < vt

4.89



242


-------
The average start rate was calculated by taking the difference between time-based emission rates (g/hr) of
Phase 1 and Phase 2 of each test and averaging the differences together as shown in Equation 6-7 and
multiplying by the length of Phase 1.

£,n=i(CCphaselJ -CCphase2J) / 1 hour \	Equation 6-7

CCstart =	x 11 minutes x f

n

/ 1 Hour \

\60 minutes)

Where:

•	CCstart= average crankcase (g/start) emission from all the test runs

•	CCphase 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 6-7 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. Thus,
we set the start crankcase emission rate to zero for THC and NOx. The average CH4 start emission rate was
positive, but because methane is estimated in MOVES as a fraction of THC, we also set it equal to zero as
shown in Table 6-13. For the crankcase extended idling rate we used the average of emissions in the idling
phase (Phase 3).

These rates also informed the base crankcase rates for NMHC and TOG. We used Equation 6-3 to calculate
TOG crankcase emissions, with the MOVES NMOG/NMHC ratio for 2010+ diesel running exhaust (1.085)
from the hcspeciation table in MOVES.6

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.169 Thus, by using tailpipe emissions
from MOVES, we yield more stable crankcase ratios.

The ratios for NMHC, TOG, and NOx are applied to the pollutants chained to them as well (Table 6-1).

243


-------
Table 6-13 Average Crankcase Emission Rates for MOVES Processes from NVFEL data

Process

CO

NOx

THC

CH4

NMHC

TOG

Crankcase Running
Exhaust (g/hr)

0.398

1.501

0.541

0.009

0.5318

0.5861

Crankcase Start Exhaust
(g/start)

0.049

0

0

0

0

0

Crankcase Extended
Idling Exhaust (g/hr)

0.246

0.710

0.453

0.006

0.4471

0.4909

No PM2.5 crankcase emission measurements were made in the NVFEL study, so the 2010 and later model
year calculations use the crankcase PM2.5 rates from ACES Phase 1 report as described for model years
2007-2009 in Section 0. More information on the PM calculations for model year 2010 and later vehicles is
provided later in this section.

We used the base exhaust emission rates from the NOx FEL 0.2 group for HHD engines in short and long-
haul single-unit and combination trucks (source types 52,53,61,62). These vehicles were chosen from the
full HDIUT data set to match the NVFEL testing vehicles.

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

Extended Idling (g/hr)



Model Year
Group

CO

NOx

THC

ch4

NMHC

TOG

Total PM2.5

Running Exhaust

2010-2013

125.42

62.51

1.50

0.57

0.93

1.58

0.15

2014+

60.06

61.87

1.47

0.56

0.91

1.55

0.12

Extended Idle Exhaust

2010-2013

39.26

42.60

2.75

1.04

1.70

2.89

0.03

2014+

39.26

42.60

1.64

0.62

1.01

1.72

0.02

The HHD tailpipe rates have two model year groups 2010-2013 and 2014+, so the ratios differ for these
model year groups as shown in Table 6-15.

As was done for the development of the 2007-2009 crankcase emission ratios documented in Section
6.1.3.2 we assume that the crankcase ratios derived for MY2010-2026 heavy heavy-duty diesel (HHD)
engines apply to the other heavy-duty regulatory classes (LHD45, MHD and Urban Bus) and source types
(other buses, transit buses, school buses, refuse trucks, and motorhomes), and that crankcase emissions
are proportional to the exhaust emissions across regulatory classes and source use types. Additionally, we
use Equation 6-6 to account for the fraction of open crankcase systems within each regulatory class shown
in Table 6-7.

Table 6-15 Crankcase/Tail pipe Ratios for Model Year 2010-2026 Heavy-Duty Diesel Regulatory Classes

Process

Pollutant

Model Year
Group

HD

baseline

Crankcase/Tailpipe ratio by regulatory
class

LHD45

MHD

HHD

Urban Bus

Running

THC

2010-2013

0.36

0

0.33

0.24

0.24

2014-2026

0.37

0

0.33

0.24

0.24

244


-------


ch4

2010-2013

0.02

0

0.014

0.011

0.011



2014-2026

0.02

0

0.015

0.011

0.011



NMHC

2010-2013

0.57

0

0.52

0.38

0.38



2014-2026

0.58

0

0.53

0.39

0.39



TOG

2010-2013

0.37

0

0.34

0.25

0.25



2014-2026

0.38

0

0.34

0.25

0.25



o
u

2010-2013

0

0

0.002

0.002

0.002



2014-2026

0

0

0.002

0.002

0.002



CO

2010-2013

0

0

0.003

0.002

0.002



2014-2026

0.01

0

0.006

0.004

0.004



NOx

2010-2013

0.02

0

0.022

0.016

0.016





2014-2026

0.02

0

0.022

0.016

0.016



PM2.5

2010-2013

0.26

0

0.24

0.17

0.17



2014-2026

0.32

0

0.29

0.21

0.21



THC

2010-2013

0

0

0

0

0



2014-2026

0

0

0

0

0



ch4

2010-2013

0

0

0

0

0



2014-2026

0

0

0

0

0



NMHC

2010-2013

0

0

0

0

0



2014-2026

0

0

0

0

0



TOG

2010-2013

0

0

0

0

0

Starts

2014-2026

0

0

0

0

0

O
u

2010-2013

0.03

0

0.03

0.02

0.02



2014-2026

0.03

0

0.03

0.02

0.02



CO

2010-2013

0.16

0

0.15

0.11

0.11



2014-2026

0.16

0

0.15

0.11

0.11



NOx

2010-2013

0

0

0

0

0





2014-2026

0

0

0

0

0



PM2.s

2010-2013

0

0

0

0

0



2014-2026

0

0

0

0

0


-------
Table 6-15 (Continued} Crankcase/Tailpipe Ratios for Model Year 2010-2060 Heavy-Duty Diesel Regulatory Classes

Process

Pollutant

Model Year
Group

HD

baseline

Crankcase/Tailpipe ratio by regulatory
class

LHD45

MHD

HHD

Urban Bus

Extended
Idle

THC

2010-2013

0.16

0

0.15

0.11

0.11

2014-2026

0.28

0

0.25

0.18

0.18

ch4

2010-2013

0.005

0

0.005

0.004

0.004

2014-2026

0.009

0

0.008

0.006

0.006

NMHC

2010-2013

0.26

0

0.24

0.17

0.17

2014-2026

0.44

0

0.4

0.29

0.29

TOG

2010-2013

0.17

0

0.15

0.11

0.11

2014-2026

0.28

0

0.26

0.19

0.19

O
u

2010-2013

0.015

0

0.013

0.01

0.01

2014-2026

0.015

0

0.013

0.01

0.01

CO

2010-2013

0.006

0

0.006

0.004

0.004

2014-2026

0.006

0

0.006

0.004

0.004

NOx

2010-2013

0.017

0

0.015

0.011

0.011

2014-2026

0.017

0

0.015

0.011

0.011

PM2.5

2010-2013

1.11

0

1.01

0.74

0.74

2014-2026

1.82

0

1.64

1.21

1.21

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

246


-------
Table 6-16. MOVES Exhaust and Crankcase Ratios for 2010-2026 MHD and HHD Diesel by Pollutant, Process, and

Model Year Group for PM2.5 Species

Process

Model Year
Group

Pollutant

Start

Running

Extended
Idle





EC

1

1

1



2010-2013

nonECnonS04PM

1

1

1



S04

1

1

1

Tailpipe



H20

1

1

1

Exhaust



EC

1

1

1



2014-2026

nonECnonS04PM

1

1

1



S04

1

1

1





H20

1

1

1





EC

0

0.17

0.74



2010-2013

nonECnonS04PM

0

0.17

0.74



S04

0

0.17

0.74

HHD



H20

0

0.17

0.74

Crankcase



EC

0

0.21

1.21



2014-2026

nonECnonS04PM

0

0.21

1.21



S04

0

0.21

1.21





H20

0

0.21

1.21





EC

0

0.24

1



2010-2013

nonECnonS04PM

0

0.24

1



S04

0

0.24

1

MHD



H20

0

0.24

1

Crankcase



EC

0

0.29

1.64



2014-2026

nonECnonS04PM

0

0.29

1.64



S04

0

0.29

1.64





H20

0

0.29

1.64

6.1,3,4 2027-2060 Model Years

For MY2027 and later vehicles, the HD2027 standards require manufacturers to use one of two options for
controlling crankcase emissions, either: 1) closing the crankcase, or 2) an updated version of the current

247


-------
requirements for an open crankcase that includes additional requirements for measuring and accounting
for crankcase emissions.

In the emissions impact analysis of the HD2027 rule, we assumed that closing the crankcase would be the
preferred option to meet the standards. We revised the crankcase emission rates accordingly by setting
the crankcase emission rates for MY2027+ HHD, MHD and LHD45 diesel vehicles to zero.

6.1.4 Glider Crankcase Emissions

Glider vehicles in MOVES are modeled using emission rates from MY 2000 heavy-duty diesel engines
(Section 2.5); the pre-2007 crankcase rates (Table 6-8 and Table 6-9) are applied to glider vehicles
(regClassID 49) for all model years.

6.2 Heavy-Duty Gasoline and CNG Crankcase Emissions

The data on heavy-duty gasoline and CNG crankcase emissions are limited. All 1969 and later spark ignition
heavy-duty engines are required to control crankcase emissions. All gasoline engines are assumed to use
positive crankcase ventilation (PCV) systems, which route the crankcase gases into the intake manifold. For
heavy-duty gasoline engines we use the same values of crankcase emission ratios as light-duty gasoline as
shown in Table 6-17; these are documented in the MOVES light-duty emission rates report.12 The HD2027
standards do not affect heavy-duty gasoline and CNG fueled engines since the rule only affects the
crankcase emissions from the compression-ignition (diesel) heavy-duty engines.

For the 1969 and later vehicles, we assume 4 percent of PCV systems fail, which would cause increased
tailpipe emissions for reasons such as misfiring of the engine, lubricating oil in the intake manifold, and
increased deterioration of the three-way catalyst. Although these processes will increase tailpipe
emissions, in MOVES we model this increase of emissions due to a failed PCV as crankcase emissions. We
assume that the elevated emissions due to a failed PCV system would be equivalent to the crankcase
emissions of a pre-1969 vehicles using Equation 6-8. The resulting fleet-wide crankcase to exhaust
emission ratios for 1969 and later vehicles are shown in Table 6-17.

Crankcase Ratioiy6y+ = PCV failure rate (4%) x Crankcase Ratioprel969	Equation 6-8

Table 6-17 Crankcase to Tailpipe Exhaust Emission Ratio for Heavy-Duty Gasoline and CNG Vehicles for THC, CO, NOx,

and PM2.5

Pollutant

pre-1969

1969 and later

HC

0.33

0.013

CO

0.013

0.00052

NOx

0.001

0.00004

PM (all species)

0.20

0.008

Due to limited information, we used the gasoline heavy-duty crankcase emission factors for heavy-duty
CNG engines because the majority of these engines are spark-ignited. However, at least one study (Clark et
al., 2017)170 suggests that CNG vehicles have open crankcase systems so we may be underestimating

248


-------
crankcase emissions, especially those of methane. We hope to revisit CNG crankcase emissions in future
versions of MOVES.

The crankcase and exhaust ratios used by the crankcase calculator for PM25 emissions from heavy-duty
gasoline and compressed natural gas vehicles are provided in Table 6-18. These values are applied to
calculate crankcase emissions associated with start exhaust as well as to running exhaust. No information
is available to estimate separate speciation between exhaust and crankcase, so the factors are the same
for all PM subspecies.

Table 6-18 MOVES Exhaust and Crankcase Ratios for Heavy-Duty Gasoline and CNG Vehicles by Pollutant, Process,
Model Year Group, and Fuel Type, and Source Type for PM2.5 Species

Pollutant

Process

1960-1968

1969-2050

EC



1

1

nonECnonS04PM

Exhaust

1

1

S04

1

1

H20



1

1

EC



0.2

0.008

nonECnonS04PM

Crankcase

0.2

0.008

S04

0.2

0.008

H20



0.2

0.008

7 Nitrogen Oxide Composition

This section discusses the values used to estimate nitric oxide (NO), nitrogen dioxide (N02) 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 N02. In other words, the molar mass of N02 (46 g/mole) is used to
calculate grams of NOx from the molar concentration of NOx.

Nitrogen oxides (NOx) are defined as NO + N02.171,172 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.171 NOz compounds are formed in the atmosphere as oxidation
products of NOx.172

Chemiluminescent analyzers used for exhaust NOx measurements directly measure NO, as NO is oxidized
by ozone to form N02 and produces florescent light. Chemiluminescent analyzers measure NOx (NO + N02)
by using a catalyst that reduces the N02to NO in the sample air stream before measurement. N02 is
calculated as the difference between NOx and NO measurements. The NOx converter within
chemiluminescent analyzers can also reduce other reactive nitrogen species (NOz), including HONO to NO.
If the concentrations of NOz-interfering species in the sample stream are significant relative to N02
concentrations, then they can bias the N02 measurements high.173

249


-------
MOVES estimates NO and N02 by applying an NO/NOx or N02/N0Xfraction to the NOx emission rates. The
NO/NO2 and N02/NOx fractions are stored in a MOVES table called nono2ratio. The nono2ratio enables the
nitrogen oxide composition to vary according to source type, fuel type, model year, and pollutant process.
However, the current NOx fractions in MOVES vary only according to fuel type, model year, and emission
process.

MOVES also estimates one important NOz species, nitrous acid (HONO), from the NOx values. HONO
emissions are estimated as a fraction (0.8 percent) of NOx emissions from all vehicle types in MOVES,
based on HONO and NOx measurements made at a road tunnel in Europe.174 HONO emissions are also
estimated using the nono2ratio MOVES table. For each source type, fuel type, and emission process, the
NO, N02, and HONO values in the nono2ratio sum to one. Future work could be conducted to update
MOVES to model NOx and HONO fractions according to regulatory class.

MOVES users should be aware that the definition of NOx in MOVES (NO+NO2+HONO) is different than the
standard NOx definition of NOx (NO + N02). 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 N02 emissions. To avoid
overcounting reactive nitrogen formation, we include HONO in the sum of NOx in MOVES. MOVES users
should consider which measure they would like to use depending on their use-case. For example, for
comparing NOx results with a vehicle emission test program, MOVES users may want to simply use NOx
(pollutantID 3), whereas MOVES users developing air quality inputs of NOx, N02, and HONO, may estimate
NOx as the sum of NO + N02 (pollutantlDs 32 and 33), rather than using the direct NOx output in MOVES
(polluantID 3).

7.1 Heavy-Duty Diesel

The heavy-duty diesel NO/NOx, N02/NOx,and HONO/NOx fractions were updated in MOVES4 using data
reported from recent emission studies, as described below. We summarized N02/N0Xfractions from three
recent studies by aftertreatment technology and model year range shown in Table 7-1.

Preble et al. (2019)139 sampled individual heavy-duty vehicle exhaust plumes at the entrance to the
Caldecott Tunnel near Oakland, California and at the Port of Oakland for multiple years. The data from
Preble et al. (2019) are also used to update the NH3 emission rates as discussed in Section 5.1, however
NH3 was only measured in 2018 at the Caldecott TunnelwhileNO and N02 were measured at both locations
for multiple years. Thirugengadam et al. (2015)50 conducted exhaust sampling of five heavy-duty diesel
vehicles measured on four different driving cycles used to represent goods movement in Southern
California. Quiros et al. (2016)48 sampled six heavy-duty diesel tractors hauling a mobile emissions
laboratory trailer. They sampled the vehicles along six routes intended to represent goods movement in
Southern California. The Advanced Collaborative Emissions Study conducted by Khalek et al. (2009)46 and
(2013)175, tested four model year 2007 and three model year 2010 heavy-duty diesel engines using an
engine dynamometer.

The N02/NOx fraction measured by aftertreatment technology and model year ranges are quite consistent
across the four different studies. This suggests that the N02/NOx obtained from the plume capture
measurements in Preble et al. (2019)139 are relevant for the wide range of operation conditions sampled in
Thirugengadam et al. (2015)50 and Quiros et al. (2016).48

Each of the studies showed that the N02/NOx increased with the introduction of diesel particulate filters
(DPF) in model years 2007-2009. This is expected because DPF aftertreatment systems are designed to

250


-------
increase the fraction of N02to facilitate passive regeneration of the DPF. A diesel oxidation catalyst
upstream of the DPF oxides NO to N02 which then oxidize soot collected on the filter.176 The DPF+SCR
aftertreatment systems introduced with MY 2010 and later engines also have higher N02/N0X fractions
than pre-DPF engines but are consistently lower than the DPF only engines.

Table 7-1. N02/N0X ratios (± 95% Confidence Intervals, if available) from heavy-duty diesel vehicles reported from

recent studies

Study

Study Description

Sample
Size

Aftertreatment

Engine
Model Year

NOz/NOx

Preble et al.
(2019)139

Caldecott Tunnel near
Oakland California,
Plume-Capture, Sample
Years: 2014, 2015, 2018

1,471

DPF + SCR

2010-2018

0.19 ±0.03

780

DPF

2007-2009

0.24 ±0.02

359

DPF Retrofit

1994-2006

0.11 ±0.02

190

No DPF

2004-2006

0.06 ±0.01

454

No DPF

1965-2003

0.03 ±0.01

Preble et al.
(2019)139

Port of Oakland,
California, Plume-
Capture, Sample Years:
2011, 2013, 2015

403

DPF + SCR

2010-2016

0.20 ±0.05

1,598

DPF

2007-2009

0.23 ±0.02

399

DPF Retrofit

1994-2006

0.15 ±0.02

199

No DPF

2004-2006

0.04 ±0.02

Thiruvengadam
et al. (2015)50

Chassis dynamometer
on four duty cycles
representative of goods
movement

1

DPF + SCR

2010-2011

~0.15

1

DPF

2011

~0.30

1

DPF

2009

~.30

Quiros et al.
(2016)48

Six good movements
routes in Southern
California sampled
using mobile laboratory

4

DPF + SCR

2013-2014

0.19 ±0.17

1

DPF (Hybrid Diesel)

2011

0.33

1

DPF

2007

0.30

Khalek et al.
(2013)175

ACES engine
dynamometer study,
16-hour cycle

3

DPF + SCR

2011

0.52 ±0.45

Khalek et al.
(2009)46

4

DPF

2007

0.54 ±0.20

We used the NO/NOxand N02/N0Xfractions from the Caldecott Tunnel (Preble et al., 2019).139 For model
years 2004-2010, the updated values are similar to the values used in previous versions of MOVES. For
model years earlier and later, the MOVES4 N02/NOx values are lower.

We did not use the DPF retrofit values because these are representative of California drayage vehicles
starting in 2010, but not the nation-wide fleet of heavy-duty pre-2007 vehicles.

MOVES3 and earlier versions use a HONO fraction of 0.8% obtained from Kurtenbach et al. (2001).174

251


-------
Table 7-2 summarizes HONO/NOx ratios from a 2023 literature review. Studies that measure HONO often
don't measure individual vehicle exhaust, thus isolating the diesel specific HONO ratio is difficult. MOVES
continues to use the value of 0.8% since it is well-within the range of the diesel-only HONO/NOx
measurements.

Table 7-2. Fleet-average and diesel specific HONO/NOx Ratios

Source

Study Type

HONO/NOx (%)

Diesel fleet (%)

Kramer et al. (2020)177

Road tunnel in the
United Kingdom

1.04

Isolated diesel vehicle ratio

0.85

66

Liang et al. (2017)178

Road tunnel in Hong
Kong

1.24

33

Xu et al. (2015)179

Ambient
measurements in
Hong Kong

1.20

33

Trinhetal. (2017)180

Chassis

dynamometer across
four drive cycles

0.16 to 1

Diesel vehicle equipped with DPF
tested

Rappengliick et al.
(2013)181

Road-side
measurements in
Houston, Texas

1.17

5-10

Kurtenbach et al.
(2001)174

Tunnel Study in
Germany

0.80

6% heavy-duty vehicles, 6%
commercial vans, 12.3% diesel
passenger vehicles

Single-vehicle Tunnel
Study

0.53

Diesel truck

Single-vehicle Tunnel
Study

0.66

Diesel passenger car

Table 7-3 shows the NOx and HONO fractions for heavy-duty diesel vehicles used in MOVES. Vehicle model
years subject to the HD2027 rule use the same fractions as model years 2010-2026, and APU exhaust
fractions are the same for all model years 2024 and later. The NO/NOxand N02/NOxfractions reported in
Preble et al. (2019) were renormalized to account for the 0.8 percent HONO emissions. The NOx fractions
are the same across all diesel source types and across all emission processes (running, start, extended idle),
except for auxiliary power units, which use the conventional NOxfractions (1950-2003) for all 1950-2023
model years because it is assumed that these APUs are not fitted with diesel particulate filters. APU
exhaust rates for the model year range 2024-2060 use the same NOx fractions as model year 2007-2009
running exhaust, because we assume they will be equipped with DPF systems but not SCR systems.

Because the nono2ratio table is classified by source type, and not regulatory class, gliders use the same
NO/N02 fractions as the other regulatory classes by model year, even though the 1960-2003 NOx fractions
are more relevant for this regulatory class. We hope to address this design limitation in future versions of
MOVES.

Finally, while the HD2027 rule as updated in MOVES4 will reduce NOxemissions from MY 2027+ HD
gasoline vehicles, we modelled no change in the NO/NO2 fractions.

252


-------
Table 7-3 N0X and HONO Fractions for Heavy-Duty Diesel Vehicles

Process

Model Year

NO

NOz

HONO

Running exhaust,
start exhaust,
extended idle
exhaust

1950-2003

0.9622

0.0298

0.008

2004-2006

0.9325

0.0595

0.008

2007-2009

0.7539

0.2381

0.008

2010-2060

0.8035

0.1885

0.008

Auxiliary power
unit exhaust

1960-2023

0.9325

0.0595

0.008

2024-2060

0.7539

0.2381

0.008

7.2 Heavy-Duty Gasoline

The NOx fractions for heavy-duty gasoline are based on the MOVES values used for light-duty gasoline
estimates. Separate values are used for running and start emission processes. As stated in the MOVES2010
report/82 the light-duty values are shifted to later model year groups to be consistent with heavy-duty
emission standards and emission control technologies. These values are shown in Table 7-4 for both light-
duty and heavy-duty gasoline vehicles. The N02fractions originally developed for MOVES2010 were
reduced by 0.008 to account for the HONO emissions.182 While the HD2027 rule will reduce NOxemissions
from HD gasoline vehicles, we modelled no change in the NOx fractions for MY 2027+.

Table 7-4 NOx and HONO Fractions for Light-Duty (Source Type 21, 31, 32) and Heavy-Duty Gasoline Vehicles (Source

Type 41, 42, 43, 51, 52, 53, 54, 61)

Light-Dutv
gasoline
model vear
groups

Heaw-Dutv
gasoline
model vear
groups

Running

Start

NO

NO,

HONO

NO

NO,

HONO

1950-1980

1960-1987

0.975

0.017

0.008

0.975

0.017

0.008

1981-1990

1988-2004

0.932

0.06

0.008

0.932

0.031

0.008

1991-1995

2005-2007

0.954

0.038

0.008

0.987

0.005

0.008

1996-2060

2008-2060

0.836

0.156

0.008

0.951

0.041

0.008

7.3 Heavy-Duty Compressed Natural Gas

We used the average N02/ NOx fractions reported from three CNG transit buses with DDC Series 50 G
engines by Lanni et al. (2003)126 with the 0.008 HONO fraction assumed for other fuel types, to estimate
the NOx fractions of NO, N02, and HONO. These assumptions yield the values in Table 7-5, which are used
for CNG heavy-duty vehicles of all model years. In the future, we hope to update these values with data
from more recent three-way catalyst CNG vehicles.48

253


-------
Table 7-5 N0X and MONO Fractions CNG Heavy-Duty Vehicles

Model Year

NO

NOz

HONO

1950-2060

0.865

0.127

0.008

254


-------
8 Appendices

A. Calculation of Accessory Power Requirements

Table A-l Accessory Load Estimates for HHD 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-2Accessory 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

255


-------
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

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 MOVES2010183 from studies
conducted between 1988 and 2007 (See Section B.2.) Slight updates to NOx and PM25T&M adjustment
factors were made for MOVES2014.34 No changes were made to the T&M assumptions or data between
MOVES2014 and MOVES3. Minor corrections to the warranty and useful life of LHD vehicles were made in
MOVES4. In MOVES4, we also updated the warranty and useful life of the MY2027+ heavy-duty vehicles
based on the HD2027 standards and we calculated new T&M percentage effects for these vehicles that
assume aftertreatment failure in a MY2027 vehicle would bring NOxtailpipe emissions to the same level as
a MY 2010 vehicle with the same failure. For MOVES5 we made minor corrections and clarifications to the
tampering documentation here and in the main body of this report.

In the future, T&M adjustment factors in MOVES should be re-evaluated and updated, particularly to
incorporate data on the durability and emissions performance of advanced aftertreatment systems on
modern heavy-duty diesel vehicles and to account for recent work surveying intentional tampering in
diesel trucks.184

This section describes the derivation of T&M emission rates applied to diesel running exhaust. The
estimation of heavy-duty gasoline deterioration is discussed in the derivation of the heavy-duty gasoline
rates (Section 3). The derivation of the T&M effects for diesel extended idle emissions are described in
Section 2.3.

B.l 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

256


-------
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.

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 8-1.

^T&M,p = X (T&M frequency; X T&M emission effectp i)	Equation 8-1

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 effectp estimated emission effect for pollutant p associated with
tampering & mal-maintenance failure i.

The emission rate at the end of useful life is then calculated using Equation 8-2.

257


-------
Ef^End of useful life,p,r,o ERZero mile,p,r,o ^	^T&M.p)	Equation 8 2

Where:

•	EREnd of useful iife,p,r,o = the heavy-duty diesel emission rate at the end of warranty for
each pollutant p, regulatory class, r, and operating mode, o

•	ERzero mjig= the zero-mile heavy-duty diesel emission rate for each pollutant p, regulatory
class, r, and operating mode, o

•	fT&M= the tampering and mal-maintenance adjustment factor for each pollutant p
(Equation 8-1)

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 VIUS185 (Vehicle Inventory and Use Survey) data, which contains data on how
quickly trucks of different regulatory classes accumulate mileage. Table B-l shows the emissions warranty
period and approximate useful life requirement period for each of the regulatory classes for pre-MY2027
vehicles. This table and the resulting values in MOVES were corrected in MOVES4 to reflect the 50,000
miles warranty requirement for LHD vehicles, and the 150,000 useful life mileage for Tier 3 LHD2b3
vehicles.

Table B-l Warranty and Useful Life Requirements by Regulatory Class For Pre-MY2027 Vehicles

Regulatory class

Warranty
requirement
mileage/age
requirementb

Calculated warranty
age3

Useful life

mileage/age

requirement

Calculated
useful life
age3

Assumed
mileage
per year

LHD2b3 (Tier 2 and
earlier)

50,000/5

2

120,000/11°

5

26,000

LHD2b3 (Tier 3)

50,000/5

2

150,000/10d

6

26,000

LHD45

50,000/5

2

110,000/10e

4

26,000

MHD

100,000/5

2

185,000/10e

5

41,000

HHD

100,000/5

1

435,000/10e

4

105,000

BUS

100,000/5

2

435,000/10f

10

44,000

aThe calculated warranty age and useful life age here are based on typical miles driven by vehicles in the
regulatory class. For example, HHD vehicles typically accumulate a large number of miles per year
(100,000+/year). Thus, HHD vehicles complete their warranty and useful life requirements based on
mileage while the vehicle age is still much below the requirement.
b 40 CFR 1037.120

c 40 CFR 86.096-2, 40 CFR 86.1805-12
d 40 CFR 86.1805-17

e 40 CFR 86.001-2 (4). The useful life mileage is the same for each regulatory class for all exhaust
pollutants (NOx, HC, CO, and PM). The useful life age requirement is generally 10 years for NOx, while it
is 8 years for the other pollutants (and for NOx in 1996-1997). However, we calculated that the mileage
requirement is the forcing requirement for all the heavy-duty regulatory classes.
f 40 CFR 86.098-2. The useful life standard is 10 years for urban buses for both NOx and PM.

258


-------
Starting from MY2027, the HD2027 standards require manufacturers to comply with new warranty and
useful life provisions as shown in Table B-2.

Table B-2 Warranty and Useful Life Requirements by Regulatory Class For MY2027+ Vehicles

Regulatory class

Warranty
requirement
mileage/age
requirement

Calculated warranty
age

Useful life

mileage/age

requirement

Calculated
useful life
age

Assumed
mileage
per year

LHD2b3

Same as Pre-MY2027+ Vehicles

LHD45

210,000/10

8

270,000/15

10

26,000

MHD

280,000/10

7

350,000/12

9

41,000

HHD

450,000/10a

4

650,000/11

6

105,000

BUS

450,000/10a

10

650,000/11

15

44,000

aThe HHD diesel and Urban Bus warranty year values in the ta
to be consistent with HD2027 final rule.

Dies were updated to 10 years in MOVES4

While both age and mileage metrics are given for these periods, whichever comes first determines the
applicability of the warranty. As a result, since the mileage limit is usually reached before the age limit, but
MOVES deals with age and not mileage, we needed to convert all the mileage values to age equivalents.
The data show that on average, heavy heavy-duty trucks accumulate mileage much more quickly than
other regulatory classes and reach the end of their warranty period more quickly. Therefore, deterioration
in heavy heavy-duty truck emissions will presumably happen at younger ages than for other regulatory
classes. Buses, on average, do not accumulate mileage as quickly. Therefore, their useful life period is
governed by the age requirement, not the mileage requirement.

We use a "scaled age effect" to calculate the age-adjusted emission rates for each age. The scaled age
effect, sa, is calculated using the age of the vehicle in comparison to the warranty and useful life
requirements, as shown in Table B-2. When the vehicle age is between the end of the warranty and the
useful life, sa is interpolated between 0 and 1 as summarized in Table B-3 below and illustrated in Figure B-
1 above.

Table B-3 Calculation of sa

Where:



age < end of warranty age

0

end of warranty age < age
< useful life

(age — end of warranty age)
(Useful life age — end of warranty age)

age > useful life

1

Since MOVES deals with age groups and not individual ages (Table 1-6), the increase in emissions by age
must be calculated by age group. For simplicity, we modeled an even age distribution within each age
group (e.g., ages 0, 1, 2, and 3 are equally represented in the 0-3 age group). We then calculated average
scaled age effects for each age group. This is important since, for example, HHD trucks reach their useful
life at four years, which means they will increase emissions through the 0-3 age group. As a result, the 0-3

259


-------
age group emission rate will be higher than the zero-mile emission rate for HHD trucks. Table B-4 andTable
B-5 show the average scaled age effect by age group for pre-2027 and 2027+ MY vehicles, respectively. In
these tables, a value of 0 indicates no deterioration, (i.e., the zero-mile emissions level (ZML)), and a value
of 1 indicates a fully deteriorated engine, or maximum emissions level, at or beyond the useful life (UL).

Table B-4 Average Scaled Age Effect, Sa For Pre-MY2027 Vehicles

Age Group

LHD2b3
(Tier 2 and
earlier)

LHD2b3
(Tier 3)

LHD45

MHD

HHD

Bus

0-3

0.0833

0.0625

0.125

0.083

0.25

0.0313

4-5

0.8333

0.6250

1

0.833

1

0.3125

6-7

1

1

1

1

1

0.5625

8-9

1

1

1

1

1

0.8125

10-14

1

1

1

1

1

1

15-19

1

1

1

1

1

1

20+

1

1

1

1

1

1

Table B-5 Average Scaled Age Effect, Sa For MY2027+ Vehicles

Age Group

LHD2b3
(Tier 2 and
earlier)

LHD2b3
(Tier 3)

LHD45

MHD

HHD

Bus

0-3





0

0

0.00

0

4-5





0

0

0.25

0

6-7

Not subject to
HD2027 rule.

Same as Pre-

0

0

1.00

0

8-9

MY2027

0.25

0.75

1.00

0

10-14

Vehicles

1

1

1.00

0.75

15-19





1

1

1.00

1

20+





1

1

1.00

1

Then, for each pollutant and age, we multiplied the zero-mile emission rate by one plus the product of the
average scaled age effect and the T&M adjustment factor.

ERp,r,a,o — ERzero miie,p,r,o x (1 + sa x ^t&m)	Equation 8-3

Where:

•	ERp r 0 a = the heavy-duty diesel emission rate for each pollutant p, regulatory class r, age
a, operating mode, o,

•	ERzero mjig= the zero-mile heavy-duty diesel emission rate for each pollutant p, regulatory
class r, operating mode, o

•	§;= average scaled age effect at age group, a

•	fT&M= the tampering and mal-maintenance adjustment factor (Equation 8-1)

260


-------
Sections B.2 through B.9 discuss the data sources and assumptions used to determine the T&M failure
frequencies and T&M emission effects used to derive the T&M adjustment factor in Equation 8-1 for each
pollutant and model year range of vehicle.

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 Memo186 (2006). The basic
EMFAC occurrence rates for tampering and mal-maintenance were developed from Radian
and EFEE reports and CARB engineering judgment.

•	Radian Study (1988). The report estimated the malfunction rates based on survey and
observation. The data may be questionable for current heavy-duty trucks due to
advancements such as electronic controls, injection systems, and exhaust aftertreatment.

•	EFEE report (1998) on PM emission deterioration rates for in-use vehicles. Their work
included heavy-duty diesel vehicle chassis dynamometer testing at Southwest Research
Institute.

•	EMFAC2000 (2000) Tampering and Mal-maintenance Rates

•	EMA's comments on ARB's Tampering, Malfunction, and Mal-maintenance Assumptions
for EMFAC 2007

•	University of California -Riverside (UCR) "Incidence of Malfunctions and Tampering in
Heavy-Duty Vehicles"

•	Air Improvement Resources, Inc.'s Comments on Heavy-Duty Tampering and Mal-
maintenance Symposium

B.3 T&M Failure Modes

EPA generally adopted the T&M failure modes developed by CARB, with a few exceptions. The high fuel
pressure category was removed. We added a failure mode for mis-fueling to represent the use of nonroad
diesel in cases when ULSD onroad diesel is required. We combined the injector failure modes into a single
group. We reorganized the EGR failure modes into "Stuck Open" and "Disabled/Low Flow." We included
the PM regeneration system, including the igniter, injector, and combustion air system in the PM filter leak
failure mode.

For model years 1994-2007, the EPA developed failure mode frequencies for model year groups that apply
to all heavy-duty diesel vehicles, including earlier model years. For model year 2007-2012, we developed
separate failure mode frequencies for heavy-duty diesel vehicles that are equipped with Lean NOx Traps
(LNT) and Selective Catalyst Reduction (SCR) systems, respectively. Beyond model year 2012, we assume all
heavy-duty vehicles are using SCR systems. Better understanding tampering and mal-maintenance effects
in contemporary vehicles is an area where additional research would be beneficial.

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.

261


-------
• 2027 and later represent the heavy-duty vehicles subject to the HD2027 standards
including further technology improvements of heavy-duty engines and after-treatment
system.187

B.5 T&M Failure Frequency Rates and Differences

EPA adopted the CARB EMFAC2007 occurrence rates, except as noted below.

Clogged Air Filter: EPA reduced the frequency rate from EMFAC's 15 percent to 8 percent. EPA reduced
this value based on the UCR results, the Radian study, and EMA's comments that air filters are a
maintenance item. Many trucks contain indicators to notify the driver of dirty air filters and the drivers
have incentive to replace the filters for other performance reasons.

Other Air Problems: EPA reduced the frequency rate from 8 percent to 6 percent based on the UCR
results.

Electronics Failed: EPA continued to use the 3 percent frequency rate for all model years beyond 2010. We
projected that the engine hardware would evolve through 2010, rather than be replaced with completely
new engine systems that would justify a higher rate of failure. For 2010 and later vehicles, the occurrence
of T&M on electronics associated with SCR and DPF aftertreatment systems is counted with the
aftertreatment specific failure modes (including "NOx aftertreatment malfunction" and "PM Filter
Disable"), rather than in the "Electronics Failed" mode.

EGR Stuck Open: EPA believes the failure frequency of this item is rare and therefore set the level at 0.2
percent. This failure will lead to drivability issues that will be noticeable to the driver and serve as an
incentive to repair.

EGR Disabled/Low Flow: EPA estimates the ERG failure rate at 10 percent. All but one major engine
manufacturer had EGR previous to the 2007 model year and all have it after 2007, so a large increase in
rates seem unwarranted. However, the Illinois EPA stated that "EGR flow insufficient" is the top OBD issue
found in their LDV l/M program188 so it cannot be ignored.

NOx Aftertreatment malfunction: EPA developed a NOx aftertreatment malfunction rate that is dependent
on the type of system used. We assumed that HHDD will use primarily SCR systems and LHDD will primarily
use LNT systems. We estimated the failure rates of the various components within each system to develop
a composite malfunction rate (Table B-6).

The individual failure rates were developed considering the experience in agriculture and stationary
industries of NOx aftertreatment systems and similar component applications. Details are included in the
chart below. We assumed that tank heaters had a five percent failure rate but were only required in one
third of the country during one fifth of the year. The injector failure rate is lower than fuel injectors, even
though they have similar technology, because there is only one required in 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.

262


-------
Table B-6 N0X 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.

263


-------
Misfuel: EPA estimated that operators will use the wrong type of fuel, such as agricultural diesel fuel with
higher sulfur levels, approximately 0.1 percent of the time.

In the future, we hope to collect updated real-world failure frequencies for newer technologies.

B.6 Tampering & Mal-maintenance Failure Frequency Rate Summary

Table B-7 T&M Failure Frequency Rate by Model Year Group

Model Year

1994-1997

1998-2002

2003-2006

2007-2009

2007-2012

2010+

NO* Afterteatment Technology:

None

None

None

None

LNT

SCR

Timing Advanced

5%

2%

2.0%

2.0%

2.0%

2.0%

Timing Retarded

3%

2%

2.0%

2.0%

2.0%

2.0%

Injector Problem (all)

28%

28%

13.0%

13.0%

13.0%

13.0%

Puff Limiter Mis-set

4%

0%

0.0%

0.0%

0.0%

0.0%

Puff Limited Disabled

4%

0%

0.0%

0.0%

0.0%

0.0%

Max Fuel High

3%

0%

0.0%

0.0%

0.0%

0.0%

Clogged Air Filter - EPA

8%

8%

8.0%

8.0%

8.0%

8.0%

Wrong/Worn Turbo

5%

5%

5.0%

5.0%

5.0%

5.0%

Intercooler Clogged

5%

5%

5.0%

5.0%

5.0%

5.0%

Other Air Problem - EPA

6%

6%

6.0%

6.0%

6.0%

6.0%

Engine Mechanical Failure

2%

2%

2.0%

2.0%

2.0%

2.0%

Excessive Oil Consumption

5%

3%

3.0%

3.0%

3.0%

3.0%

Electronics Failed - EPA

3%

3%

3.0%

3.0%

3.0%

3.0%

Electronics Tampered

10%

15%

5.0%

5.0%

5.0%

5.0%

EGR Stuck Open

0%

0%

0.2%

0.2%

0.2%

0.2%

EGR Disabled/Low-Flow - EPA

0%

0%

10.0%

10.0%

10.0%

10.0%

NOx Aftertreatment Sensor

0%

0%

0.0%

0.0%

10.0%

10.0%

Replacement NOx Aftertreatment
Sensor

0%

0%

0.0%

0.0%

1.0%

1.0%

NOx Aftertreatment Malfunction -
EPA

0%

0%

0.0%

0.0%

16.0%

13.0%

PM Filter Leak

0%

0%

0.0%

5.0%

5.0%

5.0%

PM Filter Disabled

0%

0%

0.0%

2.0%

2.0%

2.0%

Oxidation Catalyst
Malfunction/Remove - EPA

0%

0%

0.0%

5.0%

5.0%

5.0%

Mis-fuel - EPA

0.1%

0.1%

0.1%

0.1%

0.1%

0.1%

B.7 NOx T&M Emission Effects
B.7.1 Model Years 1994 through 2026

For model years 1994 through 2026, EPA developed the emission effect from each tampering and mal-
maintenance incident from CARB's EMFAC, Radian's dynamometer testing with and without the
malfunction present, Engine, Fuel, and Emissions Engineering Inc. (EFEE) results, and EPA staff testing
experience.

264


-------
EPA estimated that the lean N0X traps (LNT) in LHD are 80 percent efficient and the selective catalyst
reduction (SCR) systems in HHD are 90 percent efficient at reducing NOx.

EPA developed the NOx emission factors of the NOx sensors based on SCR systems' ability to run in open-
loop mode and still achieve NOx reductions. The Manufacturers of Emission Controls Association (MECA)
has stated that a 75-90 percent NOx reduction should occur with open loop control and >95 percent
reduction should occur with closed loop control.189 Visteon reports a 60-80 percent NOx reduction with
open loop control.190

In testing, the failure of the NOx aftertreatment system had a different impact on the NOx emissions
depending on the type of aftertreatment. The HHD vehicles with SCR systems experienced a 1000 percent
increase in NOx during a complete failure, therefore we estimated a 500 percent increase as a midpoint
between normal operation and a complete failure. The LHD vehicles with LNT systems experienced a 500
percent increase in NOx during a complete failure. We estimated a 300 percent increase as a value
between a complete failure and normal system operation. The values with 0 percent effect in shaded cells
represent areas which have no occurrence rate.

As discussed in Section 2.1.1.4.6, we estimate that 25 percent of LHD MY 2007-2009 vehicles were
equipped with LNT aftertreatment systems. For LHD2b3 MY 2010-2012 vehicles, we modeled that 25
percent of vehicles had LNT aftertreatment systems, and 75 percent had SCR systems. For LHD2b3 MY
2013+, we assume that all are equipped with SCR aftertreatment systems. For LHD45, MHD, HHD, and
Urban buses, we modeled the model year 2010 and later T&M effects assuming all engines are equipped
with SCR aftertreatment systems. We recognize this is a simplification as manufacturers produced non-SCR
equipped engines in the initial implementation years of the 2010 standard due to average, banking, and
trading, and the EPA allowance of nonconformance penalty (NCP) engines in 2012.90

B.7.2 Model Years 2027 and Later

For MY2027+ vehicles, we further adjusted the MY2010 NOx T&M emission effect to reflect the HD2027
standards. As NOx emissions become more tightly controlled with the application of advanced technologies
to meet the standards, we anticipate the NOx T&M emission effects will increase (i.e., there will be a
relatively larger impact of T&M because the emission control system is reducing a greater percentage of
the NOx produced by the engine).

To estimate the NOx T&M emission effects for the HD2027 standards, we first calculated the average zero-
mile NOx emission rate ERzero miie^ox prior to the standard based on the weighted average of the
different operating modes o, and regulatory class r, using Equation 8-4.

prp		 ^lr,o(j^Rzero mile,NOx,r,o * ^r,o )	Ex]U(lti()Tl 8-4

^^zero mile,NOX	y I

2jt*,o ^r,o

Where:

•	ERzeromiie NOx = the average heavy-duty diesel NOx emission rate

•	ERzero mile,nox,r,o = the zero-mile heavy-duty diesel NOx emission rate for regulatory class,
r, and operating mode, o

•	tr o= operation time by regulatory class and operating mode estimated by MOVES3.

Next, we estimated the NOx emission rate of MY 2010 vehicles with a tampering and mal-maintenance
failure i, using Equation 8-5, which was derived from Equation 8-2 using the fleet average emission rate
from Equation 8-4 assuming the T&M frequency is 100 percent.

265


-------
ERt&m uvox = ERzero miie..vo.Y X (l + T&M emission effecti w0x)	Equation 8-5

We then derived Equation 8-6, assuming that a NOx aftertreatment equipment failure i, in the control
scenario, would cause the average of the MY 2027+ failed emission rates, ERT&M i:Nox>to '3e t'ie same as a
NOx aftertreatment failure in the baseline MY 2010 case, Baseline ERT&M ; NOx

MY2010 ERtk.m \ Nox ~ MY2027 ERT&Myvox	Equation 8-6

MY2010 ERzeromiie WOx x (l + MY2010 T&M emission effect; WOx)
= MY2027 ERzeromiie WOx
x (l + MY2027 T&M emission effect; WOx)

By rearranging Equation 8-6, we derived Equation 8-7 to estimate the control scenario NOx T&M emissions
effects.

MY2027 T&M emission effect; WOx

MY2010 ERzero miie,WOx X (l + MY2010 T&M emission effectlw0x)] ^	Equation 8-7

MY2027 ERzeromiie WOx

The MY2027 T&M NOx emission effects for the NOx aftertreatment failures are much larger than the
MY2010 values, because the zero-mile NOx emission rate for MY2027 vehicles are lower than the MY2010
zero-mile NOx emission rates.

266


-------
Table B-8 N0X T&M Emission Effect by Model Year Group

Model Year

1994-
1997

1998-
2002

2003-
2006

2007-
2009

2007-
2012

2010-
2026

2027-
2028

2029-
2060

NOx Afterteatment
Technology:

None

None

None

None

LNT

SCR

SCR

SCR

Federal NOx Emission
Standard (g/bhp-hr)

5.0

5.0

4.0

2.0

0.2

0.2

0.05

0.05

Timing Advanced

60%

60%

60.0%

60.0%

12.0%

6.0%

6.0%

6.0%

Timing Retarded

-20%

-20%

-20.0%

-20.0%

-20.0%

-20.0%

-20.0%

-20.0%

Injector Problem (all)

-5%

-1%

-1.0%

-1.0%

-1.0%

-1.0%

-1.0%

-1.0%

Puff Limiter Mis-set

0%

0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Puff Limited Disabled

0%

0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Max Fuel High

10%

0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Clogged Air Filter - EPA

0%

0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Wrong/Worn Turbo

0%

0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Intercooler Clogged

25%

25%

25.0%

25.0%

5.0%

3.0%

3.0%

3.0%

Other Air Problem - EPA

0%

0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Engine Mechanical Failure

-10%

-10%

-10.0%

-10.0%

-10.0%

-10.0%

-10.0%

-10.0%

Excessive Oil Consumption

0%

0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Electronics Failed - EPA

0%

0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Electronics Tampered

80%

80%

80.0%

80.0%

16.0%

8.0%

8.0%

8.0%

EGR Stuck Open

0%

0%

-20.0%

-20.0%

-20.0%

-20.0%

-20.0%

-20.0%

EGR Disabled/Low-Flow -
EPA

0%

0%

30.0%

50.0%

10.0%

5.0%

5.0%

5.0%

NOx Aftertreatment SensorA

0%

0%

0.0%

0.0%

200.0%

200.0%

1294%;
1271%;
1620%;
1713%B

1301%;
1277%;
1643%;
1741%B

Replacement NOx
Aftertreatment SensorA

0%

0%

0.0%

0.0%

200.0%

200.0%

1294%;
1271%;
1620%;
1713%B

1301%;
1277%;
1643%;
1741%B

NOx Aftertreatment
Malfunction - EPAA

0%

0%

0.0%

0.0%

300.0%

500.0%

2688%;
2641%;
3339%;
3527%B

2703%;
2655%;
3386%;
3582%B

PM Filter Leak

0%

0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

PM Filter Disabled

0%

0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Oxidation Catalyst
Malfunction/Remove - EPA

0%

0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Mis-fuel - EPA

0%

0%

0%

0%

0%

0%

0%

0%

A NOx aftertreatment failure modes

BThe values in the cells are for LHD45, MHD, HHD, Bus categories, respectively

267


-------
B.7.3 N0X Adjustment Factor Calculations

Combining the NOx emission effects with the frequency rates results in the initial T&M adjustment factors
shown in the table below. This methodology estimated a small (9-14%) T&M NOx adjustment factor for
2009 and earlier models due to NOx effects of the following failure modes: electronics tampered, timing
advances, intercooler clogged, and ERG disabled/Low Flow. However, MOVES does not use the estimated
NOx T&M emission effects initially estimated for 2009 and earlier model years, and assumes no NOx
increase (with the exception of the LNT effect for LHD explained below). This is indicated in the 3rd column
of Table B-9 labeled "(Removed 2009 and earlier)." Instead, MOVES assumes NOx increases only in the
vehicles with advanced NOx exhaust aftertreatment technologies for a few reasons:

•	The Consent Decree Testing conducted by West Virginia University did not show an
increase in NOx emissions with odometer (and consequently, age) during or following the
regulatory useful life.191 Since the trucks in this program were collected from in-use fleets,
we do not believe that these trucks were necessarily biased toward cleaner engines.

•	Heavy-duty diesel manufacturers often certify zero or low deterioration factors for these
engines.

•	Starting with MY 2010 (2007 for vehicles with LNT), we expect T&M effects to become
much more significant, because a failure in the NOx aftertreatment system will
substantially increase emissions. We decided to initiate modeling of the NOx T&M
adjustment factor with the implementation of the 2010 standards.

The assumption of no T&M NOx increases for pre-2010 heavy-duty vehicles (except LHD with LNT), extends
to glider vehicles for all model years (regClassID 49).

268


-------
Table B-9 Tampering & Mai-Maintenance N0X Adjustment Factors (fr&M,Nox) (Percent) for Heavy-Duty Diesel Vehicles

without Onboard Diagnostics (OBD)

Model years

NOx

Aftertreatment
Technology

fl&M,NOx,nonOBD (Initial)

fr&M,NOx,nonOBD

(Removed 2009 and
earlier for non-LNT
engines)

1994-1997

None

10

0

1998-2002

None

14

0

2003-2006

None

8.7

0

2007-2009

None

10.7

0

2007-2012

LNT

71.5

71.5

2010-2026

SCR

87.4

87.4

2027-2028

SCR

492.2	(LHD45); 483.5
(MHD); 612.7 (HHD);

647.3	(Bus)

492.2	(LHD45); 483.5
(MHD); 612.7 (HHD);

647.3	(Bus)

2029+

SCR

494.9 (LHD45); 486.0
(MHD); 621.3 (HHD);
657.6 (Bus)

494.9 (LHD45); 486.0
(MHD); 621.3 (HHD);
657.6 (Bus)

The T&M N0X emission rates for LHD2b3 vehicles equipped with LNT aftertreatment in 2007-2009 are
calculated by first adjusting Equation 2-11 to account for T&M of LNT aftertreatment, as shown in Equation
8-8. The derivation of Equation 2-11 including the definition of normal operation frequency and DPF
regeneration frequency are discussed in Section 2.1.1.4.6

2007 — 2009 LNT NOx emissions (T&M)
Baseline Emissions

LNT normal emissions

= (normal op. frequency) x (¦

V Baseline emissions

X (1 + 2007 - 2009 LHD LNT T&.M effect)

(Baseline emissions\

+ (DPF reg. frequency) x 	-	

V Baseline emission J

= (0.90) x (0.10) x (1.715) + (0.10) x (1) x (1) = 0.2544
Where Baseline Emissions = MOVES2010 MY 2003-2006 NOx emission rates for LHD2b3

Equation 8-8

Because MOVES does not model LNT vehicles separately, we then calculated an average ratio for all 2007-
2009 LHD2b3 NOx rates (both non-LNT and LNT with T&M) over the baseline 2003-2006 NOx rates by
adjusting Equation 2-12 to account for the T&M effects of LNT, as shown in Equation 8-9.

269


-------
2007 — 2009 LHD2b3 NOx emissions (T&.M)

= (LNT market share)

Baseline emissions
2007 — 2009 LNT NOx emissions (T8lM)\

Baseline emissions	)	Equation 8-9

+ (non

/2007 — 2009 emission standards\

— LNT market share) 	-	

V	Baseline emissions	)

= (0.25) X (0.2544) + (0.75) X (0.5) = 0.4386
Where Baseline Emissions = MOVES2010 MY 2003-2006 NOx emission rates for LHD2b3

Then, the T&M effect for 2007-2009 LHD2b3 is calculated in Equation 8-10 by dividing Equation 8-9 by
Equation 2-12 and subtracting 1.

^	2007 — 2009 LHD2b3 NOx emissions (T&.M) ^

ft&m,NOx,LHD2b3,2007—2009 - 2007 - 2009 LHD2b3 NOx emissions (zero mile) ~ 1	_ .

v	y	Equation

_ /2007 - 2009 LHD2b3 NOx (T&M)\ //2007 - 2009 LHD2b3 NOx (zero mile)\	8-10

\	Baseline emissions	Jl V	Baseline emissions	J

= 0.4386/0.4225 — 1 = 1.038 — 1 = 3.8% increase due to T&M
Where Baseline Emissions = MOVES2010 MY 2003-2006 NOx emission rates for LHD2b3

For 2007-2009, LHD45 uses the same emission rates and T&M factors as LHD2b3. As noted earlier, we
assume no T&M NOx effects for pre-2010 MY vehicles in the other heavy-duty regulatory classes.

The T&M adjustment factors (fr&M,NOx) for model year 2010 and later model years incorporate the impact of
onboard diagnostic (OBD) emission effect assumptions discussed in Section B.10, and calculated with
Equation 8-15. As explained in that section, for LHD2b3 vehicles, we assume 100% OBD penetration
starting in 2010. This reduces the T&M adjustment factor by 0.33 for these years.

For 2010-2012, LHD2b3, we assume that both LNT and SCR equipped vehicles will provide the same level
of control with a 90 percent reduction from 2003-2006 levels (ignoring the PM regeneration NOx benefit
for LNT aftertreatment considered for the 2007-2009 rates for simplicity). To calculate the T&M NOx
effects for 2010-2012 (Jt&m,NOx,LHD2b3,2010-2012)' we weighted the LNT-specific and SCR-specific T&M
effects (from Table B-9) according to the market shares, and applied the 33% percent reduction for OBD as
shown in Equation 8-11:

/t&M, LHD2b3,2010-2012

= (LNT market share) x (fT&.M,NOx,LNT,nonOBD) x C/obd)
+ (SCR market share) x (_fT&LM,NOx,scR,nonOBD) x (/obd)

= (25%) x (71.5%) x (67%) + (75%) x (87.4%) x (67%) = 55.9%

Equation 3-11

270


-------
For LHD45 and heavier regulatory classes, we assume a 33 percent OBD penetration in model year 2010-
2012 as shown in Equation 8-12.

fr&M,NOx,LHD45,2010-2012

= {fT8iM,NOx,SCR,nonOBD) X (/oBd) X (PoBd)
+ (fT&.M,NOx,SCR,nonOBD) x (1 — P0Bd)

= (87.4%) X (67%) X (33%) + (87.4%) X (67%) = 77.9%

Equation 8-12

For 2013 and later model years, the T&M adjustment factors are calculated for heavy-duty vehicles
assuming that all (except for gliders) are using SCR technology and 100 percent OBD.

fr8iM,NOx,SCR,2013-2026 — (fT&.M,NOx,SCR,nonOBD) X (/oBd) X (PoBd)

= (87.4%) X (67%) X 100% = 58.6%

Equation 8-13

fT8iM,NOx,SCR,2027-2028,HHD ~ (fT8iM,NOx,SCR,nonOBD) X (/oBd) X (PoBd)

= (612.7%) X (67%) X 100% = 410.5%

Equation 8-14

The NOx Tampering & Mal-maintenance adjustment factors by regulatory class and model year groups are
summarized in Table B-10.

Table B-10 NOx T&M Adjustment Factors (fmM,Nox) by MOVES Regulatory Classes and Model Year Groups

Model Year
Group

LH2b3

(RegClass 41)

LHD45

(RegClassID 42)

MHD, HHD, Bus
(RegClassID 46,47,48)

Gliders

(RegClassID 49)

2007-2009

3.81%

3.81%

0%

0%

2010-2012

55.9%

77.9%

77.9%

0%

2013-2027

58.6%

58.6%

58.6%

0%

2027-2028

Same as 2013-
2027

329.7%

324.0% (MHD); 410.5%
(HHD); 433.7% (Bus)

0%

2029+

Same as 2013-
2027

331.6%

325.6% (MHD); 416.3%
(HHD); 440.6% (Bus)

0%

B.8 PM T&M Emission Effects

EPA developed the PM emission effects for each tampering and mal-maintenance incident from CARB's
EMFAC, Radian's dynamometer testing with and without the malfunction present, EFEE results, and
internal testing experience.

271


-------
EPA estimates that the diesel PM filter has 95 percent effectiveness. Many of the tampering and mal-
maintenance items that impact PM also have a fuel efficiency and drivability impact. Therefore, operators
will have an incentive to fix these issues.

EPA estimated that excessive oil consumption will have the same level of impact on PM as engine
mechanical failure. The failure of the oxidation catalyst is expected to cause a PM increase of 30 percent;
however, this value is reduced by 95 percent due to the PM filter effectiveness. We also considered a DOC
failure will cause a secondary failure of PM filter regeneration. We accounted for this PM increase within
the PM filter disabled and leak categories.

The values with 0 percent effect in shaded cells represent areas which have no occurrence rate.

In MOVES2014, we increased the PM emission effect for PM Filter Leaks and PM Filter Tampering for the
2007-2009 and 2010+ model year groups. The PM filter leak was increased from 600 percent to 935
percent and the PM Filter Disabled emission effect was increased from 1000 percent to 2670 percent.
These in Table B-9 effects along with the OBD effects discussed in Section B.10 results in a fleet average
PM2.5 Tampering & Mal-maintenance effect of 100 percent in 2007-2009 and 89 percent in 2010-2012
(Table 2-25). Note that we apply the same T&M Emission Effect for both EC and nonECPM.

272


-------
Table B-ll PM2.5 T&M Emission Effect by Model Year Group



1994-
1997

1998-
2002

2003-
2006

2007-
2009

2010+

Federal Emission Standard

0.1

0.1

0.1

0.01

0.01













Timing Advanced

-10%

-10%

-10%

0%

0%

Timing Retarded

25%

25%

25%

1%

1%

Injector Problem

100%

100%

100%

5%

5%

Puff Limiter Mis-set

20%

0%

0%

0%

0%

Puff Limiter Dsabled

50%

0%

0%

0%

0%

Max Fuel High

20%

0%

0%

0%

0%

Clogged Air Filter

50%

50%

30%

2%

2%

Wrong/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%

100%

5%

5%

EGR Disabled/Low Flow

0%

0%

-30%

-30%

-30%

NOx Aftertreatment Sensor

0%

0%

0%

0%

0%

Replacement NOx
Aftertreatment Sensor

0%

0%

0%

0%

0%

NOx Aftertreatment Malfunction

0%

0%

0%

0%

0%

PM Filter Leak

0%

0%

0%

935%

935%

PM Filter Disabled

0%

0%

0%

2670%

2670%

Oxidation Catalyst
Malfunction/Remove

0%

0%

0%

0%

0%

Mis-fuel - EPA

30%

30%

30%

100%

100%

273


-------
B.9 THC and CO T&M Emission Effects

EPA estimated oxidation catalysts are 80 percent effective at reducing hydrocarbons. All manufacturers
will utilize oxidation catalysts in 2007, but only a negligible number were installed prior to the PM
regulation reduction in 2007. We assumed that with Tampering and Mal-maintenance, the THC zero level
emissions will increase by 50 percent. This still represents a 70 percent reduction in THC emissions
between zero-mile 2006 emissions and fully deteriorated 2007 vehicles.

We reduced CARB's THC emission effect for timing advancement because earlier timing should reduce
THC, not increase them. The effect of injector problems was reduced to 1000 percent based on EPA's
engineering staff experience. We increased the THC emission effect of high fuel pressure (labeled as Max
Fuel High) to 10 percent in 1994-1997 years because the higher pressure will lead to extra fuel in early
model years and therefore increased THC. Lastly, we used the THC emission effect of advanced timing for
the electronics tampering (0 percent) for all model years. The values with 0 percent effect in shaded cells
represent areas which have no occurrence rate.

Table B-12 THC T&M Emission Effect by Model Year Group

Model Year

1994-
1997

1998-
2002

2003-
2006

2007-
2009

2010+

Federal HC Emission Standard (g/bhp-hr)

1.3

1.3

1.3

0.2

0.14

Timing Advanced

0%

0%

0%

0%

0.0%

Timing Retarded

50%

50%

50%

50%

10.0%

Injector Problem (all)

1000%

1000%

1000%

1000%

200.0

%

Puff Limiter Mis-set

0%

0%

0%

0%

0.0%

Puff Limited Disabled

0%

0%

0%

0%

0.0%

Max Fuel High

10%

0%

0%

0%

0.0%

Clogged Air Filter - EPA

0%

0%

0%

0%

0.0%

Wrong/Worn Turbo

0%

0%

0%

0%

0.0%

Intercooler Clogged

0%

0%

0%

0%

0.0%

Other Air Problem - EPA

0%

0%

0%

0%

0.0%

Engine Mechanical Failure

500%

500%

500%

500%

100.0

%

Excessive Oil Consumption

300%

300%

300%

300%

60.0%

Electronics Failed - EPA

50%

30%

50%

50%

10.0%

274


-------
Electronics Tampered

0%

0%

0%

0%

0.0%

EGR Stuck Open

0%

0%

100%

100%

20.0%

EGR Disabled/Low-Flow - EPA

0%

0%

0%

0%

0.0%

NOx Aftertreatment Sensor

0%

0%

0%

0%

0.0%

Replacement NOx Aftertreatment Sensor

0%

0%

0%

0%

0.0%

NOx Aftertreatment Malfunction - EPA

0%

0%

0%

0%

0.0%

PM Filter Leak

0%

0%

0%

0%

0.0%

PM Filter Disabled

0%

0%

0%

0%

0.0%

Oxidation Catalyst Malfunction/Remove -
EPA

0%

0%

0%

50%

50.0%

Mis-fuel - EPA

0%

0%

0%

0%

0%

A separate tampering analysis was not performed for CO; rather, the THC effects were assumed to apply
for CO.

Combining all of the emissions effects and failure frequencies discussed in this section, and the OBD effects
discussed in the next section, we summarized the aggregate emissions impacts over the useful life of the
fleet in the main body of the document in Table 2-28(THC and CO).

B.10 HD OBD impacts

With the finalization of the heavy-duty onboard diagnostics (HD OBD) rule, we made adjustments to 2010
and later model years to reflect the rule's implementation.

Specifically, we reduced the emissions increases for all pollutants due to an OBD tampering and mal-
maintenance factor, f0BD, which reduced the T&M adjustment factors by 33 percent. Data on the impact
of OBD were not available for heavy-duty trucks, and this number is probably a conservative estimate. This
is in addition to the substantial PM2.5 and NOx reductions for 2010 and later vehicles due to the
implementation of other standards. We assumed, since the rule phased-in OBD implementation, that 33
percent of all LHD45, MHD, HHD and Urban Bus engines would have OBD in the 2010, 2011, and 2012
model years, and 100 percent would have OBD by 2013 model year and later. For LHD2b3 vehicles, we
assumed they would have 100% OBD penetration starting in 2010. Equation 8-15 describes the calculation
of the percent increase in emission rate through useful life (T&M adjustment factors (fr&M)); where Pobd
represents the fraction of the fleet equipped with OBD (Table B-ll).

^T&M,p = fT&M,nonOBD,p X (1 — Pobd) + fT&M,nonOBD,p X ^OBD x (Pobd)	Equation 8-15

Where:

•	^T&M,p= the tampering and mal-maintenance adjustment factor for pollutant, p, that
accounts for the phase-in of OBD

•	fT&M,nonOBD,p= the tampering and mal-maintenance adjustment factor for pollutant, p, for
engines without OBD; calculated in Sections B.7 through B.9

•	Pobd= penetration of the fleet equipped with OBD, as shown in Table B-ll.

275


-------
• Wd= the effect of OBD on the T&M adjustment factor = 0.67= 33% reduction

Table B-13 Onboard Diagnostic (OBD) Assumed Phase-in (pobd) by Model Year and Regulatory Class

Model years

Regulatory Class

Pobd(%)

Pre-2010

LHD2b3, LHD45, MHD, HHD, Urban Bus

0

2010-2012

LHD2b3

100

2010-2012

LHD45, MHD, HHD, Urban Bus

33

2013+

LHD2b3, LHD45, MHD, HHD, Urban Bus

100

C. Tampering and Mai-maintenance for MY 2007 and Later Diesel Extended Idle
As discussed in Section 2.3.2.3 we assume the failure of diesel particulate filters (DPF) is the primary cause
of T&M effects on emission deterioration in 2007+ extended idle emissions. We made assumptions about
the failure rates of DPFs from in-use trucks based on consultation with several references and staff at the
California Air Resources Board (CARB) as summarized in Table C-l. We adopted the assumption shared by
CARB staff that 10 percent of 2007-2009 DPFs fail in the real-world, and 5 percent of 2010+ DPFs fail in the
real-world.

Table C-l References Used to Support In-Use DPF Failure Rate Assumption for Extended Idling Emissions

Study

Relevant Information

US EPA (2015)109

7% of 2007+ trucks in MOVES are assumed to either have a PM filter leak or have
the PM filter disabled. Current assumption for running exhaust emissions in
MOVES3.

Preble et al.
(2015)192

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)193

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)194

35% to 4% of trucks submitted warranty claims related to the PM filter between
2007 and 2011

CARB (2015)194

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
~l/2 equipped with DPFs

CARB

correspondence
(2016)

~10% of 2007-2009 DPFs and ~5% of 2010+ DPFs to fail in real-world, based on
their observations from warranty claims, snap-idle acceleration opacity tests,
and their review of the Bishop et al. (2014)193 and Preble et al. (2015)192 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

276


-------
'Deteriorated' emission rate that represents a mix of failed and properly operating systems by assigning
the 'failed' DPF emission rates a weight of 10 percent in the 2007-2009 model year group, and 5 percent
weight in the 2010-2012, and 2013+ model year groups, as shown in Table C-2. The 'Deteriorated' emission
rate represents the presumed emission rate of fully-aged heavy-duty diesel trucks. Unlike the start and
running MOVES emission rates, extended idle emission rates in MOVES are not distinguished by age. Thus,
these rates are constant with respect to age.

Table C-2 Baseline and deteriorated THC and PM2,s emission rates to account for failure of diesel particulate filters

(DPFs) by model year groups

Engine
Model
Year

Baseline

Deteriorated

THC
(g/hr)

PM25
(g/hr)

EC

(g/hr)

nonEC
(g/hr)

Failure
rate

THC
(g/hr)

PM25
(g/hr)

EC

(g/hr)

nonEC
(g/hr)

2005-
2006

8.49

0.251

0.065

0.187

-

8.49

0.251

0.065

0.187

2007-
2009

8.49

0.075

0.007

0.067

10%

8.49

0.092

0.013

0.079

2010-
2012

2.53

0.026

0.004

0.022

5%

2.83

0.037

0.007

0.030

2013+

1.38

0.012

0.002

0.010

5%

1.74

0.024

0.005

0.019

We assume that trucks that are under warranty would have substantially fewer aftertreatment failures
than older trucks. Because extended idle rates are modelled as constant with age, to estimate the fleet-
average emission rates used in MOVES, we used the 'Baseline' emission rates to represent trucks that are
within the specified 435,000 miles useful-life of the engine in the US EPA regulations. We use the
deteriorated emission rate to represent the years between the regulated "useful life" and the 1,530,000
miles that MOVES models as the mean life-time miles for a long-haul combination truck. Using the
'deterioration fraction' [1-0.435/1.53 = 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 C-3. As shown,
the MOVES EC/PM emission rates for MY 2007+ trucks are slightly higher than the 'Baseline' EC/PM
fractions in Table B-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.

277


-------
Table C-3 Emission Rates Calculated from Weighting the 'Baseline' and 'Deteriorated' Emission Rates from Table C-

2Using the Deteriorated Fraction

Engine
Model
Year

MOVES

Deteriorated
Fraction

THC
(g/hr)

PM2.5
(g/hr)

EC

(g/hr)

nonEC
(g/hr)

EC/PM

2005-2006

-

8.49

0.251

0.065

0.187

0.26

2007-2009

0.72

8.49

0.087

0.012

0.076

0.13

2010-2012

0.72

2.75

0.034

0.006

0.028

0.18

2013+

0.72

1.64

0.021

0.004

0.017

0.20

Although, 2005-2006 model year engine data was used in this analysis, the update itself is limited to the
model year 2007 and later emission rates.

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.

278


-------
Table D-l Data for Pre-2007 Extended Idle N0X Emissions

Idle NOx Rates (gram/hour) Summary

Program

Condition

#Samples

Mean NOX Emiss Rate

1991-2006 Low Speed Idle, A/C Off

McCormick, High Altitude, HDT

Low RPM, AC Off

12

85

Lim, EPA

Low RPM, No access

12

109

Irick, Clean AirTech & IdleAire



49

87

WVU - 1991-2004

Low RPM, AC Off

48

83

WVU, NCHRP



2

47

Tang, Metro NY 1984-1999



33

81

Calcagno

Low RPM, AC Off

27

120

Brodrick, UC Davis

Low RPM, AC Off

1

104

Storey

Low RPM, AC Off

4

126



Overall

188

91





1991-2006 High Speed Idle, A/C Off

Lim, EPA CCD

High RPM, No access

5

169

Calcagno

High RPM, AC Off

21

164



Overall

26

165





1991-2006 High Speed Idle, A/C On

Lim, EPA CCD

High RPM, AC On

5

212

Brodrick, UC Davis

High RPM, AC On

1

240

Calcagno

High RPM, AC On

21

223

Storey

High RPM, AC On

4

262



Overall

31

227





1975-1990 Low Speed Idle, A/C Off

WVU - 1975-1990

Low RPM, AC Off

18

48

Lim, EPA, CCD, 1985 MY

Low RPM, AC Off

1

20



Overall

19

47

1975-1990 High Speed Idle, A/C On (calculated)

Ratio of 1991-2006 "High Idle, A/C On" to "Low Idle, A/C Off"

2.5

Overall (calculated)

115.4



Calculated Extended Idle MYs 1975-1990:

69.3

Calculated Extended Idle MYs 1991-2006:

136.1

279


-------
Table D-2 Data for Pre-2007 Extended Idle HC Emissions

Idle HC Rates (gram/hour) Summary

Program

Condition

#Samples

Mean HC Emiss Rate

1991-2006 Low Speed Idle, A/C Off

McCormick, High Altitude, HDT

Low Idle, AC Off

12

10.2

WVU - 1991-2004

Low Idle, AC Off

48

9.5

Storey

Low Idle, AC Off

4

28



Overall

64

10.8









1991-2006 High Speed Idle, A/C On

Brodrick, UC Davis

High Idle, AC On

1

86

Storey

High Idle, AC On

4

48



Overall

5

55.6









1975-1990 Low Speed Idle, A/C Off

WVU - 1975-1990

Low Idle, AC Off

18

21



Overall

18

21

1975-1990 High Speed Idle, A/C On (calculated)

Ratio of 1991-2006 "High Idle, A/C On" to "Low Idle, A/C Off"

5.2

Overall (calculated)

108.2



Calculated Extended Idle MYs 1975-1990:

49.8

Calculated Extended Idle MYs 1991-2006:

25.6

280


-------
Table D-3 Data for Pre-2007 Extended Idle CO Emissions

Idle CO Rates (gram/hour) Summary

Program

Condition

#Samples

Mean CO Emiss Rate

1991-2006 Low Speed Idle, A/C Off

McCormick, High Altitude, HDT

Low Idle, AC Off

12

71

Calcagno

Low Idle, AC Off

27

37

WVU - 1991-2004

Low Idle, AC Off

48

23

Storey

Low Idle, AC Off

4

25



Overall

91

33.6









1991-2006 High Speed Idle, A/C On

Calcagno

High Idle, AC On

21

99

Brodrick, UC Davis

High Idle, AC On

1

190

Storey

High Idle, AC On

4

73



Overall

26

98.5









1975-1990 Low Speed Idle, A/C Off

WVU - 1975-1990

Low Idle, AC Off

18

31



Overall

18

31

1975-1990 High Speed Idle, A/C On (calculated)

Ratio of 1991-2006 "High Idle, A/C On" to "Low Idle, A/C Off"

2.9

Overall (calculated)

91.0



Calculated Extended Idle MYs 1975-1990:

50.8

Calculated Extended Idle MYs 1991-2006:

55.0

281


-------
Table D-4 Data for Pre-2007 Extended Idle PM Emissions

Idle PM Rates (gram/hour) Summary

Program

Condition

#Samples

Mean PM Emiss Rate

1991-2006 Low Speed Idle, A/C Off

McCormick, High Altitude, HDT

Low Idle, AC Off

12

1.8

Calcagno

Low Idle, AC Off

27

2.55

WVU - 1991-2004

Low Idle, AC Off

48

1.4

Storey

Low Idle, AC Off

4

0.3



Overall

91

1.7





1991-2006 High Speed Idle, A/C On

Calcagno

High Idle, AC On

21

4.11

Storey

High Idle, AC On

4

3.2



Overall

25

4.0



1975-1990 Low Speed Idle, A/C Off

WVU - 1975-1990

Low Idle, AC Off

18

3.8



Overall

18

3.8

1975-1990 High Speed Idle, A/C On (calculated)

Ratio of 1991-2006 "High Idle, A/C On" to "Low Idle, A/C Off"

2.3

Overall (calculated)

8.6



Calculated Extended Idle MYs 1975-1990:

5.4

Calculated Extended Idle MYs 1991-2006:

2.5

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.

282


-------
Table E-l Regression Coefficients for HD Diesel Pre-2007 PM2,s Emission Factor Model

Model-year

Speed Class (mph)

Type

Medium

Heavy Heavy-

group





Heavy-Duty

Duty

1960-87

1-25

Intercept (60)

-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 (0i)

0.02821

0.0968





Transformation









Coefficient

0.5864

0.84035





(0.5o2)





ln(PM) = p0 + ptSTP + 0.5 a2

Where:

•	60 = an intercept term for a speed class within a model year group, as shown in the table
above,

•	61 = 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).

F. Heavy-Duty Gasoline Start Emissions Analysis Figures

The figures below show heavy-duty gasoline start emissions as mentioned in Section 3.2.1.

283


-------
FTP Cold-Starts 
-------
FTP Cold-Starts (g), HD SI (HD< = 14K)
CO GEO-meon starts vs. Age by MYQ

FTP Cold-Starts (g). HD SI (HD< = 14K)
NOx GEO-meon starts vs. Age by MYQ

FTP Cold-Starts (g). HD SI (HD< = 14K)
THC GEO-meon starts vs. Age by MYG

Figure F-2 Cold-Start FTP Emissions for Heavy-Duty Gasoline Vehicles, GEOMETRIC MEANS by Model-year and Age

Groups

285


-------
CO I n_SD vs. Age by MYG

node1yeargroup

FTP Cold-Starts (g), HD SI (l-D< = 14K)
THC In„SD vs. Age by MYG

(b) THC



Vt				



10	11

nodeIycargroup

3 19901990

' 19982004

FTP Cold-Starts (g). HD SI (HD< = 14K)
NOx I n_SD vs. Age by MfG

node1ycargroup

991199? a & a 19982004

Figure F-3 Cold-start FTP Emissions for Heavy-Duty Gasoline Trucks: LOGARITHMIC STANDARD DEVIATION by Model-

year and Age Groups

286


-------
CO ARITH-meon starts vs. Age by MYG

agent id

model year group O-O-e 19601989 B-B-B 19901990 I ' I 1991199/ A-A-A 19902004

FTP Cold—Starts (g). HD SI (HD< = MIC)

THC ARITH-meon starts vs. Age by MYG

modelyeargroup OOP 19601989 BOD 19901990 I » I 19911997 ft A A 19982004

FTP Cold-Starts (g). HD St (bO< = 141^
NO* ARITH-mean starts vs. Age by MYG

agen id

mode 1 year group e-a-e 19601989 B-B-B 19901990 I ' I 19911997 A-A-A 19982004

Figure F-4 Cold-Start Emissions for Heavy-Duty Gasoline Trucks: RECALCULATED ARITHMETIC MEANS by Model-year

and Age Groups

287


-------
Table F-l Emission Standards for Heavy-Duty Spark-Ignition Onroad Engines

Regulatory Class

Model Year

Emissions Standards (g/hp-hr)





CO

THC

NMHC

NOx

NMHC + NO*

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



G. Selection of Fixed Mass Factor (fscale) values for MY 2010+ Heavy-Duty Vehicles
In M0VES3, 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/sco/e values
for LHD, MHD, and HHD weight classes. New/sco/e 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/sCa/e 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/sco;e values are used when
analyzing the real-world emissions data that leads to the base emission rates in the MOVES database. The
fscale 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/sco/e value. Glider vehicles
(regClass 49) continue to use emission rates from pre-2010 vehicles, and thus their/sco;e 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/sco/e 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/scofe
for LHD should be lower than MHD, which in turn should be lower than HHD. Our goal was to find/sco/e
values that allow the HDIUT data to cover all operating modes, thus reducing the need for gap-filling while

288


-------
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/sco/e is too big, the high power operating modes are
left vacant. On the other hand, if the/sco;e is too small, a lot of the data gets pushed to the high power
operating modes, and the high operating modes no longer capture emissions at only the most extreme and
aggressive operating conditions measured in the HDIUT data set. Both cases are sub-optimal because they
reduce the model's capability to distinguish operating modes in a meaningful way.

When analyzing the HDIUT data for various fscaie values, we estimated the number of vehicles, time, and
mass/time emission rates for each operating mode. Vehicle count and time, per operating mode, were first
cut criteria during the fscaie selection process. We used the C02 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 C02).

Tables G-l through G3 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/sco/e for each regClass, when combined with a cycle of the same number of total seconds,
should result in similar number of seconds in high power operating modes. In other words, we expect LHD,
MHD, and HHD vehicles in the HDIUT data set to have somewhat similar time distribution across power
modes. Finally, we used the C02 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/sca;e 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 C02 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 C02 emissions rates approximately 30% higher in these operating
modes. As stated earlier, our objective is to select the/SCate 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 C02 emissions rates that are robust for the high operating mode bins,/SCate
5.00 is preferable to an fSCaie of 4.00./sco/e 5.00 provides strong and expected increasing trends in C02
emission rates between operating modes 29 and 30, whereas/sco;e 6.00 does not. We believe this is due to
insufficient data in the highest operating mode bins with/sco/e 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.

289


-------
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 Tables G-l through 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/sco/e 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 /sco;e from this data set might not be as
suitable with another data set; (2) comparing closely spaced fscaie values does not necessarily provide a
clear winner across the board because there's more than one criteria (vehicle count, time, mass/time rates
for various pollutants).

In the 2019 peer-review, one of peer-reviewers asked if we have evaluated the time distribution from real-
world data using the proposed fscaie value. In Figures F-4 through G-6,we conducted 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/sco;e 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/SCate 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 representativeness of the HDIUT dataset, as well as further evaluate the MOVES default activity
assumptions, including the MOVES heavy-duty driving cycles. In the population and activity report, we
listed this as a project for consideration for future MOVES work.

290


-------
Table G-l Effect offscaie Value on Vehicle Count and Time for Light Heavy-Duty Vehicles



Number of vehicles1

Number of seconds based on a cycle with one million
seconds1,2

OpMode

fs

fs

fs

fs

fs

fs

fs

fs

fs

fs

fs

fs

2.06

4.0

5.0

6.0

9.0

17.1

2.06

4.0

5.0

6.0

9.0

17.1

0

64

64

64

64

64

64

41131

41131

41131

41131

41131

41131

1

64

64

64

64

64

64

358957

358957

358957

358957

358957

358957

11

64

64

64

64

64

64

46235

46235

46235

46235

46235

46235

12

64

64

64

64

64

64

15112

24896

29023

32878

42842

61989

13

64

64

64

64

64

64

10311

14877

16658

18290

20756

16153

14

64

64

64

64

64

43

8109

11394

12304

12431

10835

2387

15

64

64

64

64

63

18

6995

8784

8680

8092

4190

200

16

64

64

64

64

41

1

40203

20778

14065

9039

2107

1

21

64

64

64

64

64

64

45157

45157

45157

45157

45157

45157

22

59

64

64

64

64

64

7153

16366

22038

28166

47554

92823

23

64

64

64

64

64

64

9877

24704

31884

38195

49094

41467

24

64

64

64

64

64

48

12315

25290

28989

30288

26962

6425

25

64

64

64

64

64

22

13307

21400

21274

19931

11739

850

27

64

64

64

64

44

1

25844

28819

25061

18770

5671

1

28

64

64

64

44

22

2

21514

15281

8422

5020

544

5

29

64

64

42

23

0

0

16292

5808

2900

1161

0

0

30

64

43

22

13

2

1

35269

3903

1002

40

5

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

44

5

33332

86600

79892

67271

13177

7

38

64

63

62

44

22

1

46641

52178

31818

12100

978

2

39

64

62

41

23

1

1

49374

21416

6167

1940

1

2

40

64

41

23

14

1

1

113054

7852

1685

135

18

13

1	Values in bold are for final selected fSCaie- Shaded cells show instances where using an excessively high
fscale 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.

291


-------
60

70
60 -
50

40 H

8 30 -

20

io H

0

W.H n»H

0 1

0 fs = 4.00
¦ fs = 5.00
B fs = 6.00

BIH

[||

Bin Ml

a

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 offscaie Value on Coverage and Trends of operating mode Based C02for Light Heavy-Duty Vehicles

Absolute values of operating mode based emissions rates cannot be compared between series with
different fscaie values.

292


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



Number of vehicles1

Number of seconds based on a cycle with one million
seconds1,2

OpMode

fs

fs

fs

fs

fs

fs

fs

fs

fs

fs

fs

fs

4.0

5.0

6.0

7.0

9.0

17.1

4.0

5.0

6.0

7.0

9.0

17.1

0

58

58

58

58

58

58

36170

36170

36170

36170

36170

36170

1

58

58

58

58

58

58

349622

349622

349622

349622

349622

349622

11

58

58

58

58

58

58

32693

32693

32693

32693

32693

32693

12

58

58

58

58

58

58

20755

23825

26630

29346

34264

48508

13

58

58

58

58

58

58

11234

12492

13612

14397

15495

16683

14

58

58

58

58

58

58

8252

9101

9517

9898

10003

7687

15

58

58

58

58

58

30

6688

7014

7142

7000

6576

1914

16

58

58

58

58

58

16

28357

22855

18385

14645

8948

496

21

58

58

58

58

58

58

44291

44291

44291

44291

44291

44291

22

58

58

58

58

58

58

10132

14200

18827

23955

34436

78404

23

58

58

58

58

58

58

19195

25479

31609

37379

48401

49534

24

58

58

58

58

58

58

21109

27513

32401

34297

30771

19215

25

58

58

58

58

58

37

22136

24498

22647

20428

17498

5111

27

58

58

58

58

58

16

32912

28713

25621

22502

17220

1306

28

58

58

58

58

25

0

18677

15863

13842

10424

4619

0

29

58

58

46

25

10

0

12105

10075

5243

3550

625

0

30

58

39

25

14

0

0

17304

7229

3379

1034

0

0

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

30

1

56697

58308

51520

44843

11144

1

39

58

53

43

25

10

1

45255

39067

25466

8307

1628

1

40

53

39

26

14

1

0

64411

25344

7923

2440

2

0

1	Values in bold are for final selected fSCaie- Shaded cells show instances where using an excessively high
fscale 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.

293


-------
H fs = 5.00
H fs = 6.00
¦ fs = 7.00

mi

SSL

hh«

fl.

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 offscaie Value on Coverage and Trends of operating mode Based C02for Medium Heavy-Duty Vehicles

294


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



Number of vehicles

i





Number of seconds based on a cycle with one million
seconds1,2



fs

fs

fs

fs

fs

fs

fs

fs

fs

fs

fs

fs

OpMode

9.0

10.
0

11.
0

12.
0

14.
0

17.
1

9.0

10.0

11.0

12.0

14.0

17.1

0

159

159

159

159

158

159

18010

18010

18010

18010

18010

18010

1

159

159

159

159

159

159

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

2329

185

0

30

114

59

10

1

0

0

5861

1127

239

4

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

9

0

37885

33033

19207

6576

391

0

40

114

65

11

1

0

0

16174

3344

550

0

0

0

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.

295


-------
_BIB.

H fs = 9.00
¦ fs = 10.0
Bfs = 11.0

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 offscaie Value on Coverage and Trends of operating mode Based C02for Heavy Heavy-Duty Vehicles

296


-------
HHDD Opmode Distributions

Comparison of HDIUT vs. MOVES-CTI

¦ HDIUT ¦ MOVES CTI - CY2027/MY2014

n oc i i i

U.JJ

0.30

c 0.25
o

E 0.20

Ll-

CU

o 0.15
E

Q.

o 0.10
0.05

n nn ll











































il



1

-1

1



m













Li



II il 		

ll ¦¦ ¦¦ ¦¦ ¦¦ ¦¦ ¦- ¦_











1 il I

u.uu

0 1

11 12 13 14 15 16

21 22 23 24 25 27 28 29 30
Opmode



33 35

37 38 39 40

Figure G-4 Heavy Heavy-Duty Diesel Operating Mode Distribution compared between the Heavy-Duty In-Use Testing (HDIUT) Program and from M0VES3 for a

MY 2014 vehicle with anfscaie of 10 metric tons


-------
MHDD Opmode Distributions

Comparison of HDIUT vs. MOVES-CTI

¦ HDIUT ¦ MOVES CTI - CY2027/MY2014

n a c.

n /in



	|	

0.35

c

.2 0.30

fo,s

CL>

"E °-20

E

Q- 0.15
O

0.10
0.05

ll









































J





I.



ll ll ¦¦ -¦ -- ¦¦

ll ¦¦ ll !¦ ¦- II __ _ 1





ll .. .

0 1

11 12 13 14 15 16

21 22 23 24 25 27 28 29 30 33
Opmode

35 37 38 39 40

Figure G-5 Medium Heavy-Duty Diesel Operating Mode Distribution compared between the Heavy-Duty In-Use Testing (HDIUT) Program and from M0VES3 for

a MY 2014 vehicle with an fscaie of 7 metric tons

298


-------
0.45
0.40
0.35

c

° 0.30
I 0 25

CD

"8 020

Q- 0.15
O

0.10
0.05
0.00

LHDD Opmode Distributions

Comparison of HDIUT vs. MOVES-CTI

¦ HDIUT ¦ MOVES CTI - CY2027/MY2014

















































!



-1



1 |



ll ll

II

II ll	

la ll ll II ¦¦ I- _

ll II II II ..

1 i 11 12 13 14 15 16 i 21 22 23 24 25 27 28 29 30 ; 33 35 37 38 39 40

Opmode

Figure G-6 Light Heavy-Duty Diesel Operating Mode Distribution compared between the Heavy-Duty In-Use Testing (HDIUT) Program and from MOVES348 for a

MY 2014 vehicle with an fscaie of 5 metric tons.

48 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.

299


-------
H. THC and CO Emission rates from 2010 and Later Model Year Heavy-duty Vehicles from
the HDIUT

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

H.l.l LHD
0.007

0.006

0.005

3
u

0.004 :

I 0.003 :
I-

0.002
0.001

0.000

I LHD FEL 0.20, MY 2010-2013. N=52
I LHD FEL 0.35, N=0
I LHD FEL 0.50, MY 2010-2015,

N=15

j_I_i

B ¦ ¦ 11

iii

i« i I I

iiiili

1



0 1: 11 12 13 14 15 16; 21 22 23 24 25 27 28 29 30, 33 35 37 38 39 40

MOVES OpMode

Figure H-l Average LHD THC Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013 and the 0.5
NOx FEL for MY 2010-2016. Error Bars are 95% Confidence Intervals of the Mean

11

il

I LHD FEL 0.20, MY 2010-2Cjl3, N=52
I LHD FEL 0.35, N=0	.

I LHD FEL 0.50, MY 2010-2(jl5, N=15

II



	I 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 H-2 Average LHD THC Emission Rates by Operating Mode for the 0.2 NOx FEL for MY 2010-2013 and the 0.5
NOx FEL for MY 2010-2016. Error Bars are 95% Confidence Intervals of the Mean

300


-------
H.1.2 MHD

L

*1

	1	I	

I MHD FEL 0.20, MY 2010-2013, N=23

l MHD FEL 0.35, MY 2010-2013, N=23
I MHD FEL 0.50, MY 2010-2013, N=9

_

m

i

Jl

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES OpMode

Figure H-3 Average MHD THC Emission Rates by Operating Mode for the 0.2, 0.35 and 0.50 NOx FEL Groups for MY
2010-2013 Vehicles. Error Bars are 95% Confidence Intervals of the Mean

MOVES OpMode

Figure H-4 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

H.2 Comparison of THC and CO Emission Rates between MY 2010-2013 and MY 2014 in the 0.2 NO„ FEL Group
for LHD and MHD

The following figures show the comparison of the mean THC and CO emission rates between model year
2010-2013 and 2014 and later vehicles in the 0.2 NOx FEL Group for the LHD and MHD regulatory class
by operating mode. The figures for HHD are in the main report (Figure 2-42 and Figure 2-43).

301


-------
H.2.1 LHD
0.0040

0.00B5

—	0.0030

QD

—	0.0025

U

I- 0.0020
0.0015
0.0010
0.0005
0.0000

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES OpMode

Figure H-5 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









































T













ii



i I

T i

I l l

T

II

I 1

1 |



ii ii

.. ii Ii Ii 1

l

ii ii Ii Ii Ii





IL i, i.





¦ MY 2010-2013





¦ MY 2014-2016











































ii ii ii ii ii ii il



T T





i= il

T .

		 ii ii II



ii ii i' i



i

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES OpMode

Figure H-6 CO emission rates for the MY 2010-2013 and MY 2014-2016 vehicles in the LHD 0.20 NOx FEL Group

H.2.2 MHD

302


-------
0.0014

0.0012

< 0.0010
3S

^ 0.0008

t-

0.0006
0.0004
0.0002
0.0000

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 H-7 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
< 0.008

-52

g 0.006
0.004
0.002
0.000

0 1 11 12 13 14 15 16 21 22 23 24 25 27 28 29 30 33 35 37 38 39 40

MOVES OpMode

Figure H-8 CO emission rates for the MY 2010-2013 and MY 2014-2015 vehicles in the MHD 0.20 NOx FEL Group

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 3.1.1.2.2.

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

MY 2010-2013
MY 2014-2015

Juiii

303


-------
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 the following table. The following figure shows the data for
NOx. Interestingly, the trend for NOx from on-road testing is comparable to the trends from previous lab-
based testing, shown in Figure 3-23. For the on-road data, grams of NOx from starts emissions for 105-
minute soak is 1.15 times the 720-minute soak. For the same conditions, the ratios in Figure 3-23 are
approximately 1.17 and 1.37 for the data series labeled as "MOVES" and "New Data", respectively. The
trends for THC and CO are also similar between the two figures.

Table l-l Time and Pollutant Mass for Driving Assigned to Start Emissions

Soak

Time(min)

Number
of Tests

Avg. time1 for TWC to
reach 421 °C (sec)

Avg. grams of pollutant removed

NOx

co2

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

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.

304


-------
3.5

T3
dJ
>
O

E

dJ

X

O

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 1-1 Grams ofNOxfrom Start Emissions versus Soak Time

1.2 Comparison of Heavy-duty Gasoline Emission Rates by Vehicle

The figures in this subsection show the emission rates calculated from the PEMS testing data collected
on three heavy-duty gasoline vehicle. The emission rates in these figures were analyzed using an/sca/eof
5 metric tons used for the LHD2b3 and LHD45 regulatory class emission rates. The error bars are the
95% confidence intervals of the mean calculated by treating the number of routes (R) as independent
random variables.

The Isuzu NPR has the highest NOx emission rates across all operating modes. The Ford E459 and the
Ram 3500 have more similar NOx emission rates, except for the high speed and power operating modes
(opModelD 39 and 40).

305


-------
0.05

0.04

5. 0.03

bO

x
O

0.02

0.01

0.00

I 2016_Ford_E459, R=76
I 2015_lsuzu_NPR, R=83
I 2017_RAM_3500, R=38

il

I

it-

1

L .1. 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 1-2 Mean Heavy-duty Gasoline NOx Emission Rates by Operating Mode and Vehicle Calculated using fscaieof 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.

0.045

0.040 :
0.035
0.030 :

§ °-025 :

^ 0.020
H

0.015 :
0.010 :
0.005
0.000

^ a.

12016_Ford_E459, R=76
i 2015_lsuzu_NPR, R=83
12017 RAM 3500, R=38

11 12 13 14 15 16

k

liii



i l it

21 22 23 24 25 27 28 29 30 ; 33 35 37 3 8 39 40

0

I.

ii. ii« III -	is	gj. L il. ifc ill ill ii

i. i JI.

33 35 37 38 39 40

I 11 12 13 14 15 16 ; 21 22 23 24 25 27 28 29 30

MOVES OpMode

Figure 1-3 Mean Heavy-duty Gasoline THC Emission Rates by Operating Mode and Vehicle Calculated using fscaie of 5

metric tons.

306


-------
9.00
8.00
7.00
6.00

u)

m 5.00

0 4.00
3.00
2.00
1.00
0.00

0 1 ! 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 1-4 Mean Heavy-duty Gasoline CO Emission Rates by Operating Mode and Vehicle Calculated using fscaieOf 5

metric tons.

1.3 Extrapolating High-Power Operating Modes

When analyzing the heavy-duty gasoline data for the MHD and HHD regulatory classes (using an/sCa/eof 7
and 10 metric tons respectively), there was limited or no data for high power operating mode bins. The
figure below shows the mean NOx emission rates by vehicle and operating mode when using an using an
fscaieof 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.

307


-------
O.OB

0.04

"w 0.03
.25

x

O 0.02

0.01

0.00

0 1

12016_Ford_E459, R=80
I 2015_lsuzu_NPR, R=84
I 2017 RAM 3500, R=38

11 12 13 14 15 16

JL

JLj

ll

21 22 23 24 25 27 28 29 30

P- iti

33 35 37 38 39 40

MOVES OpMode

Figure 1-5 Mean Heavy-duty Gasoline NOx Emission Rates by Operating Mode and Vehicle Calculated for HHD using

an fscale of 10 metric tons

In the case of missing data, we used the emission rates of the nearest operating mode bin with data. In
case where there were limited data, we aggregated the averages with data from the next closest bin
into single averages. The figure below displays the resulting emission rates by vehicles for NOx for HHD
vehicles. For the Ford and Isuzu vehicles, the following operating modes were aggregated into single
averages: 15-16, 27-30, 38-40. For the RAM vehicle, operating modes 29-30 were aggregated. The same
aggregation was used to calculated the updated THC and CO emission rates

0.05

0.04

^ 0.03 -|

BO
X

2 0.02 -

0.01 -

0.00

12016_Ford_E459, R=80
I 2015_lsuzu_NPR, R=84
I 2017_RAM_3500, R=38

— Il.ll 11 - J. ii- i

di

¦ ll ll I

ill 11 ll

0 1 ; 11 12 13 14 15 16 ; 21 22 23 24 25 27 28 29 30 , 33 35 37 38 39 40

MOVES OpMode

Figure 1-6 Mean Heavy-duty Gasoline NOx Emission Rates by Operating Mode and Vehicle Calculated for HHD using
an fscale of 10 metric tons with Aggregated Means for High Power Bins with Limited Data

308


-------
The figure below 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

0.020
x

i 0.015
0.010
0.005
0.000

in
00

0 1

IHHDG, MY 2016

11 12 13 14 15 16

- ¦ ¦

21 22 23 24 25 27 28 29 30
MOVES OpMode

jlJL

33 35 37 38 39 40

Figure 1-8 Weighted Average Heavy-duty Gasoline NOx Emission Rates by Operating Mode for HHD Using

Production Volumes

For C02 emission rates, we used a different method than forTHC, 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 1-2.

309


-------
Table 1-2 Assumed STP Midpoint for Each Operating Mode

OpModelD

STP_midpoint

0

-

1

-

11

-

12

1.5

13

4.5

14

7.5

15

10.5

16

13.5

21

-

22

1.5

23

4.5

24

7.5

25

10.5

27

15

28

21

29

27

30

33

33

3

35

9

37

15

38

21

39

27

40

33

The figure below displays the initial mean C02 emission rates using an fscaie of 10 metric tons. Note that
the emission rates for C02 have more consistent trends than other measured pollutants; because of this,
fewer of the high power operating modes were replaced with extrapolated rates than for the THC, CO,
and NOx emission rates. In this case, we decided to only replace one operating mode with limited data
(operating mode 39 for the Isuzu vehicle), because it was only based on one route (which is why there
are no error bars).

310


-------
II II

0 1

I 2016_Ford_E459, R=80
I 2015_lsuzu_NPR, R=84
I 2017_RAM_3500, R=38

	

11 12 13 14 15 16

llL

21 22 23 24 25 27 28 29 30
MOVES OpMode

33 35 37 38 39 40

Figure 1-9 Mean Heavy-duty Gasoline C02 Emission Rates by Operating Mode and Vehicle Calculated for HHD using

an fscale of 10 metric tons

The figure below shows the mean C02 emission rates using an/sctJ;eof 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.



0

I 2016_Ford_E459, R=80
I 2015_lsuzu_NPR, R=84
I 2017_RAM_3500, R=38

LI

J1

Ji

11 12 13 14 15 16

H

11 ll



33 35 37 38 39 40

21 22 23 24 25 27 28 29 30
MOVES OpMode

Figure 1-10 Mean Heavy-duty Gasoline C02 Emission Rates by Operating Mode and Vehicle Calculated for HHD
using an fscale of 10 metric tons with Extrapolated Means for High Power Bins.

The figure below shows the weighted average C02 emission rate calculated by averaging the three
vehicles together according to their production volume sales. Note that the C02 have stronger increasing
trends with power compared to NOx.

311


-------


¦ HHDG, MY 2016



































"\











1























¦

.1

I









. 1 1

















ll









0 1

11

12 13 14 15 1

6

21 22 23 24 25 27 28 29 30

33 35 37 38 39 40

MOVES OpMode

Figure 1-11 Weighted Average Heavy-duty Gasoline C02 Emission Rates by Operating Mode for HHD Using

Production Volumes

Similar calculations were repeated for the MHD vehicles calculated using an/sca/eof 7 metric tons.

J. PM Composition Measurements from Auxiliary Power Units

Table J-l reports the organic carbon (OC), elemental carbon (EC) and total carbon (TC) measurements
conducted in the study conducted by Texas Transportation Institute (TTI, 201495). All the measurements
were collected on APU 1. TTI collected the particulate sample on quartz fiber filters, and Sunset
Laboratory Inc. analyzed the filters using thermal optical reflectance (TOR) using the IMPROVE
(Interagency Monitoring of Protected Visual Environments) procedures. Total Carbon (TC) is the sum of
Elemental Carbon (EC) and Organic Carbon (OC).

312


-------
Table J-l. Organic Carbon, Elemental Carbon, and Total Carbon Measurements from the IMPROVEJTOR measured

on APU1

Sample
ID

Min.

DR

Test

OC

(Hg/cm2)

OC

uncertainty
(Hg/cm2)

EC

(Hg/cm2)

EC

uncertainty
(Hg/cm2)

rc

(Hg/ cm2)

TC

uncertainty
(Hg/ cm2)

EC/TC
ratio

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

313


-------
References

1	89 FR 29440. April 22, 2024. Available online: https://www.govinfo.gov/content/pkg/FR-2024-04-22/pdf/2024-
06809.pdf

2	USEPA (2024). Greenhouse Gas and Energy Consumption Rates for On-road Vehicles in M0VES5. Assessment and
Standards Division. Office of Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, Ml.

3	USEPA (2024). Emission Adjustments for Onroad Vehicles in MOVES5. Office of Transportation and Air Quality. US
Environmental Protection Agency. Ann Arbor, Ml.

4	USEPA (2024). Population and Activity of Onroad Vehicles in MOVES5. Office of Transportation and Air Quality. US
Environmental Protection Agency. Ann Arbor, Ml.

5	89 FR 27842. April 18, 2024. Available online: https://www.govinfo.gov/content/pkg/FR-2024-04-18/pdf/2024-
06214.pdf

6	USEPA (2023). Speciation of Total Organic Gas and Particulate Matter Emissions from Onroad Vehicles in
MOVES4. EPA-420-R-23-006. Office of Transportation and Air Quality. US Environmental Protection Agency. Ann
Arbor, Ml. August 2023. https://www.epa.gov/moves/moves-technical-reports .

7	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, Ml. November 2020.
https://www.epa.gov/moves/moves-technical-reports

8	USEPA (2024). Evaporative Emissions from Onroad Vehicles in MOVES5. Office of Transportation and Air Quality.
US Environmental Protection Agency. Ann Arbor, Ml.

9	USEPA (2024). Brake and Tire Wear Emissions from Onroad Vehicles in MOVES5. Office of Transportation and Air
Quality. US Environmental Protection Agency. Ann Arbor, Ml. November 2020.
https://www.epa.gov/moves/moves-technical-reports.

10	US EPA. Heavy-Duty Highway Compression-Ignition Engines and Urban Buses—Exhaust Emission Standards.
https://www.epa.gov/emission-standards-reference-guide/epa-emission-standards-heavy-duty-highway-engines-
and-vehicles. Accessed August 2023.

11	40 CFR§ 86.091(2).

12	USEPA (2024). Exhaust Emission Rates for Light-Duty Onroad Vehicles in MOVES5. Office of Transportation and
Air Quality. US Environmental Protection Agency. Ann Arbor, Ml.

13. 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/.

14	Jack, Jason A. U.S. Army Aberdeen Test Center Support of Heavy Duty Diesel Engine Emissions Testing. U.S. Army
Aberdeen Test Center CSTE-DTC-AT-SL-E, Aberdeen Proving Ground, Maryland.
(https://gaftp.epa.gov/air/nei/ei_conference/EH5/sessionl/jack.pdf).

15	McClement, Dennis. Reformatting On-Road In-Use Heavy-Duty Emissions Test Data. Sierra Research,

Sacramento, CA. April 2008.

16	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.

17	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.

18	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.

19	"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.

314


-------
20	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.

21	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

22	Data Collection of Drayage Trucks in Houston-Galveston Port Area Draft Report. EP-C-06-080. May 27, 2011

23	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.

24	Rakha, Hesham and Ivana Lucic. Variable Power Vehicle Dynamics Model for Estimating Truck Accelerations.

Page 6.

25	National Renewable Energy Laboratory. Development of LNG-Powered heavy-Duty Trucks in Commercial Hauling.
NREL/SR-540-25154, Golden, CO, December 1998.

26	Goodyear. "Factors Affecting Truck Fuel Economy - Section 9" Page 5.

27	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&ETech Forum2003_Ramsay&Bunker_2. pdf

28	Society of Automotive Engineers. Commercial Truck and Bus SAE Recommended Procedure for Vehicle
Performance Prediction and Charting. SAE J2188. Revised October 2003.

29	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.

30	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.

31	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

32	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.

33	Choi, D., J. Koupal and M. Church (2012). Analysis of Recent Heavy-Duty Vehicle Emission Test Programs. MOVES
Review Workgroup, Ann Arbor, Ml. September 25, 2012.

34	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, Ml. November, 2015. https://www.epa.gov/moves/moves-technical-reports.

35	USEPA (2002). Update of Heavy-Duty Emission Levels (Model Years 1988-2004) for Use in MOBILE6. EPA-420-R-
02-018. July 2002. https://nepis.epa.gov/Exe/ZyPURLcgi?Dockey=P10022RL.txt. (Table 17)

36	USEPA Office of Transportation and Air Quality. Heavy Duty Diesel Engine Consent Decree Low NOx Rebuild
Program Summary. Washington, D.C. https://www.epa.gov/sites/default/files/2015-
01/documents/032807mstrs_lonoxsum4q07.pdf.

37	USEPA (2014). Tier 3 Vehicle Emission and Fuel Standards Program. Regulatory Impact Analysis. EPA-420-R-14-
004. February 2014. https://www.epa.gov/regulations-emissions-vehicles-and-engines/final-rule-control-air-
pollution-motor-vehicles-tier-3

38	2007/2010 Heavy-duty rulemaking. 66 FR 5002, January 18, 2001

39	USEPA (2020). Exhaust Emission Rates for Heavy-Duty On-road Vehicles in MOVES3. EPA-420-R-20-018.
Assessment and Standards Division. Office of Transportation and Air Quality. US Environmental Protection Agency.
Ann Arbor, Ml. November 2020. https://www.epa.gov/moves/moves-onroad-technical-reports.

40	Clark, Nigel et at. California Heavy Heavy-Duty Diesel Truck Emissions Characterization for Program E-55/59.

West Virginia University Research Corporation. Morgantown, WV. November 2005.

41	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.

42	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,

315


-------
Assessment and Standards Division, Ann Arbor, Ml. February 2005.
https://nepis.epa.gov/Exe/ZyPURLcgi?Dockey=P1001D6l.txt

43	Kinsey, J. S., W. A. Mitchell, W. C. Squier, K. Linna, F. G. King, R. Logan, Y. Dong, G. J. Thompson and N. N. Clark
(2006). Evaluation of methods for the determination of diesel-generated fine particulate matter: Physical
characterization results. Journal of Aerosol Science, 37 (1), 63-87. DOI:
http://dx.doi.Org/10.1016/j.jaerosci.2005.03.007.

44	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, Ml. August, 2009. https://nepis.epa.gov/Exe/ZyPURLcgi?Dockey=P10050CK.txt

45	USEPA Office of Transportation and Air Quality. Update Heavy-Duty Engine Emission Conversion Factors for
MOBILE6: Analysis of BSFCs and Calculation of Heavy-Duty Engine Emission Conversion Factors. EPA420-R-02-005,
M6.HDE.004. Assessment and Standards Division, Ann Arbor, Ml.

https://nepis.epa.gov/Exe/ZyPU RL.cgi?Dockey=P10022Ll.txt

46	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.

47	US EPA. 2012. Black Carbon Report to Congress. EPA-450/R-12-001. March 2012.
https://nepis.epa.gov/Exe/ZyPU RL.cgi?Dockey=P100EIJZ.txt

48	Quiros, D. C., A. Thiruvengadam, S. Pradhan, M. Besch, P. Thiruvengadam, B. Demirgok, D. Carder, A. Oshinuga,
T. Huai and S. Hu (2016). Real-World Emissions from Modern Heavy-Duty Diesel, Natural Gas, and Hybrid Diesel
Trucks Operating Along Major California Freight Corridors. Emission Control Science and Technology, 2 (3), 156-
172. DOI: 10.1007/s40825-016-0044-0.

49	Dixit, P., J. W. Miller, D. R. Cocker, A. Oshinuga, Y. Jiang, T. D. Durbin and K. C. Johnson (2017). Differences
between emissions measured in urban driving and certification testing of heavy-duty diesel engines. Atmospheric
Environment, 166, 276-285. DOI:

https://www.sciencedirect.com/science/article/pii/S1352231017304181Pvia%3Dihub.

50	Thiruvengadam, A., M. C. Besch, P. Thiruvengadam, S. Pradhan, D. Carder, H. Kappanna, M. Gautam, A.

Oshinuga, H. Hogo and M. Miyasato (2015). Emission Rates of Regulated Pollutants from Current Technology
Heavy-Duty Diesel and Natural Gas Goods Movement Vehicles. Environ Sci Technol, 49 (8), 5236-5244. DOI:
10.1021/acs.est.5b00943.

5140 CFR § 86.1816-18. Table 1- Fully Phased-in Tier 3 HDV Exhaust Emission Standards.

52	USEPA (2010). Technical Bulletin: Diesel Particulate Filter Operation and Maintenance. EPA-420-F-10-027. Office
of Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, Ml. May 2010.
https://www.epa.gov/sites/default/files/2016-03/documents/420fl0027.pdf

53	Graboski, Michael S., Robert L. McCormick, Janet Yanowitz, and Lisa Ryan. Heavy-Duty Diesel Vehicle Testing for
the Northern Front Range Air Quality Study. Colorado Institute for Fuels and High-Altitude Engine Research,
Colorado School of Mines, Golden, Colorado. Prepared for Colorado State University. February 1998.

54	Energy and Environmental Analysis, Inc. Documentation and Analysis of Heavy-Duty Diesel Vehicle Emission Test
Data. Prepared for New York Department of Environmental Conservation, December 2000.

55	40 CFR § 1065.260

56	40 CFR § 1065.250

57	USEPA (2012). Updates to the Greenhouse Gas and Energy Consumption Rates in MOVES2010a. EPA-420-R-12-
025. Office of Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, Ml. August, 2012.
https://nepis.epa.gov/Exe/ZyPU RL.cgi?Dockey=P100F3Z3.txt.

58	USEPA (2014). Greenhouse Gas Emissions from a Typical Passenger Vehicle. EPA-420-F-14-040a. Office of
Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, Ml. May 2014.
https://nepis.epa.gov/Exe/ZyPU RL.cgi?Dockey=P100LQ99.txt.

59	USEPA (2011). Greenhouse Gas Emission Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty
Engines and Vehicles (76 FR 57106, September 15, 2011)

60	40 Code of Federal Register Volume 76 at 57216 and 57236, September 15, 2011.

61	Greenhouse Gas Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles -
Phase 2 Final Regulatory Impact Analysis, Chapter 5. EPA-420-R-16-900. August 2016

316


-------
62	US EPA. (2024). Docket Memo Heavy-Duty Technology Resource Use Case Scenario Tool (HD TRUCS). Final Rule.
EPA-HQ-OAR-2022-0985-3877. https://www.regulations.gov/document/EPA-HQ-OAR-2022-0985-3877

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	Calcagno, James A. Evaluation of Heavy-Duty Diesel Vehicle Emissions During Cold-Start and Steady-State Idling
Conditions and Reduction of Emissions from a Truck-Stop Electrification Program. PhD Dissertation, Department of
Civil and Environmental Engineering, University of Tennessee, Knoxville, December 2005.

65	40 CFR Part 86, Appendix I (f) (2)

66	Evaluating Technologies and Methods to Lower Nitrogen Oxide Emissions From Heavy-Duty Vehicles. Final
Report. SwRI Project 19503. April, 2017. https://ww2.arb.ca.gov/sites/default/files/classic/research/apr/past/13-
312.pdf

67	EMFAC 2000 Section 6.7 START CORRECTION FACTORS. 4/20/2000. On-Road Emissions Model Methodology
Documentation.

68	Glover, E.; Carey, P. Determination of Start Emissions as a Function of Mileage and Soak Time for 1981-1993
Model-year Light-Duty Vehicles. EPA420-R-01-058 (M6.STE.003). USEPA Office of Transportation and Air Quality,
Ann Arbor, Ml. November, 2001.

69	USEPA (2017). Heavy-Duty Chassis Start Emissions Testing. Office of Transportation and Air Quality. US
Environmental Protection Agency, Ann Arbor, Ml. 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, Ml. 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, Ml. March, 2005.
https://nepis.epa.gov/Exe/ZyPU RL.cgi?Dockey=P1001DAQ.txt.

72	USEPA (2022). Control of Air Pollution from New Motor Vehicles: Heavy-Duty Engine and Vehicle Standards,
Draft Regulatory Impact Analysis. EPA-420-D-22-001. Office of Transportation and Air Quality. US Environmental
Protection Agency, Ann Arbor, Ml., March 2022. https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P10144K0.pdf

73	Keel-Blackmon, K., S. Curran and M. V. Lapsa (2016). Summary of OEM Idling Recommendations from Vehicle
Owners Manuals. Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States). Fuels, Engines and
Emissions Research Center; National Transportation Research Center (NTRC).
https://info.ornl.gov/sites/publications/Files/Pub61263.pdf.

74McCormick, Robert, M. S. Graboski, T. L. Alleman , J. Yanowitz. Idle Emissions from Heavy-Duty Diesel and Natural
Gas Vehicles at High Altitude. Journal of the Air and Waste Management Association, 50(ll):1992-8. Revised May
3, 2000.

75Lim, Han. Study of Exhaust Emissions from Idling Heavy-duty Diesel Trucks and Commercially Available Idle-
Reducing Devices. EPA420-R-02-025. US EPA Office of Transportation and Air Quality, Certification and Compliance
Division. October 2002.

76 Irick, David K. and Bob Wilson. NOx Emissions and Fuel Consumption ofHDDVs during Extended Idle. University
of Tennessee, IdleAire Technologies Inc. In: Proceedings, Coordinated Research Council 12th Annual On-Road
Vehicle Emission Workshop, San Diego, California, April 15-17, 2002.

77Lambert, Douglas, et al. Roadside Emissions Study: Preliminary Results for Stationary and On-Road Testing of
Diesel Trucks in Tulare, California. California Environmental Protection Agency. Air Resources Board. Mobile Source
Operations Division, Clean Air Technologies International, Inc May 15, 2002.

78Gautam, Mridual and Nigel N. Clark. Heavy-duty Vehicle Chassis Dynamometer Testing for Emissions Inventory,
Air Quality Modeling, Source Apportionment and Air Toxics Emissions Inventory. Phase I Interim Report, CRC
Project No. E-55/E-59, West Virginia University Research Corporation, Morgantown, July 2002.

79National Cooperative Highway Research Program (NCHRP). Heavy-duty Vehicle Emissions. NCHRP Project 25-14,
Cambridge Systematics, Inc., Battelle Laboratories, Sierra Research and West Virginia University. October 2002.
80Tang, Shida and John Munn. Internal Report- Idle Emissions from Heavy-Duty Diesel Trucks in the New York
Metropolitan Area. New York State Dept of Environmental Conservation, November 9, 2001.

81Brodrick, Dwyer. Potential Benefits of Utilizing Fuel Cell Auxiliary Power Units in Lieu of Heavy-Duty Truck Engine
Idling. Paper UCD-ITS-REP-01-01. Institute of Transportation Studies, University of California, Davis, 2001.

317


-------
82Storey, John M.E., John F. Thomas, Samuel A. Lewis, Sr., Thang Q. Dam, K. Edwards, Dean, Gerald L. DeVault, and
Dominic J. Retrossa. Particulate Matter and Aldehyde Emissions from Idling Heavy-Duty Diesel Trucks. SAE Paper
2003-01-0289. Society of Automotive Engineers, Warrendale, PA.

83	Keel-Blackmon, K., S. Curran and M. V. Lapsa (2016). Summary of OEM Idling Recommendations from Vehicle
Owners Manuals. Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States). Fuels, Engines and
Emissions Research Center; National Transportation Research Center (NTRC).
https://info.ornl.gov/sites/publications/Files/Pub61263.pdf.

84	Lutsey, N., Brodrick, C-J., Sperling, D., Oglesby, C., Transportation Research Record: Journal of the Transportation
Research Board, No 1880, TRB, National Research Council, Washington, D.C., 2004, pp. 28-38

85	Hoekzema, A. (2015). Modeling Truck Idling Emissions in Central Texas. 6800 Burleson Road, Building 310, Suite
165 Austin, Texas 78744, Capital Area Council of Governments.

86	Khan, A. S., N. N. Clark, M. Gautam, W. S. Wayne, G. J. Thompson and D. W. Lyons (2009). Idle Emissions from
Medium Heavy-Duty Diesel and Gasoline Trucks. Journal of the Air & Waste Management Association, 59 (3), 354-
359.

87	Farzaneh, M., J. Zietsman, D.-W. Lee, J. Johnson, N. Wood, T. Ramani and C. Gu (2014). TEXAS-SPECIFIC DRIVE
CYCLES AND IDLE EMISSIONS RATES FOR USING WITH EPA'S MOVES MODEL. FHWA/TX-14/0-6629-1. Texas A&M
Transportation Institute. May, 2014. http://tti.tamu.edu/documents/0-6629-l.pdf.

88	ARB (2015). EMFAC2014 Volume III - Technical Documentation. California Environmental Protection Agency, Air
Resources Board, Mobile Source Analysis Branch, Air Quality Planning & Science Division. May 12, 2015.

89	California Environmental Protection Agency. Heavy-Duty Engines and Vehicles Executive Orders. Motor Vehicle
and Engine Certification Program. Air Resources Board. Available at: https://ww2.arb.ca.gov/new-vehicle-and-
engine-certification-executive-orders

90	US EPA, Navistar Inc. Heavy-Duty Engine Recall. EPA-420-F-13-038, Ann Arbor, Ml: June 2013,
https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100GNLJ.PDF

91	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.

92	US EPA, Engine Certification Data. On-Highway Heavy Duty - Diesel and Gasoline (2012).
https://19january2017snapshot.epa.gov/compliance-and-fuel-economy-data/engine-certification-data_.html

93	Khalek, I. A., M. G. Blanks, P. M. Merritt and B. Zielinska (2015). Regulated and unregulated emissions from
modern 2010 emissions-compliant heavy-duty on-highway diesel engines. Journal of the Air & Waste Management
Association, 65 (8), 987-1001. DOI: 10.1080/10962247.2015.1051606.

94	Khan, A. S., N. N. Clark, M. Gautam, W. S. Wayne, G. J. Thompson and D. W. Lyons (2009). Idle Emissions from
Medium Heavy-Duty Diesel and Gasoline Trucks. Journal of the Air & Waste Management Association, 59 (3), 354-
359. DOI: 10.3155/1047-3289.59.3.354.

95	Zietsman, J. and J. Johnson (2014). Auxiliary Power Unit Testing for SmartWay Idle Reduction Verification. DRAFT
FOR REVIEW. EP-ll-H-000527, Auxiliary Power Unit Testing for SmartWay Idle Reduction Verification. Texas A&M
Transportation Institute. August, 2014.

96	Frey, H. C. and P.-Y. Kuo (2009). Real-World Energy Use and Emission Rates for Idling Long-Haul Trucks and
Selected Idle Reduction Technologies. Journal of the Air & Waste Management Association, 59 (7), 857-864. DOI:
10.3155/1047-3289.59.7.857.

97	TTI (2012). Development of a NOx Verification Protocol and Actual Testing of Onboard Idle Reduction
Technologies. New Technology Research and Development Program. Texas Transportation Institute. Revised:
January 2012.

98	Storey, J. M., J. F. Thomas, S. A. Lewis, T. Q. Dam, K. D. Edwards, G. L. DeVault and D. J. Retrossa (2003).
Particulate matter and aldehyde emissions from idling heavy-duty diesel trucks. SAE Technical Paper.

99	See 40 CFR 1037.106(g).

100	US EPA, Nonroad Compression-Ignition Engines: Exhaust Emission Standards, EPA-420-B-16-022, March 2015.
https://nepis.epa.gov/Exe/ZyPU RL.cgi?Dockey=P1000A05.txt

101	US EPA. Memo to Docket: Updates to MOVES for Emissions Analysis of Greenhouse Gas Emissions and Fuel
Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles - Phase 2 FRM. August 8, 2016. Pages 33-
35. Docket EPA-HQ-OAR-2014-0827-2227.

102	US EPA, Frequently Asked Questions about Heavy-Duty "Glider Vehicles" and "Glider Kits", EPA-420-F-15-904,
Ann Arbor, Ml: July 2015, https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100MUVI.PDF

318


-------
103	US EPA, Chassis Dynamometer Testing of Two Recent Model Year Heavy-Duty On-Highway Diesel Glider
Vehicles, EPA-HQ-OAR-2014-0827-2417, Ann Arbor, Ml, November 2017,

https://www. regulations. gov/document?D=EPA-HQ-OAR-2014-0827-2417

104	USEPA (2008). Mobile Source Observation Database (MSOD): User Guide and Reference. EPA420-B-08-017.
Office of Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, Ml. December, 2008.
https://nepis.epa.gov/Exe/ZyPU RL.cgi?Dockey=P10028RC.txt

105	EPA. 1999. Tier 2 Vehicle & Gasoline Sulfur Program Final Rule, https://www.epa.gov/regulations-emissions-
vehicles-and-engines/final-rule-control-air-pollution-new-motor-vehicles-tier

106	USEPA (2017). Light-Duty Vehicles, Light-Duty Trucks, and Medium-Duty Passenger Vehicles: Tier 2 Exhaust
Emission Standards and Implementation Schedule EPA-420-B-17-028. Office of Transportation and Air Quality. US
Environmental Protection Agency. September 2017.

https://nepis.epa.gov/Exe/ZyPDF.cgi/P100SMQA.PDF?Dockey=P100SMQA.PDF.

107	Development of Emission Rates for Heavy-Duty Vehicles in the Motor Vehicle Emissions Simulator MOVES2010
(131 pp, EPA-420-B-12-049, August 2012) https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100F80L.txt

108	EPA. 2001. "Control of Air Pollution From New Motor Vehicles: Heavy-Duty Engine and Vehicle Standards and
Highway Diesel Fuel Sulfur Control Requirements; Final Rule" https://www.epa.gov/regulations-emissions-
vehicles-and-engines/final-rule-control-air-pollution-new-motor-vehicles109 USEPA (2015). Exhaust Emission Rates
for Heavy-Duty On-road Vehicles in MOVES2014. EPA-420-R-15-015a. Assessment and Standards Division. Office
of Transportation and Air Quality. US Environmental Protection Agency. Ann Arbor, Ml. November, 2015.
https://www.epa.gov/moves/moves-technical-reports.

110	USEPA (2022) Final Rule and Related Materials for Control of Air Pollution from New Motor Vehicles: Heavy-
Duty Engine and Vehicle Standards, Regulatory Impact Analysis. EPA-420-R-22-035, December 2022. Chapter 3.2.

111	USEPA. Final Rulemaking to Establish Greenhouse Gas Emission Standards and Fuel Efficiency Standards for
Medium- and Heavy-Duty Engines and Vehicles: Regulatory Impact Analysis. EPA-420-R-13-901, August 2011. Page
2-30.

112	USEPA. Final Rulemaking to Establish Greenhouse Gas Emission Standards and Fuel Efficiency Standards for
Medium- and Heavy-Duty Engines and Vehicles - Phase 2: Regulatory Impact Analysis. EPA-420-R-16-900, August
2016. Page 5-14.

113	USEPA (2016). Heavy-Duty Highway Spark-Ignition Engines: Exhaust Emission Standards. EPA-420-B-16-019.
Office of Transportation and Air Quality. US Environmental Protection Agency. March 2016.
https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P1000A01.pdf.

114	Toro, C., J. Warila, D. Sonntag, D. Choi and M. Beardsley (2019). Updates to "high-power" emission rates and
start deterioration for light-duty vehicles MOVES Review Workgroup, Ann Arbor, Ml. April 10, 2019
https://www.epa.gov/sites/production/files/2019-06/documents/03-updates-ld-emission-rates-start-
deterioration-2019-04-10.pdf.

115	USEPA (2022). Control of Air Pollution from New Motor Vehicles: Heavy-Duty Engine and Vehicle Standards,
Draft Regulatory Impact Analysis. EPA-420-D-22-001. Office of Transportation and Air Quality. US Environmental
Protection Agency, Ann Arbor, Ml., March 2022. https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P10144K0.pdf

116	USEPA (2015). Exhaust Emission Rates for Light-Duty On-road Vehicles in MOVES2014. EPA-420-R-15-005.
Assessment and Standards Division. Office of Transportation and Air Quality. US Environmental Protection Agency.
Ann Arbor, Ml. October, 2015. https://www.epa.gOv/moves/moves-onroad-technical-reports#moves2014.

117	40 CFR Part 86 Appendix 1(f)(1)

118	Boyce, B. 2014. Cummins Westport - Heavy Duty Natural Gas Engines for Trucks and Buses, presented at the
Southeast Alternative Fuels Conference & Expo, October 22, Raleigh, NC, USA.

119	Ayala, A., Gebel, M., Okamoto, R., Rieger, P. et al., "Oxidation Catalyst Effect on CNG Transit Bus Emissions,"
Society of Automotive Engineers, SAE Technical Paper 2003-01-1900, 2003. (http://papers.sae.org/2003-01-1900).

120	"Central Business District (CBD)," Emission Test Cycles. DieselNet. Last Updated: September 2000, Accessed:
August 2023. (http://www.dieselnet.com/standards/cycles/cbd.php).

121	DieselNet, (http://www.dieselnet.com/standards/cycles/cbd.php).

122	Hesterberg, T.; Lapin, C.; Bunn, W.; 2008. "A Comparison of Emissions from Vehicles Fueled with Diesel or
Compressed Natural Gas." Environ. Sci. Technol. 42(17): 6437-6445.
(http://pubs.acs.org/doi/abs/10.1021/es071718i).

319


-------
123	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).

124	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).

125	Ayala, A., Kado, N., Okamoto, R., Holmen, B. et al., "Diesel and CNG Heavy-duty Transit Bus Emissions over
Multiple Driving Schedules: Regulated Pollutants and Project Overview," Society of Automotive Engineers, SAE
Technical Paper 2002-01-1722, 2002. (http://papers.sae.org/2002-01-1722).

126	Lanni, T., Frank, B., Tang, S., Rosenblatt, D. et al., "Performance and Emissions Evaluation of Compressed
Natural Gas and Clean Diesel Buses at New York City's Metropolitan Transit Authority," Society of Automotive
Engineers., SAE Technical Paper 2003-01-0300, 2003. (http://papers.sae.org/2003-01-0300).

127	McKain, D., Clark, N., Balon, T., Moynihan, P. et al., "Characterization of Emissions from Hybrid-Electric and
Conventional Transit Buses," Society of Automotive Engineers, SAE Technical Paper 2000-01-2011, 2000.
(http://papers.sae.org/2000-01-2011).

128	Clark, N., Gautam, M., Lyons, D., Bata, R. et al., "Natural Gas and Diesel Transit Bus Emissions: Review and
Recent Data," Society of Automotive Engineers, SAE Technical Paper 973203,1997.

(http://papers.sae.org/973203).

129	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).

130	Clark, et al., "Effects of Average Driving Cycle Speed on Lean-Burn Natural Gas Bus Emissions and Fuel
Economy," SAE Technical Paper 2007-01-0054, 2007

131	"Heavy-Duty Highway Compression-Ignition Engines and Urban Buses -- Exhaust Emission Standards," Emission
Standards Reference Guide, USEPA. Last Updated: 5 July 2012, Accessed: 20 July 2012.

132	US EPA OTAQ, 2012. (http://iaspub.epa.gov/otaqpub/pubsearch.jsp).

133	40 CFR 86.105-94. "Clean-fuel fleet emission standards for heavy-duty engines." Code of Federal Regulations.

134	Search for compliance documents in the "Heavy-Duty Highway Compression-Ignited Engines" category on EPA's
Transportation and Air Quality Document Index System (DIS), Last Updated: 10 August 2012, Accessed: 10 August
2012. (http://iaspub.epa.gov/otaqpub/pubsearch.jsp).

135	Hajbabaei, M., G. Karavalakis, K. C. Johnson, L. Lee and T. D. Durbin (2013). Impact of natural gas fuel
composition on criteria, toxic, and particle emissions from transit buses equipped with lean burn and
stoichiometric engines. Energy, 62 (0), 425-434. DOI: http://dx.doi.Org/10.1016/j.energy.2013.09.040.

136	Yoon, S.; Collins, J.; Thiruvengadam, A.; Gautam, M.; Herner, J.; Ayala, A. Criteria pollutant and greenhouse gas
emissions from CNG transit buses equipped with three-way catalysts compared to leanburn engines and oxidation
catalyst technologies, Journal of the Air & Waste Management Association, 2013, 63:8, 926-933,
http://dx.doi.org/10.1080/10962247.2013.800170.

137	Jeon, J., J. T. Lee and S. Park (2016). Nitrogen Compounds (NO, N02, N20, and NH3) in NOx Emissions from
Commercial EURO VI Type Heavy-Duty Diesel Engines with a Urea-Selective Catalytic Reduction System. Energy &
Fuels, 30 (8), 6828-6834. DOI: 10.1021/acs.energyfuels.6b01331.

138	Majewski, W. A. (2005). Selective catalytic reduction. Ecopoint Inc. Revision.

139	Preble, C. V., R. A. Harley and T. W. Kirchstetter (2019). Control Technology-Driven Changes to In-Use Heavy-
Duty Diesel Truck Emissions of Nitrogenous Species and Related Environmental Impacts. Environ Sci Technol, 53
(24), 14568-14576. DOI: 10.1021/acs.est.9b04763.

140	Preble, C. V., T. R. Dallmann, N. M. Kreisberg, S. V. Hering, R. A. Harley and T. W. Kirchstetter (2015). Effects of
Particle Filters and Selective Catalytic Reduction on Heavy-Duty Diesel Drayage Truck Emissions at the Port of
Oakland. Environ Sci Technol, 49 (14), 8864-8871. DOI: 10.1021/acs.est.5b01117.

141	Kean, A. J., D. Littlejohn, G. A. Ban-Weiss, R. A. Harley, T. W. Kirchstetter and M. M. Lunden (2009). Trends in on-
road vehicle emissions of ammonia. Atmospheric Environment, 43 (8), 1565-1570. DOI:
10.1016/j.atmosenv.2008.09.085.

320


-------
142	Haugen, M. J., G. A. Bishop, A. Thiruvengadam and D. K. Carder (2018). Evaluation of Heavy- and Medium-Duty
On-Road Vehicle Emissions in California's South Coast Air Basin. Environ Sci Technol, 52 (22), 13298-13305. DOI:
10.1021/acs.est.8b03994.

143	Bishop, G. A., M. J. Haugen, B. C. McDonald and A. M. Boies (2022). Utah Wintertime Measurements of Heavy-
Duty Vehicle Nitrogen Oxide Emission Factors. Environ Sci Technol, 56 (3), 1885-1893. DOI:
10.1021/acs.est.lc06428.

144	Wang, X., A. Khlystov, K.-F. Ho, D. Campbell, J. C. Chow, S. D. Kohl, J. G. Watson, S.-c. F. Lee, L.-W. A. Chen, M. Lu
and S. S. H. Ho (2019). Real-World Vehicle Emissions Characterization for the Shing Mun Tunnel in Hong Kong and
Fort McHenry Tunnel in the United States. Research Report 199. Health Effects Institute. Boston, MA. March 2019.
https://www.healtheffects.org/publication/real-world-vehicle-emissions-characterization-shing-mun-tunnel-hong-
kong-and-fort.

145	USEPA (2012). Use of Data from "Development of Emission Rates for the MOVES Model," Sierra Research, March
3, 2010. EPA-420-R-12-022. Office of Transportation and Air Quality. US Environmental Protection Agency. Ann
Arbor, Ml. August, 2012. https://nepis.epa.gov/Exe/ZyPURLcgi?Dockey=P100FlA5.txt.

146	Harvey, C. A., R. J. Garbe, T. M. Baines, J. H. Somers, K. H. Hellman and P. M. Carey (1983). A study of the
potential impact of some unregulated motor vehicle emissions. SAE transactions, 280-289.

147	Livingston, C., P. Rieger and A. Winer (2009). Ammonia emissions from a representative in-use fleet of light and
medium-duty vehicles in the California South Coast Air Basin. Atmospheric Environment, 43 (21), 3326-3333. DOI:
10.1016/j.atmosenv.2009.04.009.

148	Thiruvengadam, A., M. Besch, D. Carder, A. Oshinuga, R. Pasek, H. Hogo and M. Gautam (2016). Unregulated
greenhouse gas and ammonia emissions from current technology heavy-duty vehicles. Journal of the Air & Waste
Management Association, 66 (11), 1045-1060. DOI: 10.1080/10962247.2016.1158751.

149	Zhu, H., C. McCaffery, J. Yang, C. Li, G. Karavalakis, K. C. Johnson and T. D. Durbin (2020). Characterizing
emission rates of regulated and unregulated pollutants from two ultra-low NOx CNG heavy-duty vehicles. Fuel,
277,118192. DOI: https://doi.Org/10.1016/j.fuel.2020.118192.

150	Zhu, H., Ma, T., Toumasatos, Z., Cao, S., Karavalakis, G., Johnson, K. C., & Durbin, T. (2024). On-road NOx and
NH3 emissions measurements from in-use heavy-duty diesel and natural gas trucks in the South Coast air Basin of
California. Atmospheric Environment, 316,120179.

151	U.S. EPA. "Emission Rate and activity updates for MOVES3.R1". October 30, 2023. Available online:
https://cfpub.epa.gov/si/si_public_record_Report.cfm?Lab=OTAQ&dirEntryld=356887

152	Karavalakis, G., Hajbabaei, M., Durbin, T. D., Johnson, K. C., Zheng, Z., & Miller, W. J. (2013). The effect of
natural gas composition on the regulated emissions, gaseous toxic pollutants, and ultrafine particle number
emissions from a refuse hauler vehicle. Energy, 50, 280-291.

153	U.S. EPA. 2024. Peer Review of MOVES5.

https://cfpub. epa.gov/si/si_public_record_report.cfm?Lab=OTAQ&dirEntryld=361938

154	Karavalakis, G., Jiang, Y., Yang, J., Hajbabaei, M. et al., "Gaseous and Particulate Emissions from a Waste Hauler
Equipped with a Stoichiometric Natural Gas Engine on Different Fuel Compositions," SAE Technical Paper 2016-01-
0799, 2016, doi: 10.4271/2016-01-0799

155	Karavalakis, G., Hajbabaei, M., Jiang, Y., Yang, J., Johnson, K. C., Cocker, D. R., & Durbin, T. D. (2016). Regulated,
greenhouse gas, and particulate emissions from lean-burn and stoichiometric natural gas heavy-duty vehicles on
different fuel compositions. Fuel, 175,146-156.

156	In-Use Emissions Testing and Activity Profiles for On-Road Heavy-Duty Vehicles. March 2023. CEC-500-2023-
002. Available online: https://www.energy.ca.gov/sites/default/files/2023-03/CEC-500-2023-002.pdf

157	Jaaskelainen, H. Crankcase Ventilation. DieselNet Technology Guide. www.DieselNet.com. Copyright © Ecopoint
Inc. Revision 2012.12.

158	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.

159	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

160	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.

321


-------
161	Title 40: Code of Federal Regulations. Part 86- Protection of Environment. Control of Emissions from New and
In-Use Highway Vehicles and Engines. 86.1810-01 Subpart S—General Compliance Provisions for Control of Air
Pollution From New and In-Use Light-Duty Vehicles, Light-Duty Trucks, and Complete Otto-Cycle Heavy-Duty
Vehicles. General standards; increase in emissions; unsafe conditions; waivers.

162	40 CFR 86.004-11. "Control of Emissions from New and In-Use Highway Vehicles and Engines." Code of Federal
Regulations.

163	Khalek, I. A.; Bougher, T. L; Merrit, P. M.; Phase 1 of the Advanced Collaborative Emissions Study. CRC Report:
ACES Phase 1, June 2009.

164	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.

165	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.

166	Hare, C. T.; Baines, T. M.; Characterization of Diesel Crankcase Emissions. Society of Automotive Engineers, Off-
Highway Vehicle Meeting and Exhibition. MECA, Milwaukee. 1977.

167	Kalayci, Veli. "Spiracle™ Crankcase Filtration Systems: Technical Article". Donaldson Company, Inc. January
2011.

168	{USEPA, 2020 #3833}

169	Gerhardt, M. J., D. Sonntag, G. Brown, B. Caldwell, A. Cullen, C. Hart and S. Ludlam (2020). Crankcase Emissions
for MY2007+ Heavy-Duty Diesel Trucks. MOVES Review Workgroup, Ann Arbor, Ml. October 14, 2020.
https://www.epa.gov/moves/october-2020-moves-model-review-work-group-meeting-materials.

170	Clark, N. N., et al. (2017). Pump-to-Wheels Methane Emissions from the Heavy-Duty Transportation Sector.
Environ Sci Technol, 51 (2), 968-976. DOI: 10.1021/acs.est.5b06059.

171	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.

172	Seinfeld, J. H. and S. N. Pandis (2012). Atmospheric chemistry and physics: from air pollution to climate change,
John Wiley & Sons.

173	Dunlea, E. J., S. C. Herndon, D. D. Nelson, R. M. Volkamer, F. San Martini, P. M. Sheehy, M. S. Zahniser, J. H.
Shorter, J. C. Wormhoudt, B. K. Lamb, E. J. Allwine, J. S. Gaffney, N. A. Marley, M. Grutter, C. Marquez, S. Blanco, B.
Cardenas, A. Retama, C. R. Ramos Villegas, C. E. Kolb, L. T. Molina and M. J. Molina (2007). Evaluation of nitrogen
dioxide chemiluminescence monitors in a polluted urban environment. Atmos. Chem. Phys., 1 (10), 2691-2704.
DOI: 10.5194/acp-7-2691-2007.

174	Kurtenbach, R., K. H. Becker, J. A. G. Gomes, J. Kleffmann, J. C. Lorzer, M. Spittler, P. Wiesen, R. Ackermann, A.
Geyer and U. Piatt (2001). Investigations of emissions and heterogeneous formation of HONO in a road traffic
tunnel. Atmospheric Environment, 35 (20), 3385-3394. DOI: http://dx.doi.org/10.1016/S1352-2310(01)00138-8.

175	Khalek, I. A., M. G. Blanks and P. M. Merritt (2013). Phase 2 of the Advanced Collaborative Emissions Study.
CRC Report: ACES Phase 2. Coordinating Research Council, Inc. & Health Effects Institute. November 2013.

176	Kim, J. H., M. Y. Kim and H. G. Kim (2010). N02-Assisted Soot Regeneration Behavior in a Diesel Particulate Filter
with Heavy-Duty Diesel Exhaust Gases. Numerical Heat Transfer, Part A: Applications, 58 (9), 725-739. DOI:
10.1080/10407782.2010.523293.

177	Kramer, L. J., L. R. Crilley, T. J. Adams, S. M. Ball, F. D. Pope and W. J. Bloss (2020). Nitrous acid (HONO)
emissions under real-world driving conditions from vehicles in a UK road tunnel. Atmos. Chem. Phys., 20 (9), 5231-
5248. DOI: 10.5194/acp-20-5231-2020.

178	Liang, Y., Q. Zha, W. Wang, L. Cui, K. H. Lui, K. F. Ho, Z. Wang, S.-c. Lee and T. Wang (2017). Revisiting nitrous
acid (HONO) emission from on-road vehicles: A tunnel study with a mixed fleet. Journal of the Air & Waste
Management Association, 67 (7), 797-805. DOI: 10.1080/10962247.2017.1293573.

179	Xu, Z., T. Wang, J. Wu, L. Xue, J. Chan, Q. Zha, S. Zhou, P. K. K. Louie and C. W. Y. Luk (2015). Nitrous acid
(HONO) in a polluted subtropical atmosphere: Seasonal variability, direct vehicle emissions and heterogeneous
production at ground surface. Atmospheric Environment, 106,100-109. DOI:
https://doi.Org/10.1016/j.atmosenv.2015.01.061.

180	Trinh, H. T., K. Imanishi, T. Morikawa, H. Hagino and N. Takenaka (2017). Gaseous nitrous acid (HONO) and
nitrogen oxides (NOx) emission from gasoline and diesel vehicles under real-world driving test cycles. Journal of the
Air & Waste Management Association, 67 (4), 412-420. DOI: 10.1080/10962247.2016.1240726.

322


-------
181	Rappengliick, B., G. Lubertino, S. Alvarez, J. Golovko, B. Czader and L. Ackermann (2013). Radical precursors and
related species from traffic as observed and modeled at an urban highway junction. Journal of the Air & Waste
Management Association, 63 (11), 1270-1286. DOI: 10.1080/10962247.2013.822438.

182	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, Ml. April
2012. https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100FlA5.pdf

183	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, Ml. August, 2009. https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P10050CK.txt.

184	Belser, Evan (2020) Re: Tampered Diesel Pickup Trucks: A Review of Aggregated Evidence from EPA Civil
Enforcement Investigations. Letter to Jason E. Sloan, et al, November 20, 2020

https://www.epa.gov/sites/default/files/2021-01/documents/epaaedletterreportontampereddieselpickups.pdf

185	U.S. Census Bureau. Vehicle Inventory and Use Survey. Service Sector Statistics Division, Transportation
Characteristics Branch, https://www.census.gov/programs-surveys/vius.html.

186	Zhou, Lei. Revision of Heavy Heavy-Duty Diesel Truck Emission Factors and Speed Correction Factors. California
Air Resources Board, Sacramento. October 2006.

187	USEPA (2022) Final Rule and Related Materials for Control of Air Pollution from New Motor Vehicles: Heavy-
Duty Engine and Vehicle Standards, Regulatory Impact Analysis. EPA-420-R-22-035, December 2022. Chapter 1.

188	Illinois Environmental Protection Agency. Effectiveness of On-Board Diagnostic l/M Testing: Report to the
General Assembly, Response to Public Act 92-0682. Bureau of Air, Springfield, IL. September 2003. Page 21.

189	Manufacturers of Emission Controls Association. Technology Details - Catalytic Converters - SCR System.

190	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.

191	Darlington, T., Dennis Kahlbaum and Gregory Thompson. On-Road NOx Emission Rates from 1994-2003 Heavy-
Duty Diesel Trucks. SAE 2008-01-1299. Society of Automotive Engineers, April 2008.

192	Preble, C. V., T. R. Dallmann, N. M. Kreisberg, S. V. Hering, R. A. Harley and T. W. Kirchstetter (2015). Effects of
Particle Filters and Selective Catalytic Reduction on Heavy-Duty Diesel Drayage Truck Emissions at the Port of
Oakland. Environ Sci Technol, 49 (14), 8864-8871. DOI: 10.1021/acs.est.5b01117.

193	Bishop, G. A., R. Hottor-Raguindin, D. H. Stedman, P. McClintock, E. Theobald, J. D. Johnson, D.-W. Lee, J.
Zietsman and C. Misra (2015). On-road Heavy-duty Vehicle Emissions Monitoring System. Environ Sci Technol, 49
(3), 1639-1645. DOI: 10.1021/es505534e.

194	ARB (2015). Evaluation of Particulate Matter Filters in On-Road Heavy-Duty Diesel Vehicle Applications.
California Air Resources Board. May 8, 2015. https://ww2.arb.ca.gov/sites/default/files/2020-08/dpfeval_0.pdf.

323


-------